Search

Solvent degradation represents one of the main operational drawbacks of the post-combustion CO2 capture process. Degradation not only induces additional costs for solvent make-up, it also impacts the process efficiency and its environmental penalty due to the emission of various degradation products. There is still a gap of knowledge about the influence of process operating conditions on degradation, making it currently impossible to predict the solvent degradation rate in CO2 capture plants. Morever, the reaction mechanisms corresponding to solvent degradation are very slow, significantly complicating its study in industrial units. In the present work, appropriate experimental equipment and analytical methods are developed for accelerating the degradation of monoethanolamine solvents (MEA). The relevance of accelerated conditions is established by comparing artificially degraded solvent samples with degraded solvent samples from industrial CO2 capture pilot plants. Two approaches are evaluated implying either discontinuous or continuous gas feed, this latest being the most representative of industrial degradation. The respective influences of the gas feed composition and the gas-liquid transfer are evidenced and quantified. Finally, the present study leads to a better understanding of solvent degradation in the CO2 capture process with MEA. More generally, it also evidences that accelerated conditions at laboratory-scale may provide relevant information for the study of slow phenomena taking place in large-scale industrial processes. Further works include the development of a kinetic model for MEA solvent degradation and the extension of this methodology to other promising solvents in order to facilitate the operation and large-scale deployment of CO2 capture.

In the present work, the influence of metal ions and oxidative degradation inhibitors on the stability of monoethanolamine solvents (MEA) is studied. Solvent degradation induces additional costs and impacts the environmental balance of the CO2 capture process as well as its efficiency. The two main degradation pathways of MEA are studied under accelerated conditions: oxidative degradation with continuous gas feed and thermal degradation in batch reactors. It is confirmed that metal ions (resulting from solvent impurities and wall leaching) enhance the oxidative degradation of MEA, while they do not impact its thermal degradation. Moreover, different oxidative degradation inhibitors are tested with varying results according to the inhibitor. It appears that at the selected concentration, radical scavengers like Inhibitor A and DMTD (2,5-dimercapto-1,3,4-thiadiazole) are more efficient than chelating agents like HEDP (1-hydroxyethylidene diphosphonic acid) at inhibiting oxidative degradation. Furthermore, a...

In the present work, a kinetic model is proposed for the prediction of amine solvent degradation in the post-combustion CO 2 capture process. Solvent degradation combined to the emission of degradation products represents one of the main operational drawbacks of this process. It induces additional costs and it impacts the process efficiency and its environmental balance. In the present work, degradation is studied under accelerated conditions for the case of monoethanolamine solvent (MEA). The influence of the temperature and of the O 2 and CO 2 concentrations in the gas feed are studied, and their effect on the MEA loss and the emission of degradation products is quantified. Based on the experimental results, a kinetic model for both oxidative and thermal degradation of MEA is proposed and compared to previous attempts to model MEA degradation. The present kinetic model may be further used to develop a practical tool assessing solvent degradation in large-scale capture plants.

The present paper addresses the evolution and perspectives in the teaching of CAPE methods in the Department of Chemical Engineering at the University of Liege. The transition that happened in the 90ies with the arrival of commercial software is highlighted, as the learning outcomes evolved from the ability of building programs to solve chemical engineering problems towards the ability to use complex commercial software and to understand their limitations. Moreover, CAPE methods were extended to non-dedicated CAPE courses, which is illustrated here by the goals and challenges of their use in courses like “Reactor Engineering” and “Life Cycle Analysis”. It was observed that students sometimes assume that CAPE softwares provide straightforward and trustworthy solutions without the need of understanding their mathematical bases and assumptions. Thus, solutions to make students aware of these limitations are proposed, including the creation of an integrated project focussing on complex multi-disciplinary issues, evidencing the need for critical input from the operator.

Multi-walled carbon nanotubes (CNTs) were produced in a continuous inclined mobile-bed rotating reactor by the catalytic chemical vapour deposition of methane on a bimetallic Ni–Mo/MgO catalyst whose activity remains constant in the course of time. Measurements performed on the continuous reactor were validated to ensure that the installation worked correctly and that measurements were precise enough. The performance of the reactor was simulated using a model based on the chemical reactor engineering approach. Hypotheses of the model were verified, and a kinetic study was performed to obtain a kinetic rate expression and to determine the catalytic activity as a function of time. The purity level of produced CNTs depends on the desired properties of the product, so the operating conditions are linked to the purity level that is required. A minimal purity level corresponds to high carbon production, and a maximal purity level corresponds to high specific productivity. It was shown that operating conditions had to be fixed to reach a given specific productivity or a given carbon production, and the optimized operating conditions leading to those two opposite purity level objectives were established.

