Computer Aided Property Estimation for Process and Product Design (eBook)

Computers Aided Chemical Engineering

Rafiqul Gani, Georgios M. Kontogeorgis (Herausgeber)

eBook Download: EPUB

2004 | 1. Auflage
436 Seiten
Elsevier Science (Verlag)
978-0-08-047228-7 (ISBN)

Infos

Properties of chemical compounds and their mixtures are needed in almost every aspect of process and product design. When the use of experimental data is not possible, one of the most widely used options in the use of property estimation models.
Computer Aided Property Estimation for Process and Product Design provides a presentation of the most suitable property estimation models available today as well as guidelines on how to select an appropriate model. Problems that users are faced with, such as: which models to use and what their accuracy is, are addressed using a systematical approach to property estimation.
The volume includes contributions from leading experts from academia and industry. A wide spectrum of properties and phase equilibria types is covered, making it indispensable for research, development and educational purposes.

* This book presents the latest developments in computational modelling for thermodynamic property estimation.
* It combines theory with practice and includes illustrative examples of software applications.
* The questions users of property models are faced with are addressed comprehensively.

Inhalt

Cover 1
Preface 5
List of Contributors 7
Contents 9
Introduction to Computer Aided Property Estimation 11
Introduction to Computer Aided Property Estimation 13
Introduction 13
Classification of Properties & Models
Property Model Development 15
Property Model Use 18
Property Model Selection 20
Examples of Property Model Use 23
Example 1: Estimation of Unknown Pure Component Properties 23
Problem Definition 23
Property Model Selection 23
Model Validation & Parameter Estimation
Property Estimation 23
Example 2: New Group Definition & Parameter Estimation
Problem Definition 24
Property Model Creation/Selection 24
Model Validation and Parameter Estimation 25
Property Estimation 25
Example 3: Extraction of Heavy Compounds Using Supercritical Gases 25
Background 25
Problem Definition 26
Property Model Selection 26
Model Validation and Parameter Estimation 26
Property Estimation 27
Example 4: Density Estimation for Polymers 27
Background 27
Problem Definition 27
Property Model Selection 27
Model Validation and Parameter Estimation 29
Property Estimation 29
Example 5: Gas-Liquid Equilibria for Asymmetric Systems 30
Background 30
Problem Definition 30
Property Model Selection 30
Model Validation and Parameter Estimation 30
Property Estimation 30
Example 6: Selective Separation of Structurally Similar Compounds Using Supercritical Fluid Extraction (SCFE) 31
Background 31
Problem Definition 32
Property Model Selection 32
Model Validation and Parameter Estimation 32
Property Estimation 32
Example 7: Thermodynamic Diagrams for Heat Pump Fluids 33
Background 33
Problem Definition 33
Thermodynamic Diagram Calculation 34
Conclusions 34
References 35
Appendix 36
Role of Properties and their Models in Process and Product Design 37
Introduction 37
Roles of Properties 38
Service Role 40
Service/Advice Role 41
Integration Role 43
Property Roles and Property Model Selection 45
Service Role 46
Service/Advice Role 47
Integration Role 48
Derivative Analysis: A Connection between Model Generators and Model Users 49
Conclusions 50
References 50
Models for Properties 53
Pure Component Property Estimation: Models & Databases
Introduction 55
Models for Primary Properties 55
Primary Property Models 56
Properties & Models
Models for Secondary Properties 58
Secondary Property Models 58
Properties & Models
Secondary Properties Modeled as Primary Property 61
Functional Properties 61
Properties & Models
Databases 63
The CAPEC Database8 63
Compound Classification 63
Pure Component Data 64
Mixture Data 65
Special Data 65
Search Engine 65
References for Databases 65
Conclusions 66
References 67
Models for Liquid Phase Activity Coefficients - UNIFAC 69
Introduction 69
The UNIFAC Method 70
