Advanced Chemical Process Control: Putting Theory into Practice

✍ Scribed by Hovd M.

• Publisher: WILEY
• Year: 2023
• Tongue: English
• Leaves: 359
• Category: Library

✦ Synopsis

Advanced Chemical Process Control
Bridge the gap between theory and practice with this accessible guide
Process control is an area of study which seeks to optimize industrial processes, applying different strategies and technologies as required to navigate the variety of processes and their many potential challenges. Though the body of chemical process control theory is robust, it is only in recent decades that it has been effectively integrated with industrial practice to form a flexible toolkit. The need for a guide to this integration of theory and practice has therefore never been more urgent.
Advanced Chemical Process Control meets this need, making advanced chemical process control accessible and useful to chemical engineers with little grounding in the theoretical principles of the subject. It provides a basic introduction to the background and mathematics of control theory, before turning to the implementation of control principles in industrial contexts. The result is a bridge between the insights of control theory and the needs of engineers in plants, factories, research facilities, and beyond.
Advanced Chemical Process Control readers will also find:
Detailed overview of Control Performance Monitoring (CPM), Model Predictive Control (MPC), and more.
Discussion of the cost benefit analysis of improved control in particular jobs.
Authored by a leading international expert on chemical process control.
Advanced Chemical Process Control is essential for chemical and process engineers looking to develop a working knowledge of process control, as well as for students and graduates entering the chemical process control field.

✦ Table of Contents

Cover
Half Title
Advanced Chemical Process Control: Putting Theory into Practice
Copyright
Contents
Preface
The structure of the Book
What’s Not in the Book
References
Acknowledgments
Acronyms
Introduction
I.1 Why Is Chemical Process Control Needed?
I.2 What Knowledge Does a chemical Process Control Engineer need?
I.3 What Makes Chemical Process Control Unique?
I.4 The Structure of Control Systems in the Chemical Process Industries
I.5 Notation
1. Mathematical and Control Theory Background
1.1 Introduction
1.2 Models for Dynamical Systems
1.2.1 Dynamical Systems in Continuous Time
1.2.2 Dynamical Systems in Discrete Time
1.2.3 Linear Models and Linearization
1.2.3.1 Linearization at a Given Point
1.2.3.2 Linearizing Around a Trajectory
1.2.4 Converting Between Continuous- and Discrete-Time Models
1.2.4.1 Time Delay in the Manipulated Variables
1.2.4.2 Time Delay in the Measurements
1.2.5 Laplace Transform
1.2.6 The z Transform
1.2.7 Similarity Transformations
1.2.8 Minimal Representation
1.2.9 Scaling
1.3 Analyzing Linear Dynamical Systems
1.3.1 Transfer Functions of Composite Systems
1.3.1.1 Series Interconnection
1.3.1.2 Parallel Systems
1.3.1.3 Feedback Connection
1.3.1.4 Commonly Used Closed-Loop Transfer Functions
1.3.1.5 The Push-Through Rule
1.4 Poles and Zeros of Transfer Functions
1.4.1 Poles of Multivariable Systems
1.4.2 Pole Directions
1.4.3 Zeros of Multivariable Systems
1.4.4 Zero Directions
1.5 Stability
1.5.1 Poles and Zeros of Discrete-Time Transfer Functions
1.5.2 Frequency Analysis
1.5.2.1 Steady-State Phase Adjustment
1.5.3 Bode Diagrams
1.5.3.1 Bode Diagram Asymptotes
1.5.3.2 Minimum Phase Systems
1.5.3.3 Frequency Analysis for Discrete-Time Systems
1.5.4 Assessing Closed-Loop Stability Using the Open-Loop Frequency Response
1.5.4.1 The Principle of the Argument and the Nyquist D-Contour
1.5.4.2 The Multivariable Nyquist Theorem
