Model-Based Optimization for Petroleum Refinery Configuration Design

Model-Based Optimization for Petroleum Refinery Configuration Design An accessible, easy-to-read introduction to the methods of mixed-integer optimization, with practical applications, real-world operational data, and case studies Interest in model-based approaches for optimizing the design of petroleum refineries has increased throughout the industry in recent years. Mathematical optimization based on mixed-integer programming has brought about the superstructure optimization method for synthesizing petroleum refinery configurations from multiple topological alternatives. Model-Based Optimization for Petroleum Refinery Configuration Design presents a detailed introduction to the use of mathematical optimization to solve both linear and nonlinear problems in the refining industry. The book opens with an overview of petroleum refining processes, basic concepts in mathematical programming, and applications of mathematical programming for refinery optimization. Subsequent chapters address superstructure representations of topological alternatives, mathematical formulation, solution strategies, and various modeling frameworks. Practical case studies demonstrate refinery configuration design, refinery retrofitting, and real-world issues and considerations. Presents linear, nonlinear, and mixed-integer programming approaches applicable to both new and existing petroleum refineries Highlights the benefits of model-based solutions to refinery configuration design problems Features detailed case studies of the development and implementation of optimization models Discusses economic considerations of heavy oil processing, including cash flow analysis of refinery costs and return on capital Includes numerical examples based on real-world operational data and various commercial technologies Model-Based Optimization for Petroleum Refinery Configuration Design is an invaluable resource for researchers, chemical engineers, process and energy engineers, other refining professionals, and advanced chemical engineering students.

Table of Contents:
1 Introduction to Optimization Modeling for Petroleum Refineries 1 1.1 Background 1 1.2 Overview of Refining Processes 4 1.2.1 Atmospheric Crude Oil Distillation 5 1.2.2 Hydroprocessing 5 1.2.3 Sulfur Recovery 9 1.2.4 Reforming 9 1.2.5 Isomerization 10 1.2.6 Blending 11 1.3 Overview of Refinery Optimization Modeling 12 1.3.1 Refinery Optimization Systems, Techniques, and Tools 12 1.3.2 Modeling for Advanced Process Control 14 1.3.3 Modeling for Real-Time Optimization 15 1.3.4 Modeling for Process Simulation 17 1.3.4.1 Modeling for Dynamic Simulation 18 1.3.4.2 Modeling for Operator Training Simulation 19 1.3.5 Modeling for Planning and Scheduling 19 1.3.5.1 Systems Implementation 23 1.3.5.2 Optimization of Crude Oil Scheduling 24 1.3.5.3 Refinery Management 25 1.4 Concluding Remarks 25 References 26 2 Basic Petroleum Refinery Economics 31 2.1 Refinery Economics Overview 31 2.1.1 Refinery Profitability 31 2.1.2 Refinery Margins 32 2.1.3 Refinery Margin Calculations 33 2.1.4 Refinery Margin Trends 34 2.1.5 Refinery Margin Improvement 34 2.2 Marginal Economics for Incremental Optimization 34 2.3 Refinery Economic Analysis 36 2.3.1 Refinery Value Determination 36 2.3.2 Refinery Economic Evaluation 37 2.3.2.1 Simple Example 37 2.3.2.2 Advanced Example 38 2.3.2.3 Further Example 40 2.3.3 Refinery Contracts 41 2.4 Concluding Remarks 41 References 41 3 Superstructure Representation 43 3.1 Introduction to Superstructures 43 3.2 