Cover image for Diagnosis of process nonlinearities and valve stiction : data driven approaches
Title:
Diagnosis of process nonlinearities and valve stiction : data driven approaches
Personal Author:
Series:
Advances in industrial control
Publication Information:
Berlin, GW : Springer, 2008
Physical Description:
xx, 284 p. : ill. ; 24 cm.
ISBN:
9783540792239

Available:*

Library
Item Barcode
Call Number
Material Type
Item Category 1
Status
Searching...
30000010194027 TJ223.V3 C46 2008 Open Access Book Book
Searching...

On Order

Summary

Summary

were published in the series as the contributed volume, Process Control Performance Assessment: From Theory to Implementation with Andrzej Ordys, Damian Uduehi, and Michael Johnson as Editors (ISBN 978-1-84628-623-0, 2007). Along with this good progress in process controller assessment methods, researchers have also been investigating techniques to diagnose what is causing the process or control loop degradation. This requires the use of on-line data to identify faults via new diagnostic indicators of typical process problems. A significant focus of some of this research has been the issue of valve problems; a research direction that has been motivated by some industrial statistics that show up to 40% of control loops having performance degradation attributable to valve problems. Shoukat Choudhury, Sirish Shah, and Nina Thornhill have been very active in this research field for a number of years and have written a coherent and consistent presentation of their many research results as this monograph, Diagnosis of Process Nonlinearities and Valve Stiction. The Advances in Industrial Control series is pleased to welcome this new and substantial contribution to the process diagnostic literature. The reader will find the exploitation of the extensive process data archives created by today's process computer systems one theme in the monograph. From another viewpoint, the use of higher-order statistics could be considered to provide a continuing link to the earlier methods of the statistical process control paradigm.


Author Notes

M. A. A. Shoukat Choudhury received his B. Sc. Engineering (Chemical) from Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh in 1996. He was awarded a gold medal for his outstanding results in B. Sc. Engineering. He obtained an M. Sc. Engineering (Chemical) in 1998 from the same university. He has completed his PhD degree in process control (Chemical Engineering) at the University of Alberta, Canada. For his outstanding research performance during the course of PhD program he has been awarded several awards such as University of Alberta PhD Dissertation Fellowship, Andrew Stewart Memorial Prize and ISA Educational Foundation Scholarship. He is the principal inventor of an internation patent (applied, 2005) on "Methods for Detection and Quantification of Control Valve Stiction". The methodologies and algorithms described in this patent are implemented and available in the commercial software ProcessDoctor from Matrikon Inc. His main research interests include diagnosis of poor control performance, stiction in control valves, data compression, control loop performance assessment and monitoring, and diagnosis of plant wide oscillations.

Sirish Shah received his B.Sc. degree in control engineering from Leeds University in 1971, a M.Sc. degree in automatic control from UMIST, Manchester in 1972, and a Ph.D. degree in process control (chemical engineering) from the University of Alberta in 1976. During 1977 he worked as a computer applications engineer at Esso Chemicals in Sarnia, Ontario. Since 1978 he has been with the University of Alberta, where currently holds the NSERC-Matrikon-ASRA Senior Industrial Research Chair in Computer Process Control. In 1989, he was the recipient of the Albright & Wilson Americas Award of the Canadian Society for Chemical Engineering in recognition of distinguished contributions to chemical engineering. He has held visiting appointments at Oxford University and Balliol College as a SERC fellow in 1985-86 and at Kumamoto University, Japan as a senior research fellow of the Japan Society for the Promotion of Science (JSPS) in 1994. The main area of his current research is process and performance monitoring, system identification and design and implementation of softsensors. He has recently co-authored a book titled, Performance Assessment of Control Loops: Theory and Applications. He has held consulting appointments with a wide variety of process Industries and has also taught many industrial courses.