1.5.4.3 The Monovariable Nyquist Theorem
1.5.4.4 The Bode Stability Criterion
1.5.4.5 Some Remarks on Stability Analysis Using the Frequency Response
1.5.4.6 The Small Gain Theorem
1.5.5 Controllability
1.5.6 Observability
1.5.7 Some Comments on Controllability and Observability
1.5.8 Stabilizability
1.5.9 Detectability
1.5.10 Hidden Modes
1.5.11 Internal Stability
1.5.12 Coprime Factorizations
1.5.12.1 Inner–Outer Factorization
1.5.12.2 Normalized Coprime Factorization
1.5.13 Parametrization of All Stabilizing Controllers
1.5.13.1 Stable Plants
1.5.13.2 Unstable Plants
1.5.14 Hankel Norm and Hankel Singular Values
Problems
References
2. Control Configuration and Controller Tuning
2.1 Common Control Loop Structures for the Regulatory Control Layer
2.1.1 Simple Feedback Loop
2.1.2 Feedforward Control
2.1.3 Ratio Control
2.1.4 Cascade Control
2.1.5 Auctioneering Control
2.1.6 Split Range Control
2.1.7 Input Resetting Control
2.1.8 Selective Control
2.1.9 Combining Basic Single-Loop Control Structures
2.1.10 Decoupling
2.2 Input and Output Selection
2.2.1 Using Physical Insights
2.2.2 Gramian-Based Input and Output Selection
2.2.3 Input/Output Selection for Stabilization
2.3 Control Configuration
2.3.1 The Relative Gain Array
2.3.2 The RGA as a General Analysis Tool
2.3.2.1 The RGA and Zeros in the Right Half-Plane
2.3.2.2 The RGA and the Optimally Scaled Condition Number
2.3.2.3 The RGA and Individual Element Uncertainty
2.3.2.4 RGA and Diagonal Input Uncertainty
2.3.2.5 The RGA as an Interaction Measure
2.3.3 The RGA and Stability
2.3.3.1 The RGA and Pairing of Controlled and Manipulated Variables
2.3.4 Summary of RGA-Based Input–Output Pairing
2.3.5 Partial Relative Gains
2.3.6 The Niederlinski Index
2.3.7 The Rijnsdorp Interaction Measure
2.3.8 Gramian-Based Input–Output Pairing
2.3.8.1 The Participation Matrix
2.3.8.2 The Hankel Interaction Index Array
2.3.8.3 Accounting for the Closed-Loop Bandwidth
2.4 Tuning of Decentralized Controllers
2.4.1 Introduction
2.4.2 Loop Shaping Basics
2.4.3 Tuning of Single-Loop Controllers
2.4.3.1 PID Controller Realizations and Common Modifications
2.4.3.2 Controller Tuning Using Frequency Analysis
2.4.3.3 Ziegler–Nichols Closed-Loop Tuning Method
2.4.3.4 Simple Fitting of a Step Response Model
2.4.3.5 Ziegler–Nichols Open-Loop Tuning
2.4.3.6 IMC-PID Tuning
2.4.3.7 Simple IMC Tuning
2.4.3.8 The Setpoint Overshoot Method
2.4.3.9 Autotuning
2.4.3.10 When Should Derivative Action Be Used?
2.4.3.11 Effects of Internal Controller Scaling
2.4.3.12 Reverse Acting Controllers
2.4.4 Gain Scheduling
2.4.5 Surge Attenuating Controllers
2.4.6 Multiloop Controller Tuning
2.4.6.1 Independent Design
2.4.6.2 Sequential Design
2.4.6.3 Simultaneous Design
2.4.7 Tools for Multivariable Loop-Shaping
2.4.7.1 The Performance Relative Gain Array
2.4.7.2 The Closed-Loop Disturbance Gain
2.4.7.3 Illustrating the Use of CLDG’s for Controller Tuning
2.4.7.4 Unachievable Loop Gain Requirements
Problems
References
3. Control Structure Selection and Plantwide Control
3.1 General Approach and Problem Decomposition
3.1.1 Top-Down Analysis
3.1.1.1 Defining and Exploring Optimal Operation
3.1.1.2 Determining Where to Set the Throughput
3.1.2 Bottom-Up Design
3.2 Regulatory Control
3.2.1 Example: Regulatory Control of Liquid Level in a Deaeration Tower
3.3 Determining Degrees of Freedom
3.4 Selection of Controlled Variables
3.4.1 Problem Formulation
3.4.2 Selecting Controlled Variables by Direct Evaluation of Loss
3.4.3 Controlled Variable Selection Based on Local Analysis
3.4.3.1 The Minimum Singular Value Rule
3.4.3.2 Desirable Characteristics of the Controlled Variables
3.4.4 An Exact Local Method for Controlled Variable Selection
3.4.5 Measurement Combinations as Controlled Variables