Types of Superstructure Representation 43 3.3 State–Task Network Superstructure Representation 44 3.4 State–Equipment Network Superstructure Representation 45 3.5 Resource–Task Network Superstructure Representation 46 3.6 Superstructure Generation 47 3.7 Other Superstructure Representations 48 3.7.1 State–Space Network Superstructure Representation 48 3.7.2 Unit Operation–Port–State Superstructure Representation 48 3.7.3 Bond Graph Superstructure Representation 48 3.8 Superstructure Representation Example for Naphtha Processing 49 3.9 Chapter Summary 53 References 53 4 Modeling Framework 57 4.1 Modeling of Mixed Continuous and Integer Decision Variables 57 4.2 Superstructure Optimization Modeling 58 4.3 Constructing Superstructures 58 4.4 Modeling of Superstructure Representations 59 4.5 Modeling of Discrete Decisions and Logical Relations 60 4.5.1 Propositional Logics for Superstructure Optimization Modeling 61 4.5.2 Logical Binary Variables 62 4.5.3 Yes/No Type Binary Variables 62 4.5.4 Disjunctive Optimization Modeling 63 4.6 Modeling of Process Units and Operations 67 4.6.1 Process Design Procedure 67 4.6.2 Selecting Modeling Variables 67 4.6.3 Formulating Simple Models 68 4.6.4 Basic Unit Models 68 4.6.4.1 Mixer 68 4.6.4.2 Splitter 69 4.6.4.3 Separator 70 4.6.4.4 Valve 70 4.6.4.5 Multicomponent Splitter 70 4.6.5 Unit Operation Models 72 4.6.5.1 Compressor 72 4.6.5.2 Furnace 72 4.6.5.3 Conversion Reactor 72 4.6.5.4 Heat Exchanger 75 4.6.6 Information Flow Modeling 75 4.6.6.1 Information Flow Diagram 77 4.6.6.2 Choice of Design Variables 79 4.6.6.3 Equation Ordering 79 4.7 Modeling for Numerical Studies 84 4.8 Chapter Summary 86 References 86 5 Model Formulation and Implementation 89 5.1 Mathematical Formulation 89 5.2 Generic Optimization Model Formulation for Refinery Planning 90 5.2.1 Objective Function 91 5.2.2 Production Capacity and Expansion Constraints 91 5.2.3 Mass Balances 92 5.2.4 Demand Constraints 92 5.2.5 Availability Constraints 92 5.2.6 Non-Negativity Constraints 92 5.3 Generic Optimization Model Formulation for Refinery Design 93 5.3.1 Material Balances 93 5.3.2 Mixed-Integer Logical Constraints 93 5.3.3 Logical Constraints on Design and Structural Specifications 94 5.3.4 Logic Propositional Constraints on Design Specifications 95 5.3.4.1 Example 1 95 5.3.4.2 Example 2 100 5.3.5 Logic Propositional Constraints on Structural Specifications 101 5.3.6 Generalized Disjunctive Programming 101 5.4 Numerical Implementation for Computational Experiments 106 5.5 Computational Experiment Examples 110 5.5.1 MILP Model Results 113 5.5.2 GDP Model Results 114 5.6 Chapter Summary 123 References 123 6 Solution Strategies 125 6.1 Convex Relaxation 125 6.2 Lagrangean Decomposition 126 6.3 Global Optimization Techniques 126 6.3.1 Branch and Reduce 128 6.3.2 Spatial Branch and Bound 128 6.3.3 Hybrid Branch and Bound 128 6.3.4 Interval Analysis 129 6.3.5 Extended Cutting Plane 129 6.4 Advancements in Commercial Integer Optimization Solvers 130 6.4.1 Overview 130 6.4.2 Computational Performance of Commercial Integer Optimization Solvers 130 6.4.3 A Commercial Success Story: CPLEX Integer Optimization Solver 130 6.4.4 Solution Methods and Algorithms 131 6.4.4.1 Integer Optimization Algorithms 131 6.4.4.2 Branch and Bound 132 6.4.4.3 Presolve and Cutting Planes 134 6.4.4.4 Heuristics 135 6.4.4.5 Combined Local Search and Heuristics 136 6.4.4.6 Parallelization 136 6.4.4.7 Solution Pools 136 6.4.4.8 Tuning Tools 136 6.4.5 Application Examples 136 6.4.5.1 Example 1: Energy Optimization 137 6.4.5.2 Example 2: Financial