Table of Contents

1 Introductionp. 1
1.1 Concepts in Data-Driven Analysis of Chemical Processesp. 2
1.1.1 Linear and Nonlinear Time Seriesp. 3
1.1.2 Statistics and Randomnessp. 3
1.1.3 Frequency Content and Spectral Methodsp. 6
1.2 Nonlinearity in Control Valvesp. 8
1.3 The Layout of the Bookp. 10
1.3.1 Part I Higher-Order Statisticsp. 10
1.3.2 Part II Data Quality - Compression and Quantizationp. 10
1.3.3 Part III Nonlinearity and Control Performancep. 11
1.3.4 Part IV Control Valve Stiction - Definition, Modelling, Detection and Quantificationp. 12
1.3.5 Part V Plant-wide Oscillations - Detection and Diagnosisp. 13
1.3.6 Referencesp. 14
1.4 Summaryp. 14
Part I Higher-Order Statistics
2 Higher-Order Statistics: Preliminariesp. 17
2.1 Introductionp. 17
2.2 Time Domain Analysisp. 18
2.2.1 Momentsp. 18
2.2.2 Cumulantsp. 20
2.2.3 The Relationship Between Moments and Cumulantsp. 22
2.2.4 Properties of Moments and Cumulantsp. 22
2.2.5 Moments and Cumulants of Stationary Signalsp. 25
2.3 Spectral Analysisp. 25
2.3.1 Power Spectrum, n=2p. 26
2.3.2 Bispectrum, n=3p. 27
2.4 Summaryp. 28
3 Bispectrum and Biocherencep. 29
3.1 Bispectrump. 29
3.1.1 Estimation of the Bispectrump. 30
3.1.2 Properties of Estimators and Asymptotic Behaviourp. 32
3.1.3 Bicoherence or Normalized Bispectrump. 34
3.1.4 Properties of Bispectrum and Bicoherencep. 35
3.2 Bispectrum or Bicoherence Estimation Issuesp. 37
3.2.1 Choice of Window Functionp. 38
3.2.2 Choice of Data Length, Segment Length and Fourier Transform Lengthp. 40
3.3 Summaryp. 41
Part II Data Quality - Compression and Quantization
4 Impact of Data Compression and Quantization on Data-Driven Process Analysesp. 45
4.1 Introductionp. 45
4.2 Data Compression Methodsp. 47
4.2.1 Overview of Data Compressionp. 47
4.2.2 Box-Car (BC) Algorithmp. 47
4.2.3 Backward-Slope (BS) Algorithmp. 47
4.2.4 Combined Box-Car and Backward-Slope (BCBS) Methodp. 49
4.2.5 Swinging Door Compression Algorithmp. 49
4.2.6 The Compression Factorp. 49
4.3 Measures of Data Qualityp. 50
4.3.1 Statistical Measuresp. 50
4.3.2 Nonlinearity Measuresp. 51
4.3.3 Performance Index (Harris) Measuresp. 51
4.4 Process Data for Compression Comparisonp. 52
4.4.1 Industrial Example 1p. 52
4.4.2 Industrial Example 2p. 55
4.5 Results and Discussions for Industrial Example 2p. 56
4.5.1 Visual Observationsp. 56
4.5.2 Statistical Propertiesp. 57
4.5.3 Nonlinearity Assessmentp. 58
4.5.4 Performance (Harris) Indexp. 58
4.6 Summary of Data Quality Measuresp. 59
4.7 Automated Detection of Compressionp. 59
4.7.1 Motivationp. 59
4.7.2 Compression Detection Procedurep. 60
4.7.3 Implementation Considerationsp. 61
4.8 A Recommendation for Harmless Storing of Datap. 63
4.9 Quantizationp. 63
4.10 Summaryp. 65
Part III Nonlinearity and Control Performance
5 Measures of Nonlinearity - A Reviewp. 69
5.1 Definition of Nonlinear Systemsp. 69
5.2 Nonlinearity in Process Time Trendsp. 70
5.3 Various Measures of Nonlinearityp. 70
5.3.1 Model-Based Measures of Nonlinearityp. 71
5.3.2 Time Series-Based Measures of Nonlinearityp. 71
5.4 Summaryp. 75