General Model Equations (1st-order) 71
Combinatorial Term - In GammaiC 71
Residual Term - In GammaiR 72
Model Equations for Different Versions of UNIFAC 73
General Model Equations (KT-UNIFAC10) 73
Model Analysis 75
Numerical Analysis 76
Model Parameters 77
R & Q Parameters
Interaction Parameters 78
Performance/Limitations 78
Other Applications of UNIFAC 80
Conclusions 81
List of Symbols 81
References 82
Equations of State with Emphasis on Excess Gibbs Energy Mixing Rules 85
Introduction 85
The Van Der Waals One Fluid Mixing Rules 87
Basic Approach of EoS/GE Models 91
Brief Presentation of the Models 92
The MHV2 Model 92
The EoS 92
The Mixing Rule 92
The UNIFAC Model 92
The PSRK Model 92
The EoS 92
The Mixing Rule 93
The UNIFAC Model 93
The LCVM Model 93
The EoS 93
The Mixing Rule 93
The UNIFAC Model 93
The Wong-Sandler (WS) Model 93
The EoS 93
The Mixing Rule 94
The Predictive WS Model 94
The UNIFAC Model 94
Limitations of the WS Predictive Model 95
Input Information Needed for the Application of the Models 95
EoS Pure-Compound Parameters 95
UNIFAC Interaction Parameters 96
Performance of EoS/GE Models in VLE/GLE Prediction 96
Binary Mixtures 96
Non-Polar Mixtures 96
Polar Mixtures 99
The Case of Aqueous Mixtures Containing Hydrocarbons 100
Ternary and Multicomponent Mixtures 102
Ternary Mixtures 102
Multicomponent Mixtures 102
Performance of Cubic EOS for Solid-Gas Equilibria 107
General - Modelling Using vdW1f Mixing Rules 107
Modeling with LCVM 108
Effect of Using VLE-Based Parameters for Making SGE Predictions 114
Concluding Remarks - Future Challenges 114
List of Symbols 117
List of Abbreviations 118
References 118
Association Models - The CPA Equation of State 123
Introduction 123
The CPA Equation of State - Model Description 126
The Radial Distribution Function: From the Original to the Simplified CPA 128
The CPA Equation of State - Parameter Estimation 128
Pure Fluids 129
Mixtures: Mixing and Combining Rules 132
The CPA Equation of State - Current Status/Results 133
Discussion of the Results 133
Water / Alkanes LLE 133
VLE for Mixtures with Associating Substances 133
The Methanol/Propane System 134
Comparison with Other Models 134
LLE for Alcohol/Alkanes 136
Glycol/Alkane Liquid-Liquid Equilibria 136
Multicomponent Equilibria 139
Use of the CPA EoS 141
Challenges 141
Systems with Aromatic Hydrocarbons 141
Cross-Associating Systems 142
Critical Area of Mixtures 145
New Associating Compounds 145
Concluding Remarks 147
List of Symbols 148
List of Abbreviations 149
References 150
Models for Polymer Solutions 153
Introduction - Areas of Application 153
Choice of Solvents 154
The Rules of Thumb Based on Solubility Parameters 155
The Rule of Thumb Based on the Infinite Dilution Activity Coefficient 157
The Rule of Thumb Based on the FIory-Huggins Model 159
The Free-Volume Activity Coefficient Models 163
The Free-Volume Concept 163
The UNIFAC - FV Model 166
The Entropic-FV Model 167
Results and Discussion 168
Vapor-Liquid Equilibria 168
Athermal Systems 172
Non-Polar and Slightly Polar Systems 173
Water-Soluble Polymers and Other Hydrogen-Bonding Systems 173
Co-Polymer Solutions 174
Polymer-Mixed Solvent Systems 174
Liquid-Liquid and Solid-Liquid Equilibria 174
The Entropic-FV/UNIQUAC Model 177
Extension to Semi-Crystalline Polymers and Swelling 177
Extension of Free-Volume Models to Gas Solubilities in Elastomers 179
Equations of State for Polymers 182
List of Abbreviations 186
References 187
Property Estimation for Electrolyte Systems 191
Introduction 191
Properties & Data Analysis
Phase Diagrams 191
Thermodynamic Properties 194
Density 194
Liquid Volume 195
Dissociation Constant & Solubility Product
Improving Standard Values and Property Values at Temperatures of Interest 198
Water Activity & Osmotic Coefficient
Experimental Methods 201
Electrolyte Property Models 202
Activity Coefficient Model Library 203
Pitzer-Bromley-Debye Huckel Model 203
Electrolyte NRTL Model 205
Extended UNIQUAC Model (Aqueous and General) 209
Osmotic Coefficient-Based Model 211
Henry's Constant 212
Condition of Physical and Chemical Equilibrium 212
Solubility Index 212
