14
Advanced Instrumentation, Data Interpretation, and Control of Biotechnological Processes
Advanced Instrumentation, Data Interpretation, and Control of Biotechnological Processes
Advanced Instrumentation, Data Interpretation, and Control of Biotechnological Processes
Edited by
Jan F.M. Van Impe Katholieke Universiteit Leuven, Department of Food and Microbial Technology, Leuven, Belgium
Peter A. V anrolleghem Universiteit Gent, Department of Applied Mathematics, Biometrics and Process Control, Gent, Belgium
and
Dirk M. Iserentant Katholieke Universiteit Leuven, Department of Food and Microbial Technology, Leuven, Belgium
Springer-Science+Business Media, B.V.
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN 978-90-481-4954-4 ISBN 978-94-015-9111-9 (eBook) DOI 10.1007/978-94-015-9111-9
Printedon acid-free paper
All Rights Reserved © 1998 Springer Science+Business Media Dordrecht Originally published by Kluwer Academic Publishers in 1998. Softcoverreprint of the hardcover Ist edition 1998
No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission from the copyright owner.
To my parents J.V.I.
Vll
Contents
I Hardware Sensors 1
1 Flow Injection Analysis for On-line Monitoring of a Waste-water Treatment Plant 3 S. Isaacs and H. S¢eberg
1.1 Introduction . . . . . . . . . . . . 4 1.2 What is Flow Injection Analysis? 5
1.2.1 Basic principles . . . . 5 1.2.2 Advantages of FIA . . . . 7 1.2.3 Some design principles . . 8
1.3 FIA for Activated Sludge Process Monitoring 10 1.3.1 Sample acquisition, separation, selection and transport 13 1.3.2 Sample injection 18 1.3.3 Reagent transport 18 1.3.4 Reaction . . . . . . 1.3.5 Detection ..... 1.3.6 Standard calibration 1.3.7 Data analysis .... 1.3.8 Automatic and manual trouble prevention . 1.3.9 The analyzers .......... . 1.3.10 The problem of carry-over ... . 1.3.11 Implementation and an example
1.4 Bibliography 1.5 Appendix · ................ .
2 On-line Measurement of Viable Biomass A.J.C. Spierings
2.1 Basic Theory of Biomass Measurement ....... . 2.1.1 Capacitance, conductance and electric fields 2.1.2 Frequency of an electric field ........ .
21 21 22 25 26 27 33 35 38 39
41
42 42 46
Vlll CONTENTS
2.1.3 Effect of dead cells, non-biomass solids and immiscible liquids, bubbles and cell type . . . . . . . . . . . 53
2.1.4 Practical methods of monitoring cellular biomass . . . 59 2.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . 64
2.2.1 Biomass monitor reading versus consistency measure-ment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.2.2 Biomass monitor reading versus OD and RQ measure-ments ... 65
2.2.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . 66
3 Membrane Inlet Mass Spectrometry for the Characteriza-tion and Monitoring of Biotechnological Processes 67 F.R. Lauritsen 3.1 Introduction . . . . . . . . . 68 3.2 Mass Spectrometry . . . . . 69
3.2.1 Mass spectrometers 70 3.2.2 Ionization of volatile compounds 73 3.2.3 Tandem mass spectrometry .
3.3 Membrane Inlet Mass Spectrometry . . 3.3.1 Theory ............. . 3.3.2 Practical use of silicone membranes for the measure
ment of volatile organic compounds . . . . . 3.3.3 Membrane inlet design . . . . . . . . . . . . . .
3.4 Applications of Membrane In:let Mass Spectrometry 3.4.1 Microbial degradation of chlorinated aliphatic
pounds ..................... . 3.4.2 Fermentation of Penicillium chrysogenum . . .
