Barry A. Perlmutter - Integration and Optimization of Unit Operations_ Review of Unit Operations from R&D to Production_ Impacts of Upstream and Downstream Process Decisions-Elsevier (2022).pdf

Integration and Optimization of Unit Operations

This page intentionally left blank

Integration and
Optimization of Unit
Operations
Review of Unit Operations from R&D
to Production: Impacts of Upstream and
Downstream Process Decisions
Edited by
Barry A. Perlmutter
President, Perlmutter & Idea Development LLC, Matthews, NC, United States

Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2022 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any information storage and retrieval system, without
permission in writing from the publisher. Details on how to seek permission, further information about the
Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance
Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher (other
than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our
understanding, changes in research methods, professional practices, or medical treatment may become
necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using
any information, methods, compounds, or experiments described herein. In using such information or methods
they should be mindful of their own safety and the safety of others, including parties for whom they have a
professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability
for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or
from any use or operation of any methods, products, instructions, or ideas contained in the material herein.
ISBN: 978-0-12-823502-7
For information on all Elsevier publications
visit our website at https://www.elsevier.com/books-and-journals
Publisher: Susan Dennis
Acquisitions Editor: Anita Koch
Editorial Project Manager: Lindsay Lawrence
Production Project Manager: Kumar Anbazhagan
Cover Designer: Christian J. Bilbow
Typeset by STRAIVE, India

Contents
Contributors xiii
About the editor xv
Preface xvii
1. Crystallization
Brooke Albin
1.1 Fundamentals and laboratory scale
process development 1
1.1.1 Crystallizer design basics 1
1.1.2 Crystallizer design tradeoffs 4
1.1.3 Upstream variables affecting
crystallization 6
1.1.4 Impact on downstream
operations 7
1.2 Pilot scale crystallization studies 8
1.2.1 Objectives for a pilot plant 8
1.2.2 Scale-up criteria 9
1.3 Commercialization of crystallization
processes 10
References 11
2. Fermentation and downstream
processing: Part 1
Alan Gabelman, Ph.D., P.E.
2.1 Introduction 13
2.2 Microbiology and biochemistry
basics 13
2.3 Fermentation media and
environment 21
2.4 Growth kinetics and substrate
utilization 24
2.5 From vial to production fermenter 27
2.6 Oxygen transfer and utilization 29
2.7 Mixing in aerobic fermentation
vessels 37
2.8 Sterilization 43
2.8.1 Batch sterilization 46
2.8.2 Continuous sterilization 47
2.8.3 Filter sterilization of liquids 53
2.8.4 Filter sterilization of air 56
2.9 Heat generation 56
2.10 Scale-up 59
Nomenclature 64
References 65
3. Fermentation and downstream
processing: Part 2
3.1 Fermenter design 69
3.1.1 Fermenters without mechanical
mixers 73
3.2 Fermenter instrumentation, control and
operation 75
3.2.1 Temperature 77
3.2.2 pH 78
3.2.3 Dissolved oxygen
concentration 79
3.2.4 Mixer speed 80
3.2.5 Pressure 80
3.2.6 Gas flow rate 81
3.2.7 Liquid flow rate 82
3.2.8 Foam 83
3.2.9 Exit gas composition 85
3.2.10 Level 86
3.2.11 Substrate concentration 87
3.2.12 Power input 88
3.2.13 Redox potential 89
3.3 Continuous culture 89
3.4 Downstream processing 93
3.4.1 Monosodium glutamate 94
3.4.2 Phenethyl alcohol 100
3.5 Concluding remarks 108
Nomenclature 108
References 109
v