Solubility as a Function of Temperature and Solution pH 213
Conclusions 213
References 213
Diffusion in Multicomponent Mixtures 215
Introduction 215
Multicomponent Diffusion 216
General Facts About Diffusion 216
Diffusion Coefficients - Different Definitions and their Connections 218
Diffusion Coefficients in Rarified Gases 222
Diffusion in Liquids 223
Approaches Based on Reduction of the Number of Independent Diffusion Coefficients 224
Approaches Based on the Free Volume or Activation Energy Concepts 226
Other Approaches 228
Diffusion in Polymer Solvent Systems 228
Free Volume Basics 228
Parameter Estimation 230
Applicability of the Free-Volume Theory 233
References 236
Modelling Phase Equilibria in Systems with Organic Solid Solutions 239
Introduction 239
State of Art 240
Non-ideality of the Liquid Phase 240
Non-ideality of the Solid Phase 240
Solid Phase and Thermophysical Properties 240
Won's Model Class 241
Erickson's Model Class 241
Ungerer's Model Class 241
Calange's Model Class 242
Predictive Free Energy Model Class 242
Modelling High Pressure Wax Formation 242
Wax Formation Model 243
Low Pressure 244
The Liquid Phase Activity Coefficient 244
The Solid Phase Activity Coefficient - Predictive UNIQUAC 244
High Pressure 246
Fluid Phase Fugacities 246
Solid Phase Fugacity 247
The Solid Phase Activity Coefficient - Predictive Wilson 247
Multiphase Flash Calculations 248
Fluid Characterization 249
Characterization of n-Alkanes 249
Total n-alkane Content 249
The n-alkane Decay 249
Characterization of Solvent 250
Thermophysical Properties for the n-AIkanes 250
Results 252
Other Systems 254
References 256
An Introduction to Modeling of Gas Hydrates 261
Introduction 261
What are the Gas Hydrates? 261
Physical Properties of the Hydrates 261
History 262
Importance of Gas Hydrates 262
Methods to Prevent Hydrates 262
Types of Gas Hydrates 263
Which Components Form Hydrates? 264
Reviews 264
Thermodynamic Behaviour of Hydrates 265
Hydrate Equilibrium Curve 265
Engineering Rules 265
Binary Systems 267
Multi-Component Systems 268
Effect of Drying 268
Depressurization 269
Retrograde Behavior 269
Azeotropy 271
Inhibition 272
Structure Transition 274
Phase Equilibrium Calculations & Thermodynamic Models
Equilibrium Conditions 274
Thermodynamic Model 275
Hydrate Phase 276
Chemical Potential of Water in the Hydrate Phase 277
Composition of the Hydrate 277
Thermodynamic Properties of the Empty Lattice 278
Status and Developments of the Solid Solution Model 279
Acknowledgement 280
References 280
Appendix A: Derivation of the Hydrate Phase Equations 280
Simple Derivation Based on Chemical Reaction Theory 280
Mixed Grand Partition Function and Thermodynamic Functions 282
Evaluation of Mixed Grand Partition Function q 283
Derivation of Thermodynamic Functions 284
Appendix B: Glossary of Terms 285
Applications of Property Models and Databases 287
Molecular Simulation of Phase Equilibria for Industrial Applications 289
Introduction 289
Molecular Simulation Methods 291
Monte Carlo Simulation Methods for Phase Equilibria 292
Simulation Methods for Chemical Potential Calculation 295
Monte Carlo Moves for Long Chain Molecules 297
Force-Fields for Complex Fluids 300
Applications 303
Force-Fields for Hydrocarbons 303
Force-Fields for Water 305
Mixture Phase Equilibria 309
Conclusions 314
Acknowledgment 314
References 315
Property Models in Computation of Phase Equilibria 319
Introduction 319
Different Types of Phase Equilibrium Problems 319
Computation of VLE 321
Model Analysis 321
Computation of LLE & VLLE
LLE-phase Diagrams 325
LLE Phase Separation Calculations 326
VLLE Phase Diagrams (3-phases) 327
PT-Flash for VLLE 327
Computation of SLE 328
Application Examples: SPECS 329
Example 1 - Evaporation of Isopropanol from a PVAC Film 330
Solution 330
Using the FH Model 330
Using SPECS 331
Example 2 - The Effect of Polymer Molecular Weight on Activity Coefficients for Polymer Solutions 332
Solution 332
Example 3 - Comparison of Models for Polymer-Solvent VLE with SPECS 334
Example 4 - SRK with Various Mixing Rules using SPECS 334
Example 5 - Complex Systems with SRK & CPA Using SPECS
