You are on page 1of 9

Chapter 4

Overview of the Case Studies


In Part II of this book, we are trying to highlight the modelling practice through
a number of modelling case studies. In each case, we have provided detailed
descriptions of the models used in terms of model objectives, assumptions, model
description (including model equations), model analysis and numerical solution.
The data used for the numerical solution has also been given so that the readers may
try to compare and/or validate their solutions with ours. We have not tried to provide
a detailed derivation of the model, as in most of the case studies the models have
been taken from published literature. Unless otherwise indicated, all the models
are analysed and solved through the ICAS-MoT modelling tool (Sales-Cruz, 2006;
Heitzig et al. 2009). The model equations in each case are listed in Appendix-1
and the corresponding ICAS-MoT file can be downloaded from a website for the
book, established by Elsevier (http://www.elsevierdirect.com/companion.jsp?
ISBN=9780444531612). Also, a list of other interesting models that may be down-
loaded from the same Website is given in Appendix-2.
An overview of the models used in the case studies of this part is given
through Table 1. Examples of the workflows and data flows associated with the
modelling procedure are shown in Figures 1-3. The text below gives a brief
overview of the contents of the chapters of part-II.
Chapter 5
The chapters with the case studies have been organised into themes and,
within them, into specific topics. Chapter 5 deals with constitutive (mainly
property) models, their identification (parameter estimation) and their applica-
tion. Through Sections, 5.1, 5.2 and 5.3, we have tried to highlight the use of an
incremental modelling approach. That is, first we develop the constitutive
property models needed to study the evaporation of a chemical from a water
droplet. Then we identify the models in terms of estimating their parameter
values by matching a set of collected property-related data. Then we use these
identified models within the framework of an evaporation model. In addition,
other examples of parameter estimation involving solidliquid data and
electrolyte systems are also given. Finally, in Section 5.3.4, the use of consti-
tutive (kinetic) models to generate attainable region diagrams is highlighted.
This is similar, in principle, to the use of constitutive (property) models to
generate phase equilibrium diagrams for chemical systems.
Chapter 6
Chapter 6 deals with process simulation and optimisation. Here, we high-
light the use of the equation-oriented approach as well as the sequential modular
77
78
TABLE 1 List of Case Studies Covered in Part II

Theme Case study Short Description Chapter Section

Constitutive models Pure component Modelling of temperature dependent correlations. The 5 5.1.1
(development) property example shows the modelling of the Antoine correlation for
pure component vapour pressures as a function of
temperature.
Modelling of PVT relations through an equation of state. How 5 5.1.2
to model the molar volume of a chemical at a given
temperature and pressure through a cubic equation of state

Product and Process Modelling: A Case Study Approach


Mixture property Modelling of liquid phase activity coefficients through the 5 5.1.3
Wilson model
Constitutive models Pure component Regression of Antoine correlation parameters through 5 5.2.1
(parameter property supplied data
estimation)
Mixture properties The regression of the Wilson model parameters by matching 5 5.2.2
supplied data of activity coefficients
Mixture properties Regression of elec-NRTL model parameters by matching 5 5.2.3
supplied solubility data of an ionic liquid
Constitutive models VLE data fit Parameter estimation for the Wilson model parameters with 5 5.3.1
(application) vapourliquid equilibrium data (uses also the Antoine
correlation)
SLE calculation Use of the NRTL model within a model for SLE calculation 5 5.3.2
Evaporation from a Use of Antoine correlation, Wilson model and a VLE 5 5.3.3
droplet calculation model within a model for calculating
evaporation from a droplet

(continued)
Chapter | 4
TABLE 1 (continued)

Theme Case study Short Description Chapter Section

Attainable region Use of kinetic models within model for attainable region 5 5.3.4
diagram diagram calculation

Overview of the Case Studies


Steady-state Equation oriented Derivation of simple mass balance models and their use in 6 6.1
flowsheeting process flowsheet simulation
Process simulator Use of model library for process flowsheet modelling and 6 6.2
simulation
Modelling of dynamic Blending tank Derivation and solution of a complex model for a blending 7 7.1
systems tank
Fuel cell (DMFC) Analysis and solution of a mutli-scale dynamic model and for 7 7.2
a direct methanol fuel cell
Fluidised bed Analysis and solution of a multi-scale model for a fluidised 7 7.3
bed
Chemical reactor Analysis and solution (through ICAS-MoT) of a simple 7 7.4
dynamic model for a CSTR
Polymerisation reactor Analysis and solution of a dynamic model for a 7 7.5
polymerisation reactor (multiple solutions)
Distributed parameter Processing of oil-shale Modelling of a retorting device for the processing of oil-shale 8 8.1
modelling (discretisation along spatial direction)
Granulation process Modelling of a commercial granulator for producing 8 8.2
industrial-grade phosphate fertilisers (discretisation along
spatial direction)
Short-path evaporation Steady-state modelling of a short-path evaporation process 8 8.3
(discretisation in radial direction, integration in axial
direction)

(continued)

79
80
TABLE 1 (continued)