6 Linear or Nonlinear? A Bicoherence-Based Measure of Nonlinearityp. 77
6.1 Introductionp. 77
6.2 Bispectrum and Biocherencep. 78
6.2.1 Spurious Peaks in the Estimated Bicoherencep. 78
6.2.2 Illustrative Example 1p. 79
6.2.3 How to Choose [epsilon]?p. 80
6.3 Test of Gaussianity and Linearity of a Signalp. 81
6.3.1 Total Nonlinearity Index (TNLI)p. 85
6.4 Illustrative Example 2: Bicoherence of a Linear and a Nonlinear Signalp. 85
6.5 Illustrative Example 3: Bicoherence of a Nonlinear Sinusoid Signal with Noisep. 87
6.5.1 Mild Nonlinearity (n[subscript l] = 0.05)p. 88
6.5.2 Strong Nonlinearity (n[subscript l] = 0.25)p. 89
6.5.3 Extent of Nonlinearity and Effect of Noisep. 90
6.6 Summaryp. 91
7 A Nonlinearity Measure Based on Surrogate Data Analysisp. 93
7.1 Surrogate Time Seriesp. 93
7.1.1 Nonlinearity Detection Using Surrogatesp. 93
7.1.2 Predictability in Nonlinear Time Seriesp. 93
7.2 Algorithm for Nonlinearity Diagnosisp. 95
7.2.1 Construction of the Data Matrix for Nonlinear Predictionp. 95
7.2.2 Calculation of Prediction Errorp. 96
7.2.3 Calculation of Surrogate Datap. 96
7.2.4 Statistical Testingp. 98
7.2.5 Algorithm Summaryp. 98
7.3 Selection of the Parameter Valuesp. 99
7.3.1 Recommended Default Parameter Valuesp. 99
7.3.2 Choice of Embedding Parameters E and Hp. 99
7.3.3 Choice of Parameters C and kp. 100
7.3.4 Default Data Ensemble Size, Q and Number of Samples Per Feature, Sp. 101
7.3.5 Choice of the Number of Surrogates, Mp. 101
7.4 Data-Preprocessing and End-Matchingp. 102
7.4.1 False-Positive Results with Cyclic Datap. 102
7.4.2 End-Matchingp. 103
7.4.3 Summary of the Data-Preprocessing Stepsp. 104
7.4.4 Application to Oscillating Time Grendsp. 104
7.5 Worked Examplesp. 106
7.5.1 Identification of Nonlinear Root Causesp. 106
7.5.2 Application to the SE Asia Data Setp. 106
7.5.3 The Mechanisms of Propagation in the SE Asia Processp. 106
7.5.4 An Example with No Nonlinearityp. 108
7.6 Summaryp. 110
8 Nonlinearities in Control Loopsp. 111
8.1 Process Nonlinearityp. 111
8.1.1 Nonlinearity of a Spherical Tankp. 111
8.1.2 Nonlinearities of a Continuous Stirred Tank Reactor (CSTR)p. 115
8.2 Nonlinear Valve Characteristicp. 117
8.2.1 Linear Valvesp. 118
8.2.2 Equal Percentage Valvesp. 118
8.2.3 Square-Root Valvep. 119
8.2.4 Remarks on Nonlinear Valve Characteristicp. 120
8.3 Nonlinear Disturbancesp. 121
8.4 Summaryp. 121
9 Diagnosis of Poor Control Performancep. 123
9.1 Introductionp. 123
9.2 Problem Descriptionp. 124
9.3 Usual Causes of Poor Performancep. 125
9.4 Diagnosis of Poor Control Performancep. 126
9.4.1 Well Tuned Controllerp. 127
9.4.2 Tightly Tuned Controller or Excessive Integral Actionp. 128
9.4.3 Presence of an External Oscillatory Disturbancep. 129
9.4.4 Presence of Stictionp. 129
9.5 Industrial Case Studiesp. 129
9.5.1 Stiction in a Furnace Dryer Temperature Control Valvep. 130
9.5.2 Valve Saturationp. 131
9.5.3 Valve Problems in Some Flow Control Loopsp. 132
9.6 Summaryp. 134
Part IV Control Valve Stiction - Definition, Modelling, Detection and Quantification
10 Different Types of Faults in Control Valvesp. 137