Application Examples: ICAS 339
VLE Calculation Examples 339
Problem 1 - VLE Phase Diagram Calculation 339
Verify Binary Azeotrope and Pressure Dependence of Azeotrope 339
Select a Solvent and Study the Ternary Mixture VLE 340
LLE & VLLE Calculation Examples
Problem 2 - Investigate the Ternary System Ethanol-Water-Benzene at a Pressure of 1 atm 343
Property Model Selection 343
Calculate the LLE and VLLE Diagrams 343
SLE Calculation Examples 343
Problem 3 - Investigate the Solubility of Morphine in Single and Mixed Solvent Systems 343
Property Model Selection 343
Pure Component Property Prediction with ProPred in ICAS 343
SLE Calculation with the Utility Toolbox in ICAS 345
Conclusions 347
References 347
Application of Property Models in Chemical Product Design 349
Introduction 349
Property Model Needs in Chemical Product Design 350
Predictive Property Models for Base Properties 352
Distribution Coefficients and Octanol-Water Partition Coefficients with UNIFAC 352
Introduction 352
Estimation Methods 353
Local UNIFAC Model for Estimation of Solubility of Complex Molecules 360
Represent the Solute with the Appropriate Groups 361
Estimate the Needed Pure Component Properties 361
Solubility Data Analysis and Setting up of the Group Interaction Parameters 361
Estimation of Liquid Phase Activity Coefficients with the UNIFAC Method 361
Sensitivity Analysis of the Group Interaction Parameters 362
Estimation of the Most Sensitive Group Interaction Parameter that Fits the Experimental Data 362
Estimation of Solubility of Carbazole in Different Solvents Covered by the Regressed Parameters from Step 6 363
Product Design Examples 365
Example 1: Assessment of the Miscibility of Plasticizers in PVC 365
Problem Definition 365
Property Model Selection 365
Model Validation and Parameter Estimation 365
Property Estimation 366
Miscibility Based on Flory-Huggins Parameters 367
Conclusions & Perspectives
Example 2: Choice of Suitable Mixed Solvents in the Paint Industry 368
Problem Definition 368
Property Model Selection 368
Model Validation and Parameter Estimation 369
Property Estimation 369
Discussion 370
Example 3: Choosing Alternative Surfactants for Stabilizing an Emulsion 371
Design 372
Basic Relationships in Environmental Engineering 373
Fugacity of a Chemical in Air 373
Fugacity of a Chemical in Water 374
Fugacity of a Chemical in Soil and Sediment 374
Fugacity of a Chemical in Biota 375
The Octanol-water and Other Partition Coefficients 376
List of Symbols 378
Abbreviations 378
References 378
Computational Algorithms for Electrolyte System Properties 381
Introduction 381
Computational Algorithms 381
Algorithm for Solubility Analysis 381
Specification of System and Type of Calculation 382
System Classification 382
Generation/Solution of the Solubility Model Equations 382
Example 1: The Effect of pH on the Solubility of DUP860 in H2O 382
Example 2: The Effect of Temperature on the Solubility of Na2SO4 in H2O 383
Algorithms for Phase Diagram Generation 383
Vapor-liquid Equilibrium Phase Diagram 384
Solid-liquid Equilibrium Phase Diagram 386
Liquid-liquid Equilibrium Phase Diagram 389
Algorithms for Data Regression 392
Algorithm for Creation of Property Model Package 393
Problem Definition 393
Electrolyte System Definition 394
Property Model Selection 395
Determine Pattern Index 396
Compare Pattern 396
List of Necessary Properties and Constants 398
Parameter Tables 401
Data Regression Algorithm 401
Computational Algorithms for Process Simulation 406
The MPS Model Equations 407
Conclusions 408
References 408
Appendix A 409
Challenges and Opportunities 415
Challenges and Opportunities for Property Modeling 417
Introduction 417
Classical Separations 417
Materials and Chemical Products 419
Structured Polymers and Complex Interactions 420
Surface Phenomena 420
Multicomponent, Multiphase Mixtures and Complex Phase Diagrams 421
Emulsions 421
Pharmaceuticals and Agrochemicals 421
Polymorphism 422
Life Sciences and Biotechnology 422
Computer-Aided Systems 423
Thermodynamics in the Post-Modern World 423
Conclusions 424
References 425
Subject Index 427
Author Index 435