4. Liquid filtration
Jose M. Sentmanat
4.1 Do you need a filter? 113
4.2 Lab testing before you choose the
filter 113
4.3 Choosing the filter 116
4.3.1 Plate and frame filter press 116
4.3.2 Filter presses 116
4.3.3 Plate filters 117
4.3.4 Pressure leaf type filter 117
4.3.5 Nutsche filter 118
4.3.6 Polishing filter 118
4.4 The ABCs of liquid filtration 118
4.5 The mechanics of liquid filtration 119
4.5.1 Precoat 119
4.5.2 Filtration 120
4.5.3 Cleaning 120
4.5.4 Standby 121
4.6 Troubleshooting 121
4.7 The filter cake 121
4.8 Preventative maintenance program 122
Further reading 123
5. Cake-building filter technologies
Jose M. Sentmanat and Barry A. Perlmutter
5.1 Batch processing of filter cakes 125
5.2 Contained filter presses for cake
washing, dewatering, and drying 126
5.3 Nutsche filter and filter dryers 127
5.4 Continuousprocessingoffiltercakes 128
5.4.1 Vacuum belt filters 128
5.4.2 Horizontal vacuum belt filters 129
5.4.3 Rotary vacuum drum filters 131
5.4.4 Rotary pressure filter 131
5.4.5 Pressurized vacuum drum
filter 131
6. Centrifugation
Badrie Luckiram, BSc, MSc, CEng, MIChemE
6.1 Centrifuge choice and analysis of
available equipment 133
6.1.1 Horizontal basket centrifuges 135
6.1.2 Vertical basket centrifuges 135
6.2 Typical centrifuge operation 138
6.3 Technical considerations of equipment
selection 138
6.3.1 Design basis document 138
6.4 Other considerations of centrifuge
operation 141
6.4.1 Centrifuge inerting 141
6.4.2 General operation 141
6.4.3 Safety interlocks 142
6.4.4 Out of balance monitor 142
6.4.5 Plough parked 142
6.5 Final remarks 142
7. Dryers
Hongben Zhou
7.1 Purpose of drying 145
7.2 Dispersed solid-liquid system 145
7.3 Drying processes 147
7.4 Convective drying with hot gas 147
7.5 Conductive and radiative drying 150
7.6 Evaporation of liquid from a solid
packing 151
7.7 Drying facilities 153
7.7.1 Grain-sunning ground 153
7.7.2 Tray dryer 154
7.7.3 Belt dryer 156
7.7.4 Rotary dryer (kiln) 156
7.7.5 Fixed bed dryer 159
7.7.6 Fluidized bed dryer 161
7.7.7 Pneumatic conveyor as dryer 162
7.7.8 Spray dryer 165
7.7.9 Impact mill as dryer 166
7.7.10 Rotating vessel dryer 168
7.7.11 Plate dryer 168
7.7.12 Roller dryer 170
7.7.13 Screw conveyor as dryer 170
7.7.14 Agitated mixer as dryer 171
7.8 Troubleshooting 174
7.8.1 Heat transfer 174
7.8.2 Level of vacuum 175
7.8.3 Formation of agglomerates and
crust 175
References 176
8. Pressure filter dryer
8.1 General considerations of using a
pressure filter dryer 177
8.1.1 Pharma-specific
considerations 178
8.2 Principles of the pressure filter
dryer 179
8.3 Filter choice and analysis of available
equipment 182
8.3.1 Selection of filter dryer type 182
8.4 Technical considerations of equipment
selection 183
vi Contents

8.5 General operation of a pressure filter
dryer 183
8.5.1 GMP issues and cleaning 189
8.5.2 Filter safety interlocks 189
8.5.3 Operational issues 190
9. Process automation systems
Nick Harbud
9.1 Process automation in production
facilities 191
9.2 Process control system
(continuous process) 191
9.2.1 Controlling the process 191
9.2.2 Operating the plant 193
9.2.3 Integrating automation
systems 194
9.2.4 Enterprise interfaces 195
9.2.5 Types of process control
system 195
9.3 Process control systems
(batch process) 197
9.4 Safety instrumented systems 201
9.4.1 Identifying the hazards 203
9.4.2 Assessing the risks 203
9.4.3 High integrity pressure protection
systems 205
9.4.4 Cybersecurity risk assessment 206
9.4.5 Validation and proving 206
9.5 Alarm management systems 207
9.6 Machinery protection 209
9.6.1 Vibration monitoring system 209
9.6.2 Compressor and turbine control
systems 209
9.7 Measurement, and other fun things to
do with instruments 212
9.7.1 Diagnostics—Is it working? 213
9.7.2 Control in the field 214
9.7.3 The growth of digital
communications protocols 214
9.7.4 HART 214
9.7.5 Fieldbus 215
9.7.6 Ditching the wires 216
9.7.7 Instrument asset management
systems (IAMS) 217
9.8 The effect of technology on process
automation 217
10. Process automation life cycles
Nick Harbud
10.1 Planning for process automation 219
10.1.1 Operations and maintenance
philosophy 219
10.1.2 Identify key automation systems
and technology 219
10.1.3 Identify advanced control
schemes 220
10.1.4 Estimate system size 221
10.1.5 Site planning overall
philosophy 221
10.2 Front end engineering design 226
10.2.1 Basic automation
requirements 226
10.2.2 Advanced process control 226
10.2.3 The MAC, and why you should
use one 226
10.2.4 Other automation systems 227
10.2.5 Functional safety 228
10.2.6 Change management for
process automation 228
10.3 Delivery phase, detailed engineering,
and procurement 229
10.3.1 Process automation design
documentation 229
10.3.2 Automation system design
and software configuration 230
10.3.3 Factory acceptance testing 230
10.3.4 Shipment and site
preservation 231
10.4 Installation and commissioning 231
10.4.1 Manpower plan 231
10.4.2 Infrastructure and overheads
plan 232
10.4.3 PAS media plan 233
10.4.4 PAS change management
plan 233
10.4.5 PAS security plan 233
10.4.6 PAS integration plan 233
10.4.7 PAS maintenance plan 234
10.4.8 PAS user administration
plan 234
10.4.9 PAS turnover plan 235
10.5 Automation system operation and
obsolescence 235
10.5.1 Hardware maintenance
and obsolescence 235
10.5.2 Software maintenance and
change 235
10.5.3 Disaster recovery 236
10.6 Conclusion 237
11. Process automation platforms
Mike Williams
11.1 Background 239
11.2 Staffing of a manufacturing facility 239
11.3 Finding the balance 240
Contents vii