10.1 What Is a Control Valvep. 137
10.2 Faults in Control Valvep. 138
10.2.1 Oversized Valvep. 139
10.2.2 Undersized Valvep. 139
10.2.3 Corroded Valve Seatp. 139
10.2.4 Faulty Diaphragmp. 139
10.2.5 Packing Leakagep. 139
10.2.6 Valve Hysteresisp. 140
10.2.7 Valve Stictionp. 140
10.2.8 Large Deadbandp. 140
10.2.9 Valve Saturationp. 141
10.3 Summaryp. 141
11 Stiction: Definition and Discussionsp. 143
11.1 Introductionp. 143
11.2 What Is Stiction?p. 143
11.2.1 Definition of Terms Relating to Valve Nonlinearityp. 144
11.2.2 Discussion of the Term 'Stiction'p. 145
11.2.3 A Formal Definition of Stictionp. 146
11.3 Practical Examples of Valve Stictionp. 148
11.4 Summaryp. 151
12 Physics-Based Model of Control Valve Stictionp. 153
12.1 Introductionp. 153
12.2 Physical Modelling of Valve Frictionp. 153
12.2.1 Physics of a Control Valvep. 153
12.2.2 Friction Modelp. 154
12.2.3 Model Parametersp. 155
12.2.4 Detection of Zero Velocityp. 156
12.2.5 Model of the Pressure Chamberp. 156
12.3 Valve Simulationp. 157
12.3.1 Open-Loop Responsep. 157
12.3.2 Closed-Loop Responsep. 158
12.4 Summaryp. 160
13 Data-Driven Model of Valve Stictionp. 161
13.1 One-Parameter Data-Driven Stiction Modelp. 161
13.2 Two-Parameter Data-Driven Model of Valve Stictionp. 163
13.2.1 Model Formulationp. 163
13.2.2 Dealing with Stochastic or Noisy Control Signalsp. 166
13.2.3 Open-Loop Response of the Model Under a Sinusoidal Inputp. 166
13.2.4 Stiction in Realityp. 167
13.2.5 Closed-Loop Behaviour of the Modelp. 167
13.3 Comparison of Physics-Based Model and Data-Driven Modelp. 171
13.4 Summaryp. 171
14 Describing Function Analysisp. 173
14.1 Introductionp. 173
14.2 Describing Function Analysis for Two-Parameter Stiction Modelp. 174
14.2.1 Derivation of the Describing Functionp. 174
14.3 Asymptotes of the Describing Functionp. 177
14.4 Insights Gained from the Describing Functionp. 178
14.4.1 The Impact of the Controller on the Limit Cyclep. 179
14.5 Summaryp. 180
15 Automatic Detection and Quantification of Valve Stictionp. 181
15.1 Introductionp. 181
15.2 Stiction Detection - A Literature Reviewp. 182
15.3 Detection of Stiction Using Nonlinearity Information and the pv-op Mappingp. 183
15.3.1 Detection of Loop Nonlinearityp. 184
15.3.2 Use of pv-op Plotp. 185
15.4 Stiction Quantificationp. 187
15.4.1 Clustering Techniques of Quantifying Stictionp. 187
15.4.2 Fitted Ellipse Technique for Quantifying Stictionp. 190
15.5 An Illustrative Examplep. 192
15.5.1 Validation of the Resultsp. 193
15.6 Automation of the Methodp. 193
15.7 Simulation Resultsp. 195
15.7.1 A Worked Examplep. 195
15.7.2 Distinguishing Limit Cycles Caused by Stiction and Those Caused by a Sinusoidal Disturbancep. 196
15.7.3 Detecting Stiction When Its Impact Propagates as Disturbancep. 198
15.8 Practical Implementation Issuesp. 201
15.8.1 Bicoherence Estimationp. 201
15.8.2 Nonstationarity of the Datap. 201
15.8.3 Problems of Outliers and Abrupt Changesp. 201
15.8.4 Dealing with Short Length Datap. 202
15.8.5 Dealing with Longer Oscillationsp. 202
15.8.6 Valve Nonlinearityp. 202