3.4.3 Identification of metabolites in microbial media 3.5 Bibliography . . . . . . . . . . . . . . . . . . . . . . .
4 Flow Cytometry P. Breeuwer
4.1 Introduction . 4.2 The Flow Cytometer 4.3 Applications . . . . .
4.3.1 Assessment of cell viability and vitality 4.3.2 Measurement of intracellular pH
4.4 Conclusion 4.5 Bibliography
com-
77
81 82
86 88 91
91 94
96 101
105
106 106 110 110 116 117 118
CONTENTS ~
5 Microcalorimetric Characterization of Bacterial Inocula 121 H. Vandenhove 5.1 Introduction . . . . . . . . . . . . . . . . 122 5.2 A View on Microcalorimetry . . . . . . 123
5.2.1 Main types of microcalorimeters 124 5.2.2 Principal arrangements of microcalorimeters . 125 5.2.3 Thermodynamic considerations . . . . . . . . 126 5.2.4 Applications of microcalorimetry . . . . . . . 129
5.3 Survival of Pseudomonas fluorescens Inocula: Influence of the Physiological Growth Stage . . . . . . . . . . . 131 5.3.1 Microcalorimetry and bacterial growth . . . . 132 5.3.2 Survival in soil . . . . . . . . . . . . . . . . . 136
5.4 Microcalorimetry and Bacterial Growth Phenomena 5.4.1 Materials and methods . 5.4.2 Results and discussion 5.4.3 Conclusions
5.5 Bibliography
II Model based Control
6 On-line Dz:tta Acquisition P.A. Willems and J.P. Ottoy
6.1 Introduction ..... 6.2 Measuring Principles
6.2.1 Introduction 6.2.2 Low level electrical measurements 6.2.3 Pulses . . . .
6.3 Signal Conditioning 6.4 Data Conversion . .
6.4.1 Types of signals 6.4.2 Sample and hold 6.4.3 A/D conversion . 6.4.4 D j A conversion .
6.5 Data Transmission . . . 6.5.1 Introduction 6.5.2 Analog data transmission 6.5.3 Digital data transmission 6.5.4 Pulse trains . . . . . . . 6.5.5 Internal data transport
6.6 Controlling Devices ...... .
142 142 143 148 149
159
161
162 163 163 164 167 168 169 169 169 170 174 176 176 178 179 183 184 185
X
6.6.1 Computers .. . 6.6.2 PLC ...... . 6.6.3 Microcontrollers
6.7 Actuator Interfacing 6.8 Galvanic Isolation 6.9 Bibliography
CONTENTS
185 186 186 187 188 190
7 Data Processing for System Identification P. Van Overschee
191
7.1 Introduction ............. . 7.2 Trend Determination and Correction 7.3 Peak Shaving . . . . . . . . 7.4 Estimation of Time Delays 7.5 Filtering . . . . . 7.6 About Linearity . 7. 7 Conclusions . 7.8 Bibliography
8 Error Diagnosis and Data Reconciliation Using Linear Con-
191 193 197 199 203 205 208 209
servation Relations 211 C. Hellinga, B. Romein, K.Ch.A.M. Luyben and J.J. Heijnen
8.1 Introduction . . . . . . . . . 212 8.2 Measurement Inaccuracies .. 212
8.2.1 Stochastic variations 212 8.2.2 Systematic deviations 213 8.2.3 Derived values .. 213
8.3 Introduction to Gross Error Detection and Data Reconciliation214 8.3.1 Gross error detection ................... 214 8.3.2 Data reconciliation . . . . . . . . . . . . . . . . . . . . 216
8.4 Gross Error Detection and Data Reconciliation Using One Conservation Relation . . . . 217 8.4.1 Gross error detection . . . . . . . . . . . . . 217 8.4.2 Data reconciliation . . . . . . . . . . . . . . 219 8.4.3 Gross error detection: a statistical criterion 221
8.5 Gross Error Detection and Data Reconciliation Using Multi-ple Conservation Relations . . . . . . . . . . 221 8.5.1 Formulation of the equations . . . . . 221 8.5.2 Classification of the conversion rates . 223 8.5.3 The general equations for gross error detection and
data reconciliation 224 8.6 Gross Error Diagnosis . . . . 228
CONTENTS XI
8.6.1 Measurement errors . . 229 8.6.2 System definition errors 231
8. 7 Increasing the Test Sensitivity . 232 8.8 Detecting Errors in the Primary Measurements 233 8.9 Discussion and Conclusions . . . . . . . . . . . 234 8.10 Appendix A: Elementary Error Propagation Rules 236 8.11 Appendix B: Effect of the Number of Samples on the Esti-
mated Accuracy of a Stochastic Variable . . . . . . . . . . . . 237 8.12 Appendix C: Macrobal- a Computer Program for Data Rec-
onciliation and Gross Error Detection 239 8.13 Nomenclature 8.14 Bibliography
9 General Concepts of Bioprocess Modeling G.C. Vansteenkiste 9.1 General Introduction .. 9.2 Simulation Scenario 9.3 Modeling Methodologies 9.4 Biotechnological Processes as ill-defined Systems 9.5 Needs in Advanced Simulation of ill-defined Systems 9.6 Perspectives of the Simulation Tool ......... .