11.4 The new paradigm of autonomous
operations 240
11.5 Upgrading the level of automation 245
11.6 Where to start when considering
investment in higher levels of
autonomy 246
11.7 Conclusions 247
12. Mixing and blending
Stephanie Shira
12.1 Introduction: Why mixing
matters 249
12.2 Upstream considerations 249
12.2.1 Before the shafts 250
12.2.2 The first shaft 250
12.2.3 Distributive vs dispersive
mixing 253
12.3 The second shaft 254
12.3.1 High speed dispersion and low
speed scraping: The traditional
dual-shaft mixer 254
12.3.2 More intense dispersion
(double the shafts, quadruple
the blades of a traditional
disperser): The dual-shaft
disperser 255
12.3.3 Dual-shaft disperser case study
and performance review 258
12.4 The third shaft 258
12.5 Additional mixer design
considerations 258
12.6 Rheology considerations 260
12.7 Overmixing is just as bad as
undermixing: Know the finishing
point 261
12.7.1 Kitchen connection 261
12.7.2 Case study: “Pancake lumps”
on the production floor 261
12.7.3 Compensating behaviors result
from inadequate products 262
12.8 Reliable scale-up 262
12.8.1 Hydraulic ram discharge
press 263
12.9 Mechanical aspects and
troubleshooting 264
12.9.1 Blade health 264
12.9.2 Understanding shear (rates and
flow regimes) 265
12.10 Case study: Why push toward
efficiency? 266
12.10.1 The old way: Paradigm 266
12.10.2 The new way: Break the
paradigm 269
12.10.3 What was saved? 270
12.10.4 In conclusion: Every
perspective matters 271
References 271
Further reading 271
13. Process development and
integration by mathematical
modeling and simulation tools
Nima Yazdanpanah
13.1 Fundamentals and workflow 273
13.2 The steps for building a mathematical
model 275
13.3 Steady-state and dynamic
simulations 277
13.4 Process simulation for
optimization 277
13.4.1 Construction of the optimization
problem and its
components 279
13.5 Process development workflow for
continuous manufacturing 280
13.5.1 Process integration and steady-
state simulation 281
13.5.2 Dynamic process modeling
and control 283
13.6 Correlation between CQAs, CPPs,
CMAs 286
References 292
14. Process safety
Kaushik Basak
14.1 Lab-scale operations 293
14.1.1 Safety and hazards 293
14.1.2 Key issues for lab-scale
operation 294
14.2 Pilot plant operations 297
14.2.2 Key issues for pilot plant
operations 299
14.2.3 Pilot plant sizing, issues,
decisions, and trade-offs 301
14.3 Production scale operations 303
14.3.2 Key issues for production scale
operation 304
References 305
viii Contents

15. Process commissioning
15.1 Commissioning 307
15.2 Competency 307
15.3 Checks prior to the start of
commissioning 308
15.4 Commissioning protocols 308
15.5 Specific process engineering
responsibilities 309
15.6 Handover of the plant to the user 309
15.7 Overall recommendations for process
engineers 310
Appendix: Example Commissioning
Protocol for a new Hydrochloric Acid
Tanker Offloading Pump 310
16. Holistic process integration and
optimization: Large-scale hybrid
process applications
Ugur Tuzun
16.1 Introduction 317
16.2 Life cycles of generic activities for
large-scale bulk chemicals
production 318
16.3 Systems integration design for
specialty products manufacture and
sales 321
16.4 Gated process development with
digital interlinks 321
16.5 Digital control life cycles of integrated
large-scale production plants 327
16.5.1 Configuring
communications 327
16.5.2 Multivariable devices
communication 328
16.5.3 Loop converters 328
16.5.4 Multiplexers 328
16.6 Environmental impact monitoring
and control 329
16.6.1 Green process applications in
process industries 330
16.6.2 Industrial emissions control
strategies using digital
platforms 330
16.6.3 Digital environmental sensor
technologies 330
16.6.4 Digital platform construction
for multivariate process and
environmental datasets 331
16.6.5 Coupling environmental and
process chemistry 333
16.6.6 Environmental emissions
records and HAZOP studies 333
16.7 Systems integration of plant
operations within eco-industrial
parks 334
Acknowledgments 337
References 337
17. From idea to 1 million ton year
commercial plant
Joep Font Freide
17.1 The framework 339
17.2 The execution 341
17.2.1 Concept and laboratory
stage 341
17.2.2 Micro reactor stage 341
17.2.3 Pilot plant stage 342
17.2.4 Demonstration plant stage 343
17.3 At last: Safety first 344
18. Scale-up challenges: Examples
from refining and catalysis
Kaushik Basak
18.1 Challenges in refining scale-up 345
18.2 Challenges in catalyst scale-up 348
18.3 Decision gate for catalyst scale-up 349
References 350
19. Scale-up challenges: Wastewater
Kaushik Basak
19.1 Challenges in wastewater
treatment 351
References 353
20. Hemp/biomass process steps
Jay Van der Vlugt
20.1 Hemp cultivation overview 355
20.2 Extraction 356
20.2.1 Ethanol 356
20.2.2 Gaseous hydrocarbon
extraction 357
20.2.3 Liquid hydrocarbon
extraction 358
Contents ix