3.4.5.1 The Nullspace Method for Selecting Controlled Variables
3.4.5.2 Extending the Nullspace Method to Account for Implementation Error
3.4.6 The Validity of the Local Analysis for Controlled Variable Selection
3.5 Selection of Manipulated Variables
3.5.1 Verifying that the Proposed Manipulated Variables Make Acceptable Control Possible
3.5.2 Reviewing the Characteristics of the Proposed Manipulated Variables
3.6 Selection of Measurements
3.7 Mass Balance Control and Throughput Manipulation
3.7.1 Consistency of Inventory Control
Problems
References
4. Limitations on Achievable Performance
4.1 Performance Measures
4.1.1 Time-Domain Performance Measures
4.1.2 Frequency-Domain Performance Measures
4.1.2.1 Bandwidth Frequency
4.1.2.2 Peaks of Closed-Loop Transfer Functions
4.1.2.3 Bounds onWeighted System Norms
4.1.2.4 Gain and Phase Margin
4.2 Algebraic Limitations
4.2.1 S + T = I
4.2.2 Interpolation Constraints
4.2.2.1 Monovariable Systems
4.2.2.2 Multivariable Systems
4.3 Control Performance in Different Frequency Ranges
4.3.1 Sensitivity Integrals and Right Half-Plane Zeros
4.3.1.1 Multivariable Systems
4.3.2 Sensitivity Integrals and Right Half-Plane Poles
4.3.3 Combined Effects of RHP Poles and Zeros
4.3.4 Implications of the Sensitivity Integral Results
4.4 Bounds on Closed-Loop Transfer Functions
4.4.1 The Maximum Modulus Principle
4.4.1.1 The Maximum Modulus Principle
4.4.2 Minimum Phase and Stable Versions of the Plant
4.4.3 Bounds on S and T
4.4.3.1 Monovariable Systems
4.4.3.2 Multivariable Systems
4.4.4 Bounds on KS and KSGd
4.5 ISE Optimal Control
4.6 Bandwidth and Crossover Frequency Limitations
4.6.1 Bounds from ISE Optimal Control
4.6.2 Bandwidth Bounds from Weighted Sensitivity Minimization
4.6.3 Bound from Negative Phase
4.7 Bounds on the Step Response
4.8 Bounds for Disturbance Rejection
4.8.1 Inputs for Perfect Control
4.8.2 Inputs for Acceptable Control
4.8.3 Disturbances and RHP Zeros
4.8.4 Disturbances and Stabilization
4.9 Limitations from Plant Uncertainty
4.9.1 Describing Uncertainty
4.9.2 Feedforward Control and the Effects of Uncertainty
4.9.3 Feedback and the Effects of Uncertainty
4.9.4 Bandwidth Limitations from Uncertainty
Problems
References
5. Model-Based Predictive Control
5.1 Introduction
5.2 Formulation of a QP Problem for MPC
5.2.1 Future States as Optimization Variables
5.2.2 Using the Model Equation to Substitute for the Plant States
5.2.3 Optimizing Deviations from Linear State Feedback
5.2.4 Constraints Beyond the End of the Prediction Horizon
5.2.5 Finding the Terminal Constraint Set
5.2.6 Feasible Region and Prediction Horizon
5.3 Step-Response Models
5.4 Updating the Process Model
5.4.1 Bias Update
5.4.2 Kalman Filter and Extended Kalman Filters
5.4.2.1 Augmenting a Disturbance Description
5.4.2.2 The Extended Kalman Filter
5.4.2.3 The Iterated Extended Kalman Filter
5.4.3 Unscented Kalman Filter
5.4.4 Receding Horizon Estimation
5.4.4.1 The Arrival Cost
5.4.4.2 The Filtering Formulation of RHE
5.4.4.3 The Smoothing Formulation of RHE
5.4.5 Concluding Comments on State Estimation
5.5 Disturbance Handling and Offset-Free Control
5.5.1 Feedforward from Measured Disturbances
5.5.2 Requirements for Offset-Free Control
5.5.3 Disturbance Estimation and Offset-Free Control
5.5.4 Augmenting the Model with Integrators at the Plant Input
5.5.5 Augmenting the Model with Integrators at the Plant Output
5.5.6 MPC and Integrator Resetting
5.6 Feasibility and Constraint Handling
5.7 Closed-Loop Stability with MPC Controllers
5.8 Target Calculation
5.9 Speeding up MPC Calculations
5.9.1 Warm-Starting the Optimization
5.9.2 Input Blocking
5.9.3 Enlarging the Terminal Region
5.10 Robustness of MPC Controllers
5.11 Using Rigorous Process Models in MPC