The methanation of CO2 was performed on Rh/γ-Al2O3 catalyst at temperatures between 50 and 150 °C at 2 bar of pressure in a pulse reactor. Experiments confirm the formation of methane at low temperature and pressure. The formation of formiate species during the adsorption of CO2 can be excluded. After reaction with CO2, the catalyst is oxidized. Oxidation is not observed in the presence of CO. The CO2 is adsorbed dissociatively, forming CO (ads) and O (ads). Gem-dicarbonyl Rh(CO)2 species are more reactive than the Rh–CO linear species. The type of adsorbed species depends on the Rh oxidation state. The formation of methane by hydrogenation of CO2 and CO is carried out with 100% selectivity. The activation energy for the hydrogenation of CO2 and CO is lower than values presented in the literature which have been obtained at higher temperature. In the presence of CO2 and CO, the reaction of methanation of CO2 seems to be inhibited by CO. When oxygen is added in low amount in the reactant gas feed, a positive effect on methanation is observed. When the amount of oxygen is too high, oxygen has a negative effect. These results are in agreement with thermodynamics equilibrium calculations, except when O2 is present, confirming the importance of kinetic effects in the reaction. These results open new perspectives of application of catalysis, in order to recycle CO2 in the presence of H2.

A kinetic study was performed to describe experimental reaction rates of double-walled carbon nanotube (DWNT) synthesis by catalytic chemical vapour deposition (CCVD) over an Fe-Mo/MgO catalyst using methane as the carbon source. Initial reaction rates were determined by mass spectrometry for methane partial pressures ranging from 0.01 atm to 0.9 atm and for three temperatures: 900 °C, 950 °C and 1000 °C. Mass transfers from the bulk of the fluid to the external surface of the catalytic bed and through the catalytic bed were negligible as determined experimentally and confirmed by the methane mass balance. A detailed kinetic study was carried out to discriminate between phenomenological kinetic models and to validate the good agreement between the chosen model and the experimental data. The best model was found to involve the irreversible dissociative adsorption of methane followed by the irreversible decomposition of the adsorbed methyl group, which is the rate-determining centre. Activation energy of adsorbed methyl decomposition was found to be equal to 58 kJ mol −1 . The catalytic deactivation by coking was expressed by a decreasing function of coke content, which leads to a sigmoid equation.

The kinetics of selective hydrodechlorination of 1,2-dichloroethane into ethylene over a Pd-Ag/SiO 2 catalyst was studied using an a priori experimental design with five independent variables—temperature and partial pressures of 1,2-dichloroethane, hydrogen, ethylene and hydrogen chloride. A Langmuir–Hinshelwood model including two types of active site and the 1,2-dichloroethane adsorption as the rate-determining step was found to fit correctly with experimental data, according to the analysis of variance and the analysis of pondered residuals. The study allowed for catalytic deactivation. The rigorous experimental and statistical approach followed to carry out such a kinetic study is explained in detail.

The catalytic deactivation during multi-walled carbon nanotube (MWNT) synthesis by the CCVD process and the influence of hydrogen on it were quantified. Initial specific reaction rate, relative specific productivity and catalytic deactivation were studied. Carbon source was ethylene, and a bimetallic iron–cobalt catalyst supported on alumina was used. The catalytic deactivation was modeled by a decreasing hyperbolic law, reflecting the progressive accumulation of amorphous carbon on active sites. While the initial specific reaction rate was found not to be influenced by hydrogen, catalytic deactivation was found to be modified in the presence of hydrogen, which delayed and slowed down the deactivation by avoiding amorphous carbon deposition, thus leading to a greater relative specific productivity of carbon nanotubes.

The use of in-flight oxygen collection has shown to significantly improve space launcher performance. The conceptual approach followed by Belgian teams working on oxygen collection has been to try to widening the available design margins in order to reduce the required technological leap and limit the economical risk associated with such a development. The aim of the ESA-funded theoretical and experimental study on an air separation device is to demonstrate the possibility of performing efficient air distillation in a compact rotating column. An integration of the vehicle, propulsion system and separation unit designs is presented. The objective is to optimise the overall vehicle performance while keeping technological difficulty and system complexity at a reasonable level. Reference vehicles are presented within their specific mission profiles with an emphasis on TSTO's. Different layouts of the internal energy and mass flowsheets have been studied and were compared in order to make best use of the refrigeration capacity of the hydrogen fuel running through the propulsion system during the first phase of the flight. In those flowsheets, the separator is considered as a classical distillation tray column. That analysis provides the requirements in terms of heat exchange capacity, compression ratios and number of so-called transfer units needed in the separator. The system is intentionally kept simple to limit complexity, but the analysis is thorough and accurate, including, for example, the effect of the presence of argon. Results for a supersonic carrier are presented. A compact separation unit has been designed in order to reach those requirements. That includes internals, practical building with estimates of pressure drops, separation performance and flow limitation. Main results are given, sizing of the separator bed is provided for a carrier plane showing that such on-board separator is indeed practical.