242 243
245
245 245 247 248 249 250
10 Bioprocess Model Identification 251 P.A. Vanrolleghem and D. Dochain 10.1 Introduction . . . . . . . . . . . 252
10.1.1 Mathematical models 253 10.1.2 The model building exercise . 254 10.1.3 Current research areas w.r.t. bioprocess models 257
10.2 Case Study . . . . . . . . . 258 10.2.1 Process . . . . . . . . . . 258 10.2.2 Candidate model set . . . 259
10.3 Structure Characterization (SC) 10.3.1 A priori SC .. 10.3.2 A posteriori SC . . . . . .
10.4 Parameter Estimation ..... . 10.4.1 Theoretical identifiability 10.4.2 Practical identifiability .
10.5 Experimental Design ...... . 10.5.1 Introduction ...... . 10.5.2 Optimal experimental design for structure characteri-
261 262 264 270 271 280 288 288
zation ........................... 291
.. Xll CONTENTS
10.5.3 Optimal experimental design for parameter estimation 300 10.6 Optimal Experimental Design for the Dual Problem of Struc-
ture Characterization and Parameter Estimation 309 10.7 Conclusions . 310 10.8 Nomenclature 311 10.9 Bibliography 313
11 Optimal Control of Fed-Batch Fermentation Processes 319 J.F. Van Impe 11.1 Motivation . . . . . 320 11.2 Problem Statement . 323 11.3 Optimal Control . . 326
11.3.1 Statement of the two point boundary value problem 326 11.3.2 Extremal controls . . . . . . . . . . . . . . . . . . . 327 11.3.3 Optimal control sequence for monotonic f-L and non-
monotonic 1r • • • • • • • • . . • • • • • • • • • • • • • 328 11.4 Optimal Control Sequence with State Inequality Constraints 337
11.4.1 Substrate concentration constraint Cs(t) :S Cs,MAX 337 11.4.2 Biomass concentration constraint Cx(t) :S Cx,MAX . 338
11.5 Example . . . 340 11.6 Conclusion 342 11.7 Nomenclature 343 11.8 Bibliography 344
12 Monitoring and Adaptive Control of Bioprocesses 347 D. Dochain and M. Perrier 12.1 Introduction . . . . . . . . . . . . . . . . . . . 348 12.2 General Dynamical Model . . . . . . . . . ..
12.2.1 Example #1: PHB production process 12.2.2 General dynamical model . . . . 12.2.3 Example #2: anaerobic digestion ... 12.2.4 Example #3: yeast growth . . . . . . 12.2.5 Example #4: activated sludge process 12.2.6 Fixed bed reactors . . . . . . . . . . .
351 352 354 354 356 357 359
12.3 Dynamical Analysis of Stirred Tank Bioreactor Models . 361 12.3.1 A key state transformation 361 12.3.2 Model order reduction . . . . . . . . . . . . . . . 362
12.4 Monitoring of Bioprocesses . . . . . . . . . . . . . . . . 365 12.4.1 Asymptotic observers for single tank bioprocesses . 367 12.4.2 Application to a PHB producing process . 373
12.5 On-line Estimation of Reaction Rates . . . . . . 378
CONTENTS xm
12.5.1 Statement of the estimation problem . . . . . . . . 378 12.5.2 Observer-based estimator . . . . . . . . . . . . . . 379 12.5.3 Application to the baker's yeast fed-batch process 380
12.6 Adaptive Linearizing Control of Bioprocesses . . . 386 12.6.1 Design of the adaptive linearizing controller 386 12.6.2 Example #1: anaerobic digestion . . . 388 12.6.3 Example #2: activated sludge process 393