Theme Case study Short Description Chapter Section

Process modelling and Tennessee Eastman Analysis and solution of the simplified version of the 9 9.1
simulation Tennessee Eastman Challenge process model
Tennessee Eastman Analysis and solution of the complete Tennessee Eastman 9 9.2
Challenge process model
Batch process operations Batch crystallisation Development, analysis and solution of models of different 10 10.1
types of batch cooling crystallisation
Batch distillation Uses of a dynamic model for batch distillation combined 10 10.2
with event modelling

Product and Process Modelling: A Case Study Approach


Parameter estimation Reaction kinetics Regression of model parameters for different types of kinetic 11 11.111.3
models
Reaction kinetics Model discrimination (least squares fitting) 11 11.4
Reaction kinetics Model discrimination (maximum likelihood principle) 11 11.5
Reaction kinetics Uses of orthogonal collocation and finite difference methods 11 11.6
in parameter estimation
Bio-, agro-, pharma- Controlled release Analysis and solution of a model for controlled release from a 12 12.1
applications microcapsule
Fermentor Analysis and solution of a dynamic model for fermentor (for 12 12.2
conversion of glucose to ethanol)
Milk pasteurisation Analysis and solution of a model for a milk pasteurisation 12 12.3
process
Milling Analysis and solution of a model for a milling process (used as 12 12.4
part of tablet manufacturing process)
Granulation Analysis and solution of a model for a granulation process 12 12.4
(used as part of tablet manufacturing process)
Tablet pressing Analysis and solution of a model for a tablet pressing process 12 12.4
(used as part of tablet manufacturing process)
Chapter | 4
[(Figure_1)TD$IG]

Overview of the Case Studies


FIGURE 1 Workflow for single-scale model development.

81
82
[(Figure_2)TD$IG]

Product and Process Modelling: A Case Study Approach


FIGURE 2 Workflow for multi-scale model development.
Chapter | 4 Overview of the Case Studies 83

[(Figure_3)TD$IG]

FIGURE 3 Flow diagram of the workflow and data flow in model identification.

approach. The use of two modelling approaches, for example deriving model
equations and solving them or using already developed models from a model
library, is highlighted. In the former, we use the ICAS-MOT modelling tool to
develop, analyse and solve the model equations, whilst in the latter, we use a
commercial simulator (PRO/II) to perform the simulations.
84 Product and Process Modelling: A Case Study Approach

Chapter 7
Chapter 7 highlights models for different (dynamic) applications. Five
detailed case studies are presented. The first deals with the modelling of a
complex unit operation (blending tank). The second involves the multi-scale
dynamic modelling of the performance of a direct methanol fuel cell. The third
also involves multi-scale modelling, but this time studies a fluidised bed reactor.
The fourth presents a typical dynamic model for a chemical reactor. The final
case study in this chapter involves the dynamic modelling of a polymerisation
reactor.
Chapter 8
Through chapter 8, we highlight some aspects of distributed modelling. In
the first case study, an oil-shale pre-heating cooling unit is modelled. In the
second case study, the dynamic performance of a granulator has been modelled,
whilst in the third case study, the operation of a short-path evaporator has been
modelled. In each case the model equations have been discretised in one
(spatial) direction and integrated in the other spatial direction (Cases 8.1 and
8.3) or with respect to time (case 8.2).
Chapter 9
The modelling issues related to the well-known Tennessee-Eastman chal-
lenge problem are highlighted in Chapter 9 through two case studies, involving
different forms of the models proposed in the published literature. Both these
models are quite complex, and the use of a systematic step-by-step solution
strategy helps to get the solutions, in these cases multiple solutions.
Chapter 10
Chapter 10 covers aspects of modelling, and simulation of batch operations.
Two case studies are presented. In the first case study, a detailed step-by-step
procedure to develop a model for the study of batch crystallisation operations is
presented; then a model is developed and analysed using this procedure, and
finally solved for a specific chemical system (sucrose crystallisation). In
the second case study, an example of events modelling is provided through
the multiple running of a batch distillation operation. Here, using an available
dynamic model for a process, an events modelling feature is added to evaluate
the multiple operations associated with batch distillation.
Chapter 11
Chapter 11 returns to the topic of constitutive models, in this case, different
types of parameter estimation problems related to identification of kinetic
models. Examples cover simple steady-state least squares regression to non-
linear least squares fitting, to model selection, to the use of maximum likelihood
principle and to finally, use of orthogonal collocation for dynamic optimisation.
Chapter 12
The final set of examples covered in chapter 12 highlight the theme
modelling, for bio, agro and pharma applications. We start with an example
from the agro-chemical sector and highlight the modelling of the perfor-
mance of a microcapsule for controlled release. The second example deals
Chapter | 4 Overview of the Case Studies 85

with a bio-process, that is, a bio-fermentation process for the conversion of


glucose to ethanol. This model is able to monitor the performance of the
fermentor with respect to a number of key operational variables. The last
three examples come from the pharma sector. They deal with the operational
steps in the production of drug tablets. First the model of a milling process is
given, then a model for a granulation process and finally, a model for a tablet
pressing operation. In all three cases, the models have been developed with
the objective to support process-product monitoring through application of
PAT (process analytical technology). The models supplement to the collected
data for monitoring the product quality under a specific condition of
operation.

You might also like