20.2.4 Subcritical and supercritical
carbon dioxide 359
20.2.5 Cosolvent injection 360
20.2.6 Solvent-less processes 360
20.2.7 Dry sifting 360
20.2.8 Cold water (kief) extraction 361
20.2.9 Distillation 362
20.3 Innovations and other extraction
technologies 364
20.3.1 Ultrasonic processing 364
20.3.2 Hybrid microwave 365
20.3.3 Targeted cannabinoid salt
precipitation 365
20.3.4 Winterization-purification 367
20.3.5 Organic solvent
nanofiltration 367
20.4 Cannabinoid isolation 368
20.4.1 Decarboxylation 370
20.5.1 Hazardous installation
requirements 370
20.5.2 Contamination and other
process issues 371
References 372
21. Techno-economic analyses
Ron Leng and John Anderson
21.1.1 Uses of a techno economic
assessment 373
21.1.2 Decision making 374
21.2 Technology assessment 376
21.2.1 Definition of new
technology 376
21.2.2 Feasibility: The first screen 377
21.2.3 Technology scalability to
full-scale manufacturing 377
21.2.4 Technical success
parameters 377
21.2.5 Types of technology risk 378
21.2.6 Risk management plan 379
21.2.7 Licensed technology 382
21.2.8 Investment in a start-up
technology 383
21.2.9 Duplication of existing
technology: A caution 383
21.2.10 Types of projects 383
21.2.11 Types of process
technology 384
21.2.12 Batch vs. continuous
mode 385
21.2.13 Technology package 386
21.3 Making cost-of-manufacturing
estimates during the early stages of a
project 387
21.3.1 Identifying variable and fixed
costs 387
21.3.2 Variable costs 388
21.3.3 Fixed costs 392
21.4 Putting the costs together: Example
problems 397
21.5 Handling uncertainties during early
project stages 399
21.6 Combining costs with revenues to
compute economic indicators 405
21.6.1 Introduction to economic
indicators 405
21.6.2 There are only two key
questions 405
21.6.3 Risk and reward: Is there any
data? 405
21.6.4 Financial indicators:
Definitions 405
21.6.5 Internal rate of return (IRR) or
discounted cash flow percent
(DCF%) 406
21.6.6 Final summary 409
References 411
22. Project management
Venkata Ramanujam and Bob Barnes
22.2 The project engineering process 413
22.2.1 Integrating course work in
chemical process
engineering 415
22.3 Predictive tools 418
22.4 Industries served by process
engineers 419
22.5 Process plant components 419
22.6 Process safety and process engineering
work flow 420
22.7 Putting it all together with practical
knowledge 421
22.7.1 Selecting the site or living with
the selection handed to you 421
22.7.2 Site issues 423
22.7.3 Common concerns: Funding,
control of site 424
22.7.4 Community issues: Tax
incentives, sales tax, resources,
and workforce supply 425
22.8 Engineering: In-house resources and
EPC firms 425
x Contents

22.8.1 Forming the team 425
22.8.2 Selecting the engineering,
procurement, and construction
(EPC) firm 425
22.8.3 The all-important P&ID
development 426
22.8.4 Controls and control room
concerns 426
22.8.5 QA/QC needs 426
22.8.6 Facilities and equipment for
operations and
maintenance 427
22.8.7 Hazard analysis: Is it required or
just a good practice 427
22.8.8 Project management 427
22.8.9 Scheduling 427
22.9 Project execution 428
22.9.1 Organization and
planning 428
22.9.2 Sitework and utility supply 428
22.9.3 Foundations and steel
erection 428
22.9.4 Setting equipment 429
22.9.5 Piping 429
22.9.6 Power distribution 429
22.9.7 Control networking and field
instruments 430
22.9.8 Project controls: Schedule
and budget 430
22.9.9 Operator training 430
22.9.10 Commissioning, qualification
batches and testing and
start-up 431
23. Decommissioning
Barry A. Perlmutter
23.1 Options for decommissioning 433
23.2 How to begin decommissioning 433
23.2.1 Decontamination 433
23.2.2 Final steps of the
decommissioning project 436
Index 437
Contents xi