5.12 Misconceptions, Clarifications, and Challenges
5.12.1 Misconceptions
5.12.1.1 MPC Is Not Good for Performance
5.12.1.2 MPC Requires Very Accurate Models
5.12.1.3 MPC Cannot Prioritize Input Usage or Constraint Violations
5.12.2 Challenges
5.12.2.1 Obtaining a Plant Model
5.12.2.2 Maintenance
5.12.2.3 Capturing the Desired Behavior in the MPC Design
Problems
References
6. Some Practical Issues in Controller Implementation
6.1 Discrete-Time Implementation
6.1.1 Aliasing
6.1.2 Sampling Interval
6.1.3 Execution Order
6.2 Pure Integrators in Parallel
6.3 Anti-Windup
6.3.1 Simple PI Control Anti-Windup
6.3.2 Velocity Form of PI Controllers
6.3.3 Anti-Windup in Cascaded Control Systems
6.3.4 A General Anti-Windup Formulation
6.3.5 Hanus’ Self-Conditioned Form
6.3.6 Anti-Windup in Observer-Based Controllers
6.3.7 Decoupling and Input Constraints
6.3.8 Anti-Windup for “Normally Closed” Controllers
6.4 Bumpless Transfer
6.4.1 Switching Between Manual and Automatic Operation
6.4.2 Changing Controller Parameters
Problems
References
7. Controller Performance Monitoring and Diagnosis
7.1 Introduction
7.2 Detection of Oscillating Control Loops
7.2.1 The Autocorrelation Function
7.2.2 The Power Spectrum
7.2.3 The Method of Miao and Seborg
7.2.4 The Method of Hägglund
7.2.5 The Regularity Index
7.2.6 The Method of Forsman and Stattin
7.2.7 Prefiltering Data
7.3 Oscillation Diagnosis
7.3.1 Manual Oscillation Diagnosis
7.3.2 Detecting and Diagnosing Valve Stiction
7.3.2.1 Using the Cross-Correlation Function to Detect Valve Stiction
7.3.2.2 Histograms for Detecting Valve Stiction
7.3.2.3 Stiction Detection Using an OP–PV Plot
7.3.3 Stiction Compensation
7.3.4 Detection of Backlash
7.3.5 Backlash Compensation
7.3.6 Simultaneous Stiction and Backlash Detection
7.3.7 Discriminating Between External and Internally Generated Oscillati
7.3.8 Detecting and Diagnosing Other Nonlinearities
7.4 Plantwide Oscillations
7.4.1 Grouping Oscillating Variables
7.4.1.1 Spectral Principal Component Analysis
7.4.1.2 Visual Inspection Using High-Density Plots
7.4.1.3 Power Spectral Correlation Maps
7.4.1.4 The Spectral Envelope Method
7.4.1.5 Methods Based on Adaptive Data Analysis
7.4.2 Locating the Cause for Distributed Oscillations
7.4.2.1 Using Nonlinearity for Root Cause Location
7.4.2.2 The Oscillation Contribution Index
7.4.2.3 Estimating the Propagation Path for Disturbances
7.5 Control Loop Performance Monitoring
7.5.1 The Harris Index
7.5.2 Obtaining the Impulse Response Model
7.5.3 Calculating the Harris Index
7.5.4 Estimating the Deadtime
7.5.5 Modifications to the Harris Index
7.5.6 Assessing Feedforward Control
7.5.7 Comments on the Use of the Harris Index
7.5.8 Performance Monitoring for PI Controllers
7.6 Multivariable Control Performance Monitoring
7.6.1 Assessing Feedforward Control in Multivariable Control
7.6.2 Performance Monitoring for MPC Controllers
7.7 Some Issues in the Implementation of Control Performance Monitoring
7.8 Discussion
Problems
References
8. Economic Control Benefit Assessment
8.1 Optimal Operation and Operational Constraints
8.2 Economic Performance Functions
8.3 Expected Economic Benefit
8.4 Estimating Achievable Variance Reduction
8.5 Worst-Case Backoff Calculation
A. Fourier–Motzkin Elimination
B. Removal of Redundant Constraints
Reference
C. The Singular Value Decomposition
D. Factorization of Transfer Functions into Minimum Phase Stable and All-Pass Parts
D.1 Input Factorization of RHP Zeros
D.2 Output Factorization of RHP Zeros
D.3 Output Factorization of RHP Poles
D.4 Input Factorization of RHP Poles
D.5 SISO Systems
D.6 Factoring Out Both RHP Poles and RHP Zeros
Reference
E. Models Used in Examples
E.1 Binary Distillation Column Model
E.2 Fluid Catalytic Cracker Model
References
Index