15.8.7 Filtering of the Datap. 203
15.8.8 Segmenting Data for pv-op Plotp. 204
15.9 Summaryp. 204
16 Industrial Applications of the Stiction Quantification Algorithmp. 205
16.1 Industrial Case Studiesp. 205
16.1.1 Loop 1: A Level Loopp. 205
16.1.2 Loop 2: A Linear-Level Control Loopp. 207
16.1.3 Loop 3: A Flow Control Loopp. 208
16.1.4 Loop 4: Flow Control Loop Cascaded with Level Controlp. 209
16.1.5 Loop 5: A Pressure Control Loopp. 210
16.1.6 Loop 6: A Composition Control Loopp. 210
16.1.7 Loop 7: A Cascaded Flow Control Loopp. 211
16.1.8 Loop 8: A Temperature Control Loopp. 212
16.1.9 Loops 9 and 10p. 212
16.2 Online Compensation for Stictionp. 213
16.3 Summaryp. 215
17 Confirming Valve Stictionp. 217
17.1 Methods to Confirm Valve Stictionp. 217
17.2 Gain Change Method for Confirming Valve Stictionp. 218
17.2.1 Distinguishing Stiction from External Oscillatory Disturbancep. 218
17.3 Describing Function Analysisp. 222
17.3.1 Comparison of Describing Function Analysis (DFA) Results with Simulation Resultsp. 225
17.4 Industrial Examplep. 225
17.5 Summaryp. 226
Part V Plant-wide Oscillations - Detection and Diagnosis
18 Detection of Plantwide Oscillationsp. 229
18.1 Introductionp. 229
18.2 What is an Oscillation?p. 230
18.2.1 Units of Frequencyp. 230
18.2.2 Examples of Oscillatory Signalsp. 230
18.3 Detection of Oscillation(s) in a Single Time Seriesp. 231
18.3.1 The Power Spectrump. 231
18.3.2 Hagglund's IAE Methodp. 231
18.3.3 Autocovariance (ACF) Based Methodp. 232
18.3.4 Other Methodsp. 237
18.4 What are Plant-wide Oscillations?p. 237
18.5 Classification of Plant-wide Oscillations or Disturbancesp. 237
18.5.1 Time scalesp. 237
18.5.2 Oscillating and Non-oscillating Disturbancesp. 238
18.6 Detection of Plant-wide Oscillationsp. 238
18.6.1 High-Density Plotsp. 238
18.6.2 ACF-Based Methodp. 239
18.6.3 Power Spectral Correlation Map (PSCMAP)p. 239
18.6.4 Spectral Envelope Methodp. 240
18.6.5 Spectral Decomposition Methodsp. 241
18.7 Summaryp. 250
19 Diagnosis of Plant-wide Oscillationsp. 253
19.1 Root Cause Diagnosis of Plant-wide Oscillationp. 253
19.1.1 Finding a Nonlinear Root Cause of a Plant-Wide Disturbancep. 253
19.1.2 Finding a Linear Root Cause of a Plant-wide Disturbancep. 256
19.2 Industrial Case Study 1 - Eastman Chemical Plantp. 257
19.2.1 Data Descriptionp. 258
19.2.2 Reduction of the Problem Sizep. 258
19.2.3 Detection of Plant-wide Oscillation by PSCMAPp. 259
19.2.4 Nonlinearity Analysis Using Biocherence-Based Indicesp. 260
19.2.5 Diagnosis of the Problem in Loop LC2p. 262
19.3 Industrial Case Study 2 - SE Asia Refinery Data Analysisp. 263
19.3.1 Oscillation Detection by PSCMAPp. 264
19.3.2 Oscillation Detection by Spectral Envelopep. 265
19.3.3 Oscillation Diagnosisp. 266
19.4 Industrial Case Study 3 - Mitshubishi Chemical Corporationp. 266
19.4.1 Scope of the Analysis and Data Setp. 268
19.4.2 Oscillation-Detection Resultsp. 268
19.4.3 Oscillation Diagnosisp. 268
19.4.4 The Results of Maintenance on the PC1 and LI1 Loopsp. 271
19.5 Summaryp. 272
Referencesp. 273
Copyright Acknowledgementsp. 281
Indexp. 283