Contributors
Numbers in parentheses indicate the pages on which the authors’
contributions begin.
Brooke Albin (1), Research & Development, MATRIC
(Mid-Atlantic Technology, Research & Innovation
Center), South Charleston, WV, United States
John Anderson (373), Engineering & Process Sciences,
Dow Chemical, Midland, MI, United States
Bob Barnes (413), Project & Process Consultant, Prova-
tions LLC, Gregory, TX, United States
Kaushik Basak (293, 345, 351), Principal Engineer
(SMPO), Shell plc., Shell Technology Centre,
Bangalore, India
Joep Font Freide (339), FFTechnology, Guildford, United
Kingdom
Alan Gabelman, Ph.D., P.E. (13, 69), Gabelman Process
Solutions, LLC, West Chester, OH, United States
Nick Harbud (191, 219), C.Eng., F.I.Chem.E., Newbury,
United Kingdom
Ron Leng (373), Engineering & Process Science, Dow
Chemical, Midland, MI, United States
Badrie Luckiram, BSc, MSc, CEng, MIChemE (133,
177, 307), Pharmaceutical & Process Engineer, London,
United Kingdom
Barry A. Perlmutter (125, 433), Perlmutter & Idea Devel-
opment LLC, Matthews, NC, United States
Venkata Ramanujam (413), McDermott Inc., Houston,
TX, United States
Jose M. Sentmanat (113, 125), Liquid Filtration Specialist,
LLC, Conroe, TX, United States
Stephanie Shira (249), Myers Mixers, Cudahy, CA, United
States
Ugur Tuzun (317), Churchill College, University of
Cambridge, Cambridge, United Kingdom
Jay Van der Vlugt (355), Cannabinoid Sciences, Nectar
Health Sciences Inc., Victoria, BC, Canada
Mike Williams (239), Process Automation, ARC
Advisory, Dedham, MA, United States
Nima Yazdanpanah (273), Engineering and Development,
Procegence, Chevy Chase, MD, United States
Hongben Zhou (145), BHS-Sonthofen Process Tech-
nology GmbH & Co. KG, Herrsching, Germany
xiii

About the editor
Barry A. Perlmutter is President of Perlmutter & Idea Development LLC (P&ID). He has 40years of technical engi-
neering and business marketing experience in solid-liquid separation, filtration, centrifugation, and process drying. His
skills focus on process solutions, innovation strategy, and business development and market expansion. Barry has published
and presented worldwide and is responsible for introducing many European technologies into the marketplace. He is an
author of Elsevier’s Solid-Liquid Filtration - Practical Guides in Chemical Engineering handbook and a new e-book
Framework for Selecting Automated Solid-Liquid Filtration Technologies for Clarification Applications.
Barry began his career with the US Environmental Protection Agency and then entered the world of solid-liquid sep-
aration at Pall Corporation. For 11years, he continued at Rosenmund Inc. as VP of Engineering and Sales including Comber
and Guedu Dryers and Ferrum Centrifuges. From the process industries, Barry joined Process Efficiency Products, now part
of Amiad USA, as a Director of Marketing and Sales for the manufacturing of filtration, separation and adsorption
technologies for cooling tower and HVAC water, process fluids, and water and wastewater treatment. He then became
President & Managing Director of BHS-Filtration Inc. (BHS-Sonthofen Inc.) where he grew the filtration, drying, mixing,
and recycling business of BHS for more than 20years including the integration of AVA GmbH dryers. His current
company, P&ID, allows Barry to provide consulting services for process and project development with operating
companies and business development, marketing, and sales strategies for process technology suppliers.
He received his BS degree in Chemistry from Albany State (NY) University, MS degree from the School of Engineering
at Washington University, St. Louis, and an MBA from the University of Illinois, Chicago.
xv

Preface
Over my career of 40years in the process industry, writing has always been a passion for me. It represents an opportunity to
convey concepts, ideas, and technical information in a manner that makes sense to the audience. While I never had any
formalized journalism or writing training, this skill somehow developed on its own through my continuing learning,
reading, and speaking/presenting on the topics of solid-liquid separation, centrifugation, drying, and other process
equipment and technologies as well as business development and innovation. This work has spanned over 40 countries
on 6 continents.
I began writing when I was a young Environmental Scientist with the US Environmental Protection Agency (USEPA).
During those years, I issued Code of Federal Register rules and justifications, approved, of course, by the Branch Manager
and eventually the Regional Administrator. Several of my reports are still available should you be eager to read “EPA 905/
5-81-002: Economic Impact of Implementing VOC Group II Rules in Ohio” or “EPA 905/9-82-005: Air Quality Non-
Attainment Areas in Region 5.”
From the USEPA, I joined Pall Corporation and continued my writing in their marketing group where I issued my first
filtration paper in 1982, WER 5300—Principles of Filtration. This paper had to be approved by Dr. Pall before it was issued.
My writing continued, and one of my tag lines was “like the Sheriff in the Wild West, my role is to bring order from chaos in
the filtration industry.”
My technical and marketing application articles—more than 150 to date—culminated in 2015 with the publication of
my first book for Elsevier, the handbook of Solid-Liquid Filtration. Part of Elsevier’s Practical Guides in Chemical Engi-
neering, where each book provides a focused introductory view on a single subject, the Handbook required almost 1 year to
write. The fun and challenge of that task have further been rewarded with year-to-date sales of more than 900 copies.
Now here we are at the current book, Integration and Optimization of Unit Process Operations. On the strength of the
handbook’s market acceptance, Elsevier asked me to propose a second book. They suggested that based upon my
experience, I edit a book unique to the chemical process industry (CPI). I welcomed the opportunity.
The problem in the marketplace, as I see it, is the type of engineers trained. In the early 1970s, companies wanted staff
with an “I-shaped” skill level. Someone with “I-shaped” skills has a deep (vertical) expertise in one area and practically no
experience or knowledge in other areas. This person is typically known as a specialist. In the 1980s, the industry wanted
“T-shaped” professionals. The vertical bar on the T represents strong knowledge in a specific discipline. The horizontal bar
represents a wide (horizontal) yet shallow knowledge in other areas. This allows the person to be able to collaborate across
other disciplines and acquire new skills or knowledge. Now, however, with the rapid proliferation of technological
advances and the cross-disciplinary nature of work, we need “Key-shaped” engineers who have several areas of expertise
with varying degrees of depth. This book addresses this need.
First, what this book is not is another textbook for designing equipment and technology. There are many references,
university courses, etc., for this work and teaching the “nuts and bolts” of pumps, heat exchangers, distillation towers,
thermodynamics, etc.
This book takes a different approach to share up-to-date and practical information on chemical unit operations from the
R&D stage to scale-up and demonstration to commercialization and optimization. At each stage, the information presented
differs as the technology and issues faced at the lab scale change in commercialization and optimization. This book takes a
broader view and encourages a “Key-shaped” approach to chemical engineering.
As the chemical industry changes and becomes more integrated worldwide, information exchange is needed. This
exchange must include not only principles of operation, but also practical knowledge transfer. This book addresses
this need.
Engineers must be able to ask questions of I-shaped and T-shaped professionals to develop creative solutions. This book
addresses the needs of engineers who want to increase their skill levels in various disciplines so that they can develop,
commercialize, and optimize processes.
xvii