12.7 Conclusions . 397 12.8 Bibliography 397
13 Optimal Adaptive Control of Fed-Batch Fermentation Pro-cesses 401 J.F. Van lmpe and G. Bastin
13.1 Introduction ............... . 13.2 Optimal Adaptive Control: Motivation .
13.2.1 Problem statement ....... .
402 404 404
13.2.2 Case study: the penicillin G fed-batch fermentation 407 13.2.3 Optimal control strategy . . 409 13.2.4 Heuristic control strategies 410 13.2.5 Linearizing control . . . 413 13.2.6 The stability problem . . . 415 13.2. 7 The monitoring problem . . 418
13.3 Optimal Adaptive Control: On-line Measurements of Cs and Cx . . . . . . . . . . . . . . . . . 419 13.3.1 Mathematical description ........... , . . 419 13.3.2 Simulation results . . . . . . . . . . . . . . . . . . 421
13.4 Optimal Adaptive Control: On-line Measurements of Cs . 423 13.4.1 Mathematical description . . . . . . . . . . . . . . 423 13.4.2 Simulation results . . . . . . . . . . . . . . . . . . 424
13.5 Optimal Adaptive Control: On-line Measurements of CER . 426 13.5.1 Mathematical description 426 13.5.2 Simulation results 430
13.6 Conclusions . 431 13.7 Nomenclature 433 13.8 Bibliography 434
14 Predictive Control in Biotechnology using Fuzzy and Neural Models 437 H. te Braake, R. Babuska, E. van Can and C. Hellinga
14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 438 14.2 Non-linear Model-based Predictive Control Structure . 441
XIV
14.3 Modeling with Neural Networks .. 14.3.1 Neural network structure .
CONTENTS
443 443
14.3.2 Training of neural networks 445 14.3.3 RAWN training method . . 446 14.3.4 Modeling non-linear dynamic systems with neural net-
works . . . . . . . . . 448 14.4 Modeling with Fuzzy Logic . . . . 449
14.4.1 Fuzzy model structure . . . 450 14.4.2 Fuzzy inference mechanism 452 14.4.3 Identification of fuzzy models from data 453 14.4.4 Prediction with fuzzy models . . . . . . 453
14.5 Application to Pressure Control in a Fermentor 454 14.5.1 Experimental setup . . . . . . . . . . . . 454 14.5.2 Building a neural network model for the pressure process455 14.5.3 Fuzzy modeling of the pressure process . . . . . . 457 14.5.4 Controlling the fermentor with non-linear MBPC 459
14.6 Discussion and Conclusions 461 14.7 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . 462
XV
Preface
This book is the outgrowth of the COMETT II Course on Advanced Instrumentation, Data Interpretation, and Control of Biotechnological Processes organized by the Katholieke Universiteit Leuven and the Universiteit Gent, and held at Gent, Belgium, October 1994.
The editors of the present volume were very fortunate to find all invited speakers prepared to write state-of-the-art expositions based on their lectures. Special thanks are due to all of them. The result is an account of recent advances in instrumentation, data interpretation, and model based optimization and control of bioprocesses. For anyone interested in this emerging field, this text is of value and provides comprehensive reviews as well as new and important trends and directions for the future, motivated and illustrated by a wealth of applications.
The typesetting of all this material represented a tremendous amount of work. I am most grateful to my wife, Myriam Uyttendaele, and to Kurt Gheys, who did most of the proof-reading. Their efforts have increased a lot the uniformity in style and presentation of the different manuscripts. Many thanks also to the co-editors, for their continued support.
Kluwer Academic Publishers is gratefully acknowledged for publishing this book, thus contributing to the transfer of the latest research results into large scale industrial applications.
Leuven, august 1997
Jan F.M. Van Impe
xvn
Introduction
The scope ofthe field of biotechnological processes is very wide, covering such processes as fermentations for production of high-valued specialist chemicals (e.g., pharmaceuticals), high-volume production of foods and feeds (yoghurt, cheese, beer, ... ) as well as biological waste treatment, handling solid (composting), liquid (activated sludge) and gaseous wastes (biofilters). Compared to other engineering disciplines, the introduction of modern optimization and control strategies is lagging behind. Two main reasons can be identified. First, the living organisms (or part thereof) that are central to these processes make the mathematical modeling of the processes a difficult task, and, since models are central to the development of control systems, therefore also the on-line control problem is complex. The other difficulty stems from the absence, in most cases, of cheap and reliable instrumentation suited to real-time monitoring.
In this book a number of advanced techniques are introduced to deal with these problems. In the first part modern on-line hardware sensors are discussed in detail (FIA, viable biomass measurement, membrane inlet mass spectrometry, flow cytometry, micro calorimetry). In the second part several aspects of model based optimization and control are dealt with, starting from on-line data acquisition, processing, error diagnosis and data reconciliation, over bioprocess modeling ~nd identification, and development of combined hardware-software sensors, up to on-line (optimal) adaptive and model based predictive control algorithms.
The book is directed to engineers, researchers, and students in the field of process control and systems theory as applied to industrial biotechnological processes, as well as to bioengineers who have some background in control engineering and would like to apprehend better how advanced control theory applies to biological processes.
The Editors