This study is an example of practical application of kinetic data treatment for simultaneous model discrimination and parameter estimation. The study is applied to the hydrogenation of ethylene on a copper-magnesia catalyst and brings a deepened analysis about the experimental strategy by comparing several alternative strategies using a priori and sequential experimental designs. The best model corresponds to a Langmuir–Hinshelwood mechanism with non-competing adsorption of hydrogen and ethylene and where the rate-determining step is either the addition of molecularly adsorbed hydrogen or the addition of the second atom of hydrogen adsorbed dissociatively. Furthermore, the important question of knowing in practice how many designs and how many measurements per design are actually necessary to determine accurate kinetic and physico-chemical parameters, is addressed. A data correction procedure is also presented that takes catalyst deactivation into account.

A kinetic study was performed to describe the initial specific rate of multi-walled carbon nanotube synthesis by catalytic chemical vapor deposition (CCVD) on a bimetallic cobalt–iron catalyst at high temperature using ethylene decomposition to solid carbon and gaseous hydrogen. The study uses a mass spectrometer that allows reaction rate to be inferred from the exhaust gas composition measurements. The aim is to obtain a better understanding of the elementary steps involved in the production of carbon nanotubes so as to derive phenomenological kinetic models in agreement with experimental data. The best models assume the elimination of the first hydrogen atom from adsorbed ethylene as rate determining step and involve a hydrogen adsorption weak enough to be neglected. It was proved that hydrogen partial pressure has no influence on initial reaction rate of carbon nanotube synthesis with the catalyst used for this study. Activation energy and ethylene adsorption enthalpy were found to be equal to around 130 and −130 kJ mol −1 , respectively.

A systematic method to design optimal redundant sensor networks applying data validation theory is described. Sensor networks are optimised thanks to a genetic algorithm. The design objective is to estimate the process key variables within a required accuracy. The solution time is reduced by means of two parallelisation techniques both using the MPI library: the global parallelisation method and the distributed genetic algorithm method. Results are presented for an ammonia synthesis loop.

Data reconciliation and principal component analysis are two recognised statistical methods used for plant monitoring and fault detection. We propose to combine them for increased efficiency. Data reconciliation is used in the first step of the determination of the projection matrix for principal component analysis (eigenvectors). Principal component analysis can then be applied to raw process data for monitoring purpose. The combined use of these techniques aims at a better efficiency in fault detection. It relies mainly in a lower number of components to monitor. The method is applied to a modelled ammonia synthesis loop.

Milk fat exhibits seasonal variations in composition and properties which are undesirable for many subsequent applications. Thus techniques are sought to process dairy products in order to achieve a consistent quality. A brief introduction to milk fat presents its most important particularities, especially composition, seasonal variations, solid fat content, crystallisation and polymorphism. The differential scanning calorimetry (DSC) analysis allows to estimate the solid fat content. Different methods have been developed to estimate more and more precisely the solid fat content which is certainly an important parameter in the description of the textural properties of butterfat. The industrial crystallisation process is modelled on the basis of pilot plant data. That model allows to approach the particular temperature profile in the scraped surface heat exchangers and to give a first explanation of the involved phenomena.

Modern spreadsheet programs offer a wide range of calculation, charting and dialog options, allowing the development of effective teaching aids. Examples of worksheet illustrating numerical analysis, thermodynamics and unit operations are shown. All offer complete interaction between the problem statement and the presentation of results.

In view of the constant growth of electricity usage and public pressure to reduce the dependence on nuclear power plants in the energy supply, solutions are sought to increase the capacity of power plants using fossil fuels. Highly efficient cycles are available : gas turbines combined with waste heat boilers and steam cycles are able to achieve efficiencies above 50–55%. However building new plants requires a large amount of capital. Alternative proposals are based on upgrades of existing plants : capital savings are expected by reusing part of the facilities. In the present study, three parallel proposals are compared on the basis of exergy efficiency; cost of investment and flexibility of operation are also discussed. They are compared with classical Rankine cycle and state of the art combined cycles.