Some theory is included to provide the necessary background of the specific unit operation, but as stated previously, this
is not the main emphasis. Each chapter discusses practical aspects and illustrates the impacts of upstream process decisions
on downstream operations. Chapters also include troubleshooting at each process stage and suggest questions to ask to
develop creative solutions to process problems.
The engineer using this book will be able to take the content and apply it to the task at hand. For example, if you are
working on a process and need information on electrical and controls, you will find this. If you are a new project manager,
you will find a chapter on how to develop a project from beginning to final acceptance and start-up. Whether you are a start-
up or producing millions of tons/year, you will find the necessary guidance. I hope that this will be your “go-to book” along
the way as you grow and expand your skills and career.
The organization of the chapters follows that of a chemical operating company no matter the size of the operation. It
begins with crystallization and fermentation. Then, there are discussions of the process equipment followed by automation,
mixing and blending, process modeling and safety, and commissioning. We then discuss optimization, project man-
agement, techno-economic analysis, and “putting it all together.” The book concludes with a chapter on decommissioning
which is important, as processes change, products change, and the market itself changes.
Two more topics in the book deserve a separate mention. First, there is a chapter on hemp, cannabis, CBD or canna-
bidiols, and biomass. This is a new and flourishing industry, and many of the readers of this book may be drawn into this
process area. Finally, we discuss sustainability and holistic integration and optimization of chemical processes and con-
sumer product manufacturing. This chapter explores the impacts of environmental, socio-ecological, and economic issues
on decision making requiring the application of holistic systems modeling in process and product design to evaluate the
related consequences.
Finally, the text, as you will see, varies from chapter to chapter, as all contributing authors come from differing back-
grounds and experiences. This, I believe, it one of the greatest strengths of this book. Besides the United States, we have
authors from India, Germany, United Kingdom, Ireland, Netherlands, and Canada. Their experience encompasses process
engineers, technology suppliers, plant managers, academia, governmental agencies, consultants, and start-up to Fortune
500 companies. Each author brings a unique approach to problem solving and plant operations. An approach and expertise
they have so graciously taken the time to share.
As one author commented: “We, as a community, really have a responsibility to help and support younger engineers
and/or people who are thinking of going into the profession. We particularly need to mentor people from non-traditional
backgrounds who just need some encouragement and support, otherwise there is the danger of them becoming discouraged
and falling away from the profession. We need diversity in this profession.”
This book embraces that diversity. Thank you to all the authors who spent time researching and writing to contribute
your chapters. You are the backbone of this book. I have enjoyed working with you and truly hope that our paths will
cross again.
Thanking everyone I’ve worked with over my 40years for their guidance, influence, help, and assistance would take a
book itself. As I reflect on my career and the many worldwide friends that I have had the pleasure of meeting over all these
years, I am truly grateful to each one of you. Let me say that the word “friends” in my mind are colleagues, customers,
competitors, suppliers, publishers, editors, and many others who have helped me to succeed. I have been fortunate through
hard work, long hours, and a personal goal of making each and every one of our contacts an informative and productive
experience to build many long-lasting relationships and, more importantly, invaluable friendships over all these years. It
has been these relationships that keep me striving to give back to our engineering community.
A heartfelt thank you also to my parents, my wife Michelle, and my family, friends, trainers, and yogis for supporting me
all these years and being part of my life. You all have heard the stories, and while you may not have fully understood all, you
have been there for me forever. Thank you, thank you, and thank you again.
I now must give one final thanks to Jenn Goddu who started with me in 2014 as my technical associate, editor, friend,
and all-around writer as I publish, blog, post, and tweet. Her skills are completely beyond reproach.
And, to the readers of this book, I hope that the information from the experiences of the contributing authors will help
you to succeed in your careers and personal growth. Thank you.
Barry A. Perlmutter, Editor
Perlmutter & Idea Development LLC, Matthews, NC, United States
xviii Preface