Textauszug

Chapter 1

Introduction to Computer Aided Property Estimation

Georgios M. Kontogeorgis; Rafiqul Gani

1.1 INTRODUCTION

Computer aided property estimation implies the use of mathematical models for the calculation of the needed properties. Depending on the type of property, the size-scale factor, the required accuracy, the application, etc., the mathematical models have varying degrees of complexity. For example, they may be very simple polynomial functions (correlations) representing various pure component temperature dependent properties, a non-linear set of algebraic equations representing various cubic equations of state, or, a very large set of differential-algebraic equations representing the behavior of atoms within a specified boundary. In all cases, an appropriate mathematical model is derived or selected and solved with the help of computers to obtain the necessary property values. These mathematical models, representing the behavior of atoms, molecules and/or solutions, often include parameters that have been regressed to match the observed behavior of a set of systems (atoms, molecules and/or solutions). Therefore, computer aided property estimation also includes model development and model parameter estimation in addition to property estimation.

The essential steps in computer aided property estimation are to select an appropriate model for the desired property and to solve the corresponding model equations to estimate the property, provided all the necessary model parameters are available. A pre-estimation step is to derive or develop the needed property model, which can be time consuming and expensive. Therefore, the availability of computerized libraries of property models from which the needed model can be selected and/or adopted is commonly practiced. The method of solution of the model equations, of course, depends on the type of the model (that is, the set of equations representing the model). However, for single value property estimations involving simple polynomial functions, hand calculators may be used. For repetitive calculations of the same property and the same system within an iterative loop and/or same property for different systems, use of a computer-based calculation option is more appropriate. On the other side of the scale-size factor, for single value property estimations involving complex mathematical models (such as molecular modeling), the method of solution is computer intensive and availability of powerful computers is an advantage. Parameter estimation is necessary when the available parameters do not give acceptable results or are simply not available. In either case, experimental data must be available to fine-tune or regress the model parameters.

In this chapter, we will give an overview of computer aided property estimation in terms of types of properties, types of models and types of solution approaches together with a discussion on methods for model development.

1.2 CLASSIFICATION OF PROPERTIES & MODELS

Properties are classified in this chapter in terms of the scale-size factor, the function-use factor and the dependence factor according to a defined hierarchy. The scale-size factor is used at the inner-level, where, properties are classified in terms of scale & size into microscopic, mesoscopic and macroscopic. Microscopic properties refer to properties of atoms while macroscopic properties refer to properties of molecules. At each scale-size, the properties are further classified in terms of single atoms/molecules or multiple atoms/molecules of different types. At the next level, each type of property from the inner level is classified in terms of the function-use factor as physical, chemical, transport, environmental and so on. That is, the function and/or use of the property define its class. At the outer level, properties are classified as primary (single value property, which can be determined only from the structural information), secondary (property that cannot be explicitly calculated only from the structural information and is usually a function of other properties) or functional (properties that are dependent on the intensive variables, temperature, pressure and/or composition, in addition to the structural information). Figure 1 illustrates this outer-level classification.

Figure 1 Outer-level classification of properties.

A model is needed for estimation of any property from any level and sub-level. The type of model to be used depends on the three classification factors. At the macroscopic level, Eqs. 1-4 highlight examples of primary, secondary and functional property models, respectively.

c=231.239*Log(ΣklkLk+ΣiAiNi+ΣjBjMj)

  (1)

c=(Pc*Vc)/(83.14*Tc)

  (2)

vap=10[A−B/(T+C)]

  (3)

nγi=1−lnEi-(xi*δij/Ei+xj*δji/Ej)

  (4)

In the above models, Eq. 1 illustrates a 3rd-order group contribution based additive model for a single value primary pure component property (critical temperature - Tc) where each of the summation terms indicates group contribution for the corresponding group-order. The theoretical basis here is that the accuracy of property estimation increases with the increase in the number of additive terms. Equation 2 illustrates a secondary property of a pure component or mixture (critical compressibility factor, Zc, as a function of critical temperature, critical pressure and critical volume). Theoretically, if the primary property expressions are introduced into the expression for the secondary property, the corresponding secondary property can be estimated only from the molecular structural information and the parameters for the additive terms. Note, however, that the parameters are regressed (usually) only with the primary property experimental data. The secondary property is usually derived from observed relationships between properties of a class of systems. Equation 3 represents a temperature dependent correlation for the vapor pressure, Pvap, as a function of temperature, T. This class of properties change as the condition of the state (defined by the intensive properties – temperature, pressure and/or composition) change. Therefore, the behavior of systems under different conditions needs to be modeled. Equation 4 represents mixture property - activity coefficient for component i in a binary liquid mixture.

The models are classified in this chapter in terms of theoretical, semi-empirical and empirical (see Figure 2). Most of the commonly used property models belong to the semi-empirical type. Molecular modeling approaches belong to the theoretical class while correlation functions belong to the empirical type.

Figure 2 Classification of property estimation methods.

The main feature of all property models is that regressed values for a set of model parameters are needed in order to estimate the property from the model equations. If these parameters are available in the form of regressed values, then the estimation can be performed. Otherwise, the unavailable model parameter values will need to be regressed or another property model will need to be selected. In some cases, the property model parameter can be predicted, giving rise to truly predictive property models. Thus, property model development and use consists of a model parameter estimation and/or verification step. The application range of the property model, therefore, depends not only on the theoretical aspects (such as the behavior of the system) but also on the data used for the regression of the model parameters.

1.3 PROPERTY MODEL DEVELOPMENT

Development of models in general and property models in particular, is a cyclic process as shown in Figure 3. One starts with a definition of the model requirements and a hypothesis, solves the model equations, checks the calculated values against collected experimental data, and if the comparison is not favorable, changes the theory and/or the model parameters (thereby generating a new model) and repeats the cycle. As with other types of models, decisions need to be made in terms of the type of model to be developed for a specific property, the required accuracy, the application range (in terms of systems, conditions, problems), and the expected users. For many property model developments, these decisions are inter-related. For example, property model development, formulated as, develop a model for the estimation of the average density of polymers defines the type of model, the application range and the expected users. On the other hand, a property model development problem defined simply as “develop a model to predict the activity coefficients of liquid solutions” needs further information in terms of types of systems and conditions for which the model will be applicable. Otherwise, the problem definition implies a model applicable to all types of systems and under all conditions, which is almost impossible to develop with the current knowledge.

Figure 3 Steps in the development of property models.

The verification/validation step of property modeling should also check for thermodynamic consistency, such as the Gibbs-Duhem conditions (see Eqs. 5-7), the relation between the normal boiling point and critical temperature (see Eq. 8) and the condition that some property values have a...

Produkt-Details

• Erscheint lt. Verlag: 30.6.2004
• Sprache: englisch
• Themenwelt: Sachbuch/Ratgeber
• Themenwelt: Mathematik / Informatik ► Mathematik
• Themenwelt: Naturwissenschaften ► Biologie ► Botanik
• Themenwelt: Naturwissenschaften ► Chemie ► Physikalische Chemie
• Themenwelt: Naturwissenschaften ► Chemie ► Technische Chemie
• Themenwelt: Naturwissenschaften ► Physik / Astronomie ► Festkörperphysik
• Themenwelt: Naturwissenschaften ► Physik / Astronomie ► Thermodynamik
• Themenwelt: Technik ► Umwelttechnik / Biotechnologie
• Themenwelt: Wirtschaft ► Betriebswirtschaft / Management ► Logistik / Produktion
• ISBN-10: 0-08-047228-1 / 0080472281
• ISBN-13: 978-0-08-047228-7 / 9780080472287