Coupling of a partial oxidation reactor with a gas turbine is proposed to increase the capacity and efficiency of thermal power plants. The exhaust of the partial oxidation gas turbine is a preheated fuel that can be burned with little modification in the existing steam geenerator. The feasibility study requires the development of a reliable model for an existing 250 MW power plant. Data reconciliation is found valuable for obtaining precise mass and energy balances and for screening out unreliable measurements. The identification of model parameters from reconciled values is then straightforward. The same validation software is later applied to rating simulation of the modified plant, by merging the partial oxidation gas turbine in the model.

Solvent degradation may be a major drawback for the large-scale implementation of post-combustion CO2 capture due to amine consumption and emission of degradation products. However, its influence on the process operations has rarely been studied. In the present work, a kinetics model describing solvent oxidative and thermal degradation has been developed based on own experimental results for the benchmark solvent, i.e. 30 wt% monoethanolamine (MEA) in water. This model has been included into a global Aspen Plus model of the CO2 capture process. The selected process modelling approaches are described in the present work. Using the resulting simulation model, optimal operating conditions can be identified to minimize both the energy requirement and the solvent degradation in the process. This kind of process model assessing solvent degradation may contribute to the design of large-scale CO2 capture plants to consider not only the process energy penalty, but also its environmental penalty. Indeed, both aspects are relevant for the large-scale deployment of the CO2 capture technology.

A methodology was developed to improve energy usage in the process industries by optimizing plant operation. The first step was to build fine tuned site-scale models of energy intensive production plants and of combined heat and power utility distribution networks (heat, power, cooling), and later to optimize them using various numerical techniques. The methodology not only comprises these models, but also the selected optimization methods, their implementation in algorithms suited for the problems to be solved and the evaluation of the results in terms of their applicability in real industrial problems. The detailed models were tuned on the basis of actual plant data. They were used to calculate accurate energy balances, and thus to identify sources of energy waste. Finally they allowed optimization of energy usage by adjusting operating conditions at a medium time scale (a few hours to a couple of days). Energy savings in the range 5%–10% could be obtained for energy intensive processes, where even a low percentage represents a considerable amount.

Analysis of the results of a data reconciliation program is made easier by extracting more information from the Jacobian matrix of the constraint equations. Standard deviation for all state variables (measured or not measured) is related to the standard deviation of measurements. Distinction between variables that are actually corrected by the validation process, and those that are merely derived from a single measurement is straightforward. Based on this information, decisions can be taken: deletion of unnecessary measurements, addition of new measurement points and their optimal selection, or identification of key measurements for which any enhancement of accuracy would result in significant improvement in the quality of the process validation.

A dynamic model of a post-combustion capture pilot plant is developed using Aspen Plus Dynamics. An innovative process control strategy is studied for regulating the water balance of the process. A washing section where the flue gas from the absorber is washed with cold water is included to the process in order to reduce the emissions of amine to the air. Control of the water balance in the solvent loop is successfully achieved by changing the washing water temperature. In previous publications regarding CO2 capture pilot plants, the regulation of the water balance always required a water make-up flow which appears here as unnecessary. Rejection of disturbances and different load reduction scenarios are tested to confirm the efficiency of this strategy. Potential operational problems of this control strategy are identified and solved.

An original algorithm has been developed which applies several new concepts to the problem of local models. The algorithm is fully simulator independent and user-transparent. It uses a sequential least squares procedure to build appropriate approximating formulae from a general model that contains all the terms necessary to represent any particular mixture. No a priori information on the type of mixture (ideal/non ideal, narrow/wide boiling, …) or operating conditions of the process is necessary. Though, it is found that sophisticated simulators and integrators can suffer from the discontinuities introduced when parameters are updated or the local model is changed. In this regard, a regressive least squares method would lead to better results. Finally, local models algorithms are proposed as a new tool for the interpolation of properties in raw experimental tables.

A compressor, an expander and an arbitrary separator model have been added to the BELSIM-COL70 steady state simulation package for multi-stage separation processes. The performance of the dogleg method applied to the global equation system of an extended version of Cavett's problem is compared to the classical Wegstein method and the dogleg method applied to the torn variables only. The results of the simulation of a C3-splitter are compared to industrial measurements. Finally, the algorithm is applied to an azeotrope separation process with demixion in the liquid phase.

Carbon capture and storage is a technology that can contribute to face the challenge of rising energy demand combined with a growing environmental awareness. In the present work, the CO 2 capture process with monoethanolamine (MEA) is modeled using the simulation tool Aspen Plus. Two different modeling approaches are studied and compared: the equilibrium and the rate-based approaches. An optimization of key process parameters is performed and process modifications are studied with the objective of improving the global process energy efficiency.

In this paper, neural networks approaches are compared for predicting the high pressure (HP) steam flow rate from a Kraft recovery boiler. We apply two types of neural networks: a static multilayer perceptron and a dynamic Elman's recurrent neural network. Starting from a one-day database of raw process data related to the boiler, the goal is to model and predict the next 12-hours of HP steam flow production from the boiler to the steam turbine. The results illustrate the potential of the dynamic approach in this task.

A data mining methodology, the random forests, is applied to analyze pollutant emission from the recovery boiler of a Kraft pulping process. Starting from a large database of raw process data, the goal is to identify the input variables that explain the most output variations.

The method previously proposed to estimate the uncertainty of validated variables in steady state data reconciliation has been extended to dynamic data reconciliation. The approach used in this article to estimate a posteriori variances in the case of dynamic date validation is based on the one described in [4]. Orthogonal collocations are used to discretize ODE. The influence of the window parameters on posteriori variances is studied. Results are presented for a perfectly mixed reactor with heat exchange and first order kinetic.

The method previously proposed to estimate the uncertainty of validated variables in steady state data reconciliation has been extended to dynamic data reconciliation. The approach used in this article to estimate a posteriori variances in the case of dynamic date validation is based on the one described in [2] for the stationary case. Orthogonal collocations are used to discretise ODE. Results are presented for an adiabatic reactor with first order kinetic.

The experience gained in the development of a CAPE-OPEN 1.0 thermo socket for the BELSIM-VALI software is presented. VALI is a data validation and reconciliation software that provides consistent mass and energy balances, reliable and accurate Key Performance Indicators and soft sensors. It is not a simulation software per definition, but rather a powerful equation based software that is used online, to monitor and manage the operations of different processes. VALI has its own thermodynamic database of chemical compounds (> 800 fluids, > 650 solids) and methods. After demands of our clients it was decided to make VALI compliant with other thermodynamic packages via a CAPE-OPEN 1.0 thermo socket. The source code of the VALI physical property modules had to be modified, to call the Material Object components instead of built-in thermodynamic functions (thermo socket). In the user interface, we had to develop a CAPE-OPEN thermo plugs viewer to allow the selection of a CAPE-OPEN thermo plug from our Graphical User Interface (GUI). Several case studies were analysed, with performance comparison between the native thermodynamic model, and properties obtained from several CAPE-OPEN thermoplugs. We will particularly analyse here the modelling of a gas liquefaction system and a distillation case study.

Data Reconciliation is the first mathematical method that did address the problem of data validation for linear problems. It exploits information redundancy and (linear) conservation laws to extract accurate and reliable information from measurement data and from the process knowledge. It allows producing a single consistent set of data representing actual process operation, assuming the plant is operated in steady state..

A general approach is proposed for designing the cheapest sensor network able to detect and locate a set of specified faults. The method is based on the sensitivity of process residuals with respect to faults. A genetic algorithm is used to select the sensors and their locations. Results are shown for two water networks.

Process monitoring has to consider the problem of measurement uncertainty.A model based approach (data validation) is compared to data driven techniques for an industrial application.

Preface. Foreword. List of Contributors. Volume 1. 1 Introduction. Section 1 Computer-aided Modeling and Simulation. 1 Large-Scale Algebraic Systems (Guido Buzzi Ferraris and Davide Manca). 1.1 Introduction. 1.2 Convergence Tests. 1.3 Substitution Methods. 1.4 Gradient Method (Steepest Descent). 1.5 Newton's Method. 1.6 Modified Newton's Methods. 1.7 Quasi-Newton Methods. 1.8 Large and Sparse Systems. 1.9 Stop Criteria. 1.10 Bounds, Constraints, and Discontinuities. 1.11 Continuation Methods. 2 Distributed Dynamic Models and Computational Fluid Dynamics (Young-il Lim and Sten Bay Jorgensen). 2.1 Introduction. 2.2 Partial Differential Equations. 2.3 Method of Lines. 2.4 Fully Discretized Method. 2.5 Advanced Numerical Methods. 2.6 Applications. 2.7 Process Model and Computational Fluid Dynamics. 2.8 Discussion and Conclusion. 3 Molecular Modeling for Physical Property Prediction (Vincent Gerbaud and Xavier Joulia). 3.1 Introduction. 3.2 What is Molecular Modeling? 3.3 Statistical Thermodynamic Background. 3.4 Numerical Sampling Techniques. 3.5 Interaction Energy. 3.6 Running the Simulations. 3.7 Applications. 3.8 Conclusions. 4 Modeling Frameworks of Complex Separation Systems (Michael C. Georgiadis, Eustathios S. Kikkinides, and Margaritis Kostoglou). 4.1 Introduction. 4.2 A Modeling Framework for Adsorption-Diffusion-based Gas Separation Processes. 4.3 Modeling of PSA Processes in gPROMS. 4.4 Efficient Modeling of Crystallization Processes. 4.5 Modeling of Grinding Processes. 4.6 Concluding Remarks. 5 Model Tuning, Discrimination, and Verification (Katalin M. Hangos and Rozalia Lakner). 5.1 Introduction. 5.2 The Components and Structure of Process Models. 5.3 Model Discrimination: Model Comparison and Model Transformations. 5.4 Model Tuning. 5.5 Model Verification. 6 Multiscale Process Modeling (Ian T. Cameron, Gordon D. Ingram, and Katalin M. Hangos). 6.1 Introduction. 6.2 Multiscale Nature of Process and Product Engineering. 6.3 Modeling in Multiscale Systems. 6.4 Multiscale Model Integration and Solution. 6.5 Future Challenges. 7 Towards Understanding the Role and Function of Regulatory Networks in Microorganisms (Krist V. Gernaey, Morten Lind, and Sten Bay Jorgensen). 7.1 Introduction. 7.2 Central Dogma of Biology. 7.3 Complexity of Regulatory Networks. 7.4 Methods for Mapping the Complexity of Regulatory Networks. 7.5 Towards Understanding the Complexity of Microbial Systems. 7.6 Discussion and Conclusions. Section 2 Computer-aided Process and Product Design. 1 Synthesis of Separation Processes (Petros Proios, Michael C. Georgiadis, and Efstratios N. Pistikopoulos). 1.1 Introduction. 1.2 Synthesis of Simple Distillation Column Sequences. 1.3 Synthesis of Heat-integrated Distillation Column Sequences. 1.4 Synthesis of Complex Distillation Column Sequences. 1.5 Conclusions. 2 Process Intensification (Patrick Linke, Antonis Kokossis, and Alberto Alva-Argaez). 2.1 Introduction. 2.2 Process Intensification Technologies. 2.3 Computer-Aided Methods for Process Intensification. 2.4 Concluding Remarks. 3 Computer-aided Integration of Utility Systems (Francois Marechal and Boris Kalitventzeff). 3.1 Introduction. 3.2 Methodology for Designing Integrated Utility Systems. 3.3 The Energy Conversion Technologies Database. 3.4 Graphical Representations. 3.5 Solving the Energy Conversion Problem Using Mathematical Programming. 3.6 Solving Multiperiod Problems. 3.7 Example. 3.8 Conclusions. 4 Equipment and Process Design (I. David, L. Bogle, and B. Eric Ydstie). 4.1 Introduction. 4.2 The Structure of Process Models. 4.3 Model Development. 4.4 Computer-aided Process Modeling and Design Tools. 4.5 Introduction to the Case Studies. 4.6 Conclusions. 5 Product Development (Andrzej Kraslawski). 5.1 Background. 5.2 Definition Phase. 5.3 Product Design. 5.4 Summary. Volume 2. Section 3 Computer-aided Process Operation. 1 Resource Planning (Michael C. Georgiadis and Panagiotis Tsiakis). 1.1 Introduction. 1.2 Planning in the Process Industries. 1.3 Planning for New Product Development. 1.4 Tactical Planning. 1.5 Resource Planning in the Power Market and Construction Projects. 1.6 Solution Approaches to the Planning Problem. 1.7 Software Tools for the Resource Planning Problem. 1.8 Conclusions. 2 Production Scheduling (Nilay Shah). 2.1 Introduction. 2.2 The Single-Site Production Scheduling Problem. 2.3 Heuristics/Metaheuristics: Specific Processes. 2.4 Heuristics/Metaheuristics: General Processes. 2.5 Mathematical Programming: Specific Processes. 2.6 Mathematical Programming: Multipurpose Plants. 2.7 Hybrid Solution Approaches. 2.8 Combined Scheduling and Process Operation. 2.9 Uncertainty in Planning and Scheduling. 2.10 Industrial Applications of Planning and Scheduling. 2.11 New Application Domains. 2.12 Conclusions and Future Challenges. 3 Process Monitoring and Data Reconciliation (Georges Heyen and Boris Kalitventzeff). 3.1 Introduction. 3.2 Introductory Concepts for Validation of Plant Data. 3.3 Formulation. 3.4 Software Solution. 3.5 Integration in the Process Decision Chain. 3.6 Optimal Design of Measurement System. 3.7 An Example. 3.8 Conclusions. 4 Model-based Control (Sebastian Engell, Gregor Fernholz, Weihua Gao, and Abdelaziz Toumi). 4.1 Introduction. 4.2 NMPC Applied to a Semibatch Reactive Distillation Process. 4.3 Control of Batch Chromatography Using Online Model-based Optimization. 4.4 Control by Measurement-based Online Optimization. 4.5 Nonlinear Model-based Control of a Reactive Simulated Moving Bed (SMB) Process. 4.6 Conclusions. 5 Real Time Optimization (Vivek Dua, John D. Perkins, and Efstratios N. Pistikopoulos). 5.1 Introduction. 5.2 Parametric Programming. 5.3 Parametric Control. 5.4 Hybrid Systems. 5.5 Concluding Remarks. 6 Batch and Hybrid Processes (Luis Puigjaner and Javier Romero). 6.1 Introduction. 6.2 The Flexible Recipe Concept. 6.3 The Flexible Recipe Model. 6.4 Flexible Recipe Model for Recipe Initialization. 6.5 Flexible Recipe Model for Recipe Correction. 6.6 Final Considerations. 7 Supply Chain Management and Optimization (Lazaros G. Papageorgiou). 7.1 Introduction. 7.2 Key Features of Supply Chain Management. 7.3 Supply Chain Design and Planning. 7.4 Analysis of Supply Chain Policies. 7.5 Multienterprise Supply Chains. 7.6 Software Tools for Supply Chain Management. 7.7 Future Challenges. Section 4 Computer-integrated Approaches in CAPE. 1 Integrated Chemical Product-Process Design: CAPE Perspectives (Rafiqul Gani). 1.1 Introduction. 1.2 Design Problem Formulations. 1.3 Issues and Needs. 1.4 Framework for Integrated Approach. 1.5 Conclusion. 2 Modeling in the Process Life Cycle (Ian T. Cameron and Robert B. Newell). 2.1 Cradle-to-the-Grave Process and Product Engineering. 2.2 Industrial Practice and Demands in Life-Cycle Modeling. 2.3 Applications of Modeling in the Process Life Cycle: Some Case Studies. 2.4 Challenges in Modeling Through the Life Cycle. 3 Integration in Supply Chain Management (Luis Puigjaner and Antonio Espuna). 3.1 Introduction. 3.2 Current State of Supply Chain Management Integration. 3.3 Agent-based Supply Chain Management Systems. 3.4 Environmental Module. 3.5 Financial Module. 3.6 Multiagent Architecture Implementation and Demonstration. 3.7 Concluding Remarks. 4 Databases in the Field of Thermophysical Properties in Chemical Engineering (Richard Sass). 4.1 Introduction. 4.2 Overview of the Thermophysical Properties Needed for CAPE Calculations. 4.3 Sources of Thermophysical Data. 4.4 Examples of Databases for Thermophysical Properties. 4.5 Special Case and New Challenge: Data of Electrolyte Solutions. 4.6 Examples of Databases with Properties of Electrolyte Solutions. 4.7 A Glance at the Future of the Properties Databases. 5 Emergent Standards (Jean-Pierre Belaud and Bertrand Braunschweig). 5.1 Introduction. 5.2 Current CAPE Standards. 5.3 Emergent Information Technology Standards. 5.4 Conclusion (Economic, Organizational, Technical, QA). Section 5 Applications. 1 Integrated Computer-aided Methods and Tools as Educational Modules (Rafiqul Gani and Jens Abildskov). 1.1 Introduction. 1.2 Integrated Approach to CAPE. 1.3 Educational Modules. 1.4 Conclusion. 2 Data Validation: a Technology for Intelligent Manufacturing (Boris Kalitventzeff, Georges Heyen, and Miguel Mateus). 2.1 Introduction. 2.2 Basic Aspects of Validation: Data Reconciliation. 2.3 Specific Assets of Information Validation. 2.4 Advanced Features of Validation Technology. 2.5 Applications. 2.6 Conclusion. 3 Facing Uncertainty in Demand by Cost-effective Manufacturing Flexibility (Petra Heijnen and Johan Grievink). 3.1 Introduction. 3.2 The Production Planning Problem. 3.3 Mathematical Description of the Planning Problem. 3.4 Modeling the Profit of the Production Planning. 3.5 Modeling the Objective Functions. 3.6 Solving the Optimization Problem. 3.7 Sensitivity Analysis of the Optimization. 3.8 Implementation of the Optimization of the Production Planning. 3.9 Conclusions and Final Remarks. Authors' Index. Subject Index.

[Abstract] In the context of an ESA-funded study on an air separation device intended to provide in-flight Oxygen Collection capability to future launchers, experimental and system level investigations are performed. Formerly hosted by the Belgian Royal Military Academy, it is now sustained by the Universite Libre de Bruxelles. Looking at the world wide picture of space launcher studies, oxygen collection is seldom studied by space researchers and engineers. There are few really rational causes to this situation. Aerospace engineers are often reluctant to invest themselves in very different fields, which is strongly required here. Besides, the research projects are often cornered by funding constraints towards very high technology options. This often pushes resulting preliminary studies away from practically and economically viable new applications. To make design trades more accessible over such an heterogeneous set of concepts, simple models are required. The work provided here is the result of simplifying both the cycle, to limit technological constraints while retaining most of the performance, and to provide an approximate modelling while retaining a sensible analysis and sufficiently accurate predictions. The vehicle considered here has been presented in previous articles, but the result has a wider scope. Strong beneficial effects comes from combining the propulsion cycle of the first stage (possibly subsonic) with the separation process, leading to both oxygen collection and even improved propulsion efficiency for the first stage. This approach attempts to draw a reasonable bottom line for the separation plant performance and for the required system complexity. Simplified, but widely accepted, methods are presented for analysing the various aspects of the separation plant performance. Although the model has some drawbacks that can be corrected using a limited set of more accurate predictions, the mass and energy balance can be solved accurately. Options that have a strong impact on performance –mainly expressed by the collection ratioare analysed: use of significant hydrogen pressurisation and use of para-ortho conversion that improves cooling capacity of hydrogen. Although the system is brought to minimal complexity and some hydrogen capacity is wasted for system heat integration simplicity, the retained performance are found well within minimal requirements to sustain operation and are consistent with previous more accurate computations. The analysis therefore allows to assess a viable bottomline for the propulsion system complexity and technological level and to predict performance of oxygen collecting two stage to orbit vehicles with a rather simple analysis.

The use of in-flight Oxygen Collection has shown to significantly improve space launcher performance. The conceptual approach followed by the Royal Military Academy of Brussels (RMA) has tried to widen the available design margins in order to reduce the required technological leap and limit the economical risk associated with such a development. The aim of the ESA-funded theoretical and experimental study on an air separation device is to demonstrate the possibility of performing efficient air distillation in a compact rotating column. An integration of the vehicle, propulsion system and separation unit designs is presented aiming to optimise the overall vehicle performance while keeping technological difficulty and system complexity at a reasonable level. Reference vehicles are presented in their specific mission profiles with an emphasis on TSTO’s. Different layouts of the internal energy and mass flowsheets are compared, in order to make best use of the refrigeration capacity of the hydrogen fuel running though the propulsion system during the first phase of the flight considering the separator as a classical distillation column. This analysis provides the requirements in terms of heat exchange capacity, compression ratios and number of so-called transfer units needed in the separator. Here, the system is intentionally kept simple, to limit complexity, but the analysis is thorough and accurate, including, for example, the effect of the presence of argon. An analysis of the separation unit to reach those requirements is proposed. That includes internals, practical building with estimates of pressure drop, separation performance and flow limitation. Analysis of size reduction of the distillation unit from usual 1-g column to the high-g unit is provided as well as the scale up methodology of laboratory results. First experimental results obtained with our centrifugally enhanced distillation separation system are presented and perspectives for a larger on-board operational unit proposed.

Multi-objective optimisation (MOO) has been used with an equation solver data reconciliation software to develop a tool for sensor system design based on modifying the sensitivity matrix of a simulated process. MOO enables searching for the best trade-off between two conflicting objectives: the cost of the system and the precision of key performance indicators (KPI) (variables that have to be measured or calculated). This methodology has been applied to design the sensor system of a two stage experimental air-water heat pump. Proper knowledge of modelling equations and constants helps to improve the estimation of the precision of variables, and lowers the cost of the system. Compared to single objective optimisation, the MOO strategy increases the number of solutions, yet the precision function still relates to different objectives for each KPI, and its formulation is shown to have an impact on the trade-off obtained.

A general method to design optimal sensor networks able to estimate process key variables within a required precision was recently proposed. This method is based on the data reconciliation theory and the sensor network is optimised thanks to a genetic algorithm. To reduce the solution time, two ways of parallelizing the algorithm have been compared: the global parallelization and the distributed genetic algorithms. Both techniques allow reducing the elapsed time but the second one is more efficient.