Chapter 1
Crystallization
Brooke Albin
Research & Development, MATRIC (Mid-Atlantic Technology, Research & Innovation Center), South Charleston, WV, United States
Crystallizer process design requires attention to many varied factors. This chapter discusses fundamentals and laboratory
scale process development, pilot scale crystallization studies, and commercialization of crystallization processes to provide
an overview of the considerations in this area of solids processing.
1.1 Fundamentals and laboratory scale process development
The design of an industrial crystallization unit depends greatly on the characteristics of the feed supplied from the upstream
process and has direct consequences for downstream operations. For this reason, both narrow and broad perspectives are
needed to ensure a design that will be implemented successfully. The crystallizer operation must be robust within the entire
range of operating conditions it is subjected to. For example, if upstream concentration varies, the crystallizer must be able
to respond to that in some way in order to continue to operate smoothly and not cause upsets downstream.
The design of a crystallizer starts in the laboratory. The lab setting offers maximum flexibility for making changes to
design, adjusting conditions, and closely observing behavior of the system. Laboratory crystallizer equipment is often con-
structed of glass, which provides a significant advantage in early stages of process development when much can be learned
by visual inspection. Nucleation, crystal growth, agitation, slurry thickness, and tendency for fouling can all be studied
in situ. Ranges for operating conditions can start to be approximated often within the first few tests, and many items of
concern can be identified at this stage so proper design considerations can be made.
1.1.1 Crystallizer design basics
Crystallization is achieved by exploiting differences in solubility of components in a solution. It can be a useful method for
separating components or purifying a particular material. It is often used for recovering a solid product of high purity, but in
some cases, the objective is to remove solid impurities from a liquid stream. In either case, the separation occurs when
supersaturation is generated to solidify one component in pure form. This process is governed by the physical properties
of the components in the solution. A phase diagram is the ideal starting point for developing any crystallization process.
A phase diagram for a typical binary eutectic system is shown in Fig. 1.1. It shows the solid-liquid equilibrium data
(solubility curves) for each of the major components in solution. This provides important information regarding the con-
ditions required for crystallizing the desired component, and it establishes the limits of what crystallization can achieve in
terms of yield (recovery). For a solution of a given concentration, the diagram indicates the temperature at which crystal-
lization will begin. If temperature is lowered further, more solids will form, leaving a less concentrated liquid (mother
liquor) behind. The theoretical yield can be determined by performing a material balance that accounts for the starting
concentration of the solution and ending concentration of the mother liquor at a given set of conditions. The eutectic indi-
cates the point at which both components will crystallize and separation cannot be achieved. This represents the conditions
at which the maximum theoretical yield can be attained.
Consider the example in Fig. 1.1, which assumes a feed solution containing 80wt% of component A. Pure crystals of
component A will start to form as the solution is cooled to 0°C. Crystals of component A will continue to form as tem-
perature is lowered until the solution is cooled to its eutectic point at 40°C. At this point, both A and B will crystallize,
producing a solid phase that contains a 2:1 ratio of B:A.
Phase diagrams can sometimes be found in the literature, but it is preferred that solubility data are generated experi-
mentally to ensure they are representative. This especially applies to feeds that contain greater than 1%–2% of impurities or
have more than two components, since the physical properties are more likely to be affected (Bamforth, 1965).
Integration and Optimization of Unit Operations. https://doi.org/10.1016/B978-0-12-823502-7.00022-0
Copyright © 2022 Elsevier Inc. All rights reserved. 1

To determine solubility experimentally, the solution is subjected to a temperature sufficient for inducing crystallization and
is allowed time to equilibrate. A solid-liquid separation is performed, and the resulting mother liquor is analyzed to
determine the residual concentration. Thus, by repeating this at various temperatures, freezing temperature versus concen-
tration can be plotted. Less preferably, a reasonable estimate of theoretical solubility data can be calculated using the van’t
Hoff Equation, where x2 is the soluble mass fraction, DHf is the heat of fusion, R is the ideal gas constant, T is saturation
temperature, and TM is the pure melting point temperature.
ln x2 ¼
DH f
RT
T
TM
1

Theoretical data can provide useful insight in the preliminary stages of development, but it is recommended that data are
later verified in the laboratory.
When developing a crystallization process, it is also essential to define the metastable and labile zones. Fig. 1.2 depicts
these zones on a phase diagram. The metastable zone indicates the range of conditions in which a solution can sustain
supersaturation. Chemical solutions with large metastable zones require concentrations that greatly exceed equilibrium
before nucleation spontaneously occurs. Supersaturated solutions occurring in the metastable zone are stable unless dis-
turbed. The labile zone represents the region beyond the threshold concentration at which the supersaturated solution will
no longer remain stable and nucleation will automatically “kick-off.” When highly supersaturated solutions self-nucleate,
they are likely to undergo a rapid, uncontrolled crystallization that tends to produce fine crystals of poor quality. In
commercial processes, this generally must be avoided. Therefore, seeding is frequently used to enable a more controlled
crystallization that prevents excessive nucleation and promotes good crystal growth. Seeding involves the addition of
homogeneous or heterogeneous crystals to act as nucleation sites for growing crystals. To be effective, seed crystals should
be added while operating in the metastable zone.
These concepts are often presented in the classic undergraduate chemistry lab experiment in which a supersaturated
solution is prepared and scratching the side of a beaker or flask initiates crystal formation. In this example, the scratching
frees tiny glass particles that can act as seed crystals. Once crystals begin to form, the solute concentration decreases and,
given enough time, the solution will reach equilibrium and supersaturation is eliminated.
FIG. 1.1 A typical binary eutectic phase diagram.
2 Integration and optimization of unit operations