Optimization 137 6.4.5.3 Example 3: Manufacturing Optimization 137 6.4.5.4 Concluding Remarks 138 6.5 Chapter Summary 139 References 139 7 Industrial Case Studies with Business-Centric Techno-Commercial Considerations 145 7.1 Industrial Case Study 1: Refinery Configuration for Heavy Oil Processing 145 7.1.1 Background 145 7.1.2 Problem Statement 146 7.1.3 Model Formulation 147 7.1.4 Numerical Example 148 7.1.5 Concluding Remarks 151 7.2 Industrial Case Study 2: Refinery Configuration for Whole Complex Processing 152 7.2.1 Model Formulation 152 7.2.1.1 Superstructure Representation 156 7.2.1.2 Logic Propositions 162 7.2.1.3 Objective Function 164 7.2.2 Computational Results 165 7.2.2.1 Computational Results and Discussion 166 7.2.2.2 Model Validation 171 7.2.2.3 Application Extension to Refinery Upgrade Studies 176 7.2.2.4 Sensitivity Analysis 176 7.2.3 Concluding Remarks 176 7.3 Industrial Case Study 3: Refinery Configuration for Naphtha Upgrading 177 7.3.1 Problem Statement 178 7.3.2 Propositional Logics and Logic Cuts in Process Synthesis Problems 178 7.3.3 Logical Constraints 178 7.3.3.1 General Formulation 178 7.3.3.2 Logical Constraints on Processing Alternatives of Naphtha for Petroleum Refineries 182 7.3.4 Computational Experience 182 7.3.5 Concluding Remarks 183 7.4 Chapter Summary 186 References 186 8 Industrial Case Studies with Environmental-Centric Techno-Commercial Considerations 191 8.1 Industrial Case Study 1: Refinery Configuration with Environmental Considerations 191 8.1.1 Background 191 8.1.2 Problem Statement 192 8.1.3 Model Formulation 192 8.1.3.1 Superstructure Representation 192 8.1.3.2 Material Balance Constraints 192 8.1.3.3 Logical Constraints 194 8.1.3.4 Logic Propositions 194 8.1.3.5 Environmental Performance Assessment for Risk Evaluation of Flowsheets 196 8.1.3.6 Objective Function 197 8.1.4 Numerical Example 197 8.1.5 Concluding Remarks 198 8.2 Industrial Case Study 2: Refinery Configuration with Heat Integration 198 8.2.1 Problem Statement 198 8.2.2 Superstructure Representation 199 8.2.3 Modeling and Computational Strategy 201 8.2.4 Model Formulation 202 8.2.4.1 Flowsheet Optimization 202 8.2.4.2 Heat Integration Constraints 205 8.2.4.3 Objective Function 206 8.2.5 Computational Results 206 8.2.6 Concluding Remarks 209 8.3 Chapter Summary 211 References 212 9 Industrial Case Studies with Engineering-Centric Techno-Commercial Considerations 215 9.1 Industrial Case Study 1: Refinery Configuration for High-Octane Fuel Production 215 9.1.1 Catalytic Reforming Process 216 9.1.2 Data Reconciliation Method 216 9.1.3 Problem Statement 217 9.1.4 Model Formulation 217 9.1.4.1 Data Reconciliation Model 218 9.1.4.2 Feed Characterization 219 9.1.4.3 Reactor Representation 220 9.1.4.4 Reactor Pressure Balance 221 9.1.4.5 Reaction Kinetic Tuning 221 9.1.4.6 Reactor Switch in Cyclic Reformer 221 9.1.4.7 Measurement Models 223 9.1.5 Results and Discussion 224 9.1.5.1 Key Process Variables 224 9.1.5.2 Tuning Strategies 225 9.1.5.3 Reformate Yields 226 9.1.5.4 Reactor Total Endotherms 226 9.1.6 Concluding Remarks 226 9.2 Industrial Case Study 2: Refinery Configuration for Low-Benzene Fuel Production 227 9.2.1 Problem Statement 227 9.2.2 Superstructure Representation 227 9.2.3 Model Formulation 229 9.2.4 Preliminary Computational Results 234 9.3 Chapter Summary 234 References 234 Summary and Conclusions 237 Index 239

About the Author :
Cheng Seong Khor, PhD, is formerly a senior lecturer at Chemical Engineering Department, Universiti Teknologi PETRONAS (UTP), Malaysia and now a data scientist at PETRONAS.