By studying the phase diagram, one can infer that changing either temperature or concentration is necessary to accom-
plish supersaturation and initiate solids formation. In fact, this is the design basis of most crystallizer units. Traditional
industrial crystallization, the focus of this section, commonly operates on the basis of evaporation, indirect cooling, or
evaporative (direct) cooling. The best approach for a given system is determined by its solubility data. When the solubility
changes very little as a function of temperature, evaporation will be required for generating supersaturation.
On the other hand, when the solubility is a strong function of temperature, cooling is appropriate. A rule of thumb is to
choose evaporative crystallization when solubility decreases less than 0.005g/g°C and to use cooling crystallization when
the decrease is greater than 0.005g/g°C (Genck, 2011). Evaporative cooling is usually preferred to minimize the tendency
for fouling. However, sometimes evaporative cooling is not practical because the required vapor pressure is too low, the
material is unstable at the temperature required for evaporation, or due to other limitations. In that case, indirect cooling
should be considered. Other less common types of crystallization include salting out, reactive crystallization (precipi-
tation), and crystallization from the melt.
Crystallization processes can be designed to operate in a batch mode or continuous mode. Batch operation offers the
most flexibility and is suited well to smaller production units. Continuous operating mode has a much narrower range of
operating conditions, but it is often more cost-effective for large scale production and can be simpler to operate since it runs
at steady state. Units that produce roughly 10–15MM lbs/year or more are often candidates for continuous operation.
Rarely, batch-automatic operation is employed in some specialized systems, such as the falling film melt crystallizer.
In this case, the crystallizer itself operates in batch mode but can be integrated seamlessly into a continuous process by
using multiple holding tanks and involving complex sequencing.
The most common continuous crystallizer is a forced circulation (FC) crystallizer. This type of crystallizer involves the
simplest design and is used for straightforward processes that do not have restrictive particle size requirements. Forced
circulation crystallizers typically produce particle sizes ranging from 105 to 500mm (Genck, 2004). When larger crystal
size is important, other designs such as a draft-tube baffle (DTB) or OSLO (Krystal) are more suitable. The draft-tube baffle
design produces crystal sizes on the order of 300–4000mm, and an OSLO crystallizer gives a crystal size range of
180–4000mm (Genck, 2004). Both designs incorporate classified product removal methods that separate product based
on settling velocities to yield larger crystal size and narrower crystal size distribution (CSD). Fines destruction can also
be implemented in either of the designs to further control the CSD. Perry’s Chemical Engineers’ Handbook provides a
thorough review of various crystallizer designs. “A Clearer View of Crystallizers” by Genck is recommended for additional
reading. Vendor websites also provide useful content on this topic.
FIG. 1.2 Phase diagram showing the stable, meta-
stable, and labile zones.
Crystallization Chapter 1 3

1.1.2 Crystallizer design tradeoffs
The primary design criteria for a crystallizer include purity, yield (recovery), and capacity. Since these aspects of a crys-
tallization process are interrelated, prioritization is important for striking the right balance in developing a viable process.
For example, extreme purity may be attainable, but it often comes at the cost of yield and throughput. Extra recrystallization
steps might be required to reach the high purity, resulting in loss of yield with each additional processing step. Operations
will become more intensive and throughput will decrease. Likewise, if yield is the main focus, it might make sense to com-
promise on purity and throughput. Since high rates of recycle might be needed to achieve the yield, impurities would be
expected to go up, so achieving high purity becomes increasingly difficult. Another obvious factor is cost. If all three cri-
teria are maximized, both capital costs and operating costs can easily become prohibitive.
Nonetheless, there are measures than can and should be taken to promote optimization of the overall system in
terms of purity, yield, and capacity factors. To begin with, a crystallization process must be developed to facilitate
sufficient crystal growth. Some processes demand strict specifications for the CSD. Even if it is not specified,
acceptable particle size and shape are almost always critical factors for ensuring good performance in the crystallizer
and downstream.
Both nucleation (the production of new particles) and growth (increasing the size of existing particles) are critical to the
final CSD that is obtained. The level of supersaturation needs to be controlled carefully to give acceptable rates of nucle-
ation versus growth. When particles form, the goal is to relieve supersaturation primarily by depositing molecules on
existing particles. This results in particle growth instead of producing more small individual crystals. To achieve this, crys-
tallization should be carried out in the metastable region. Ideally, a constant, minimal amount of supersaturation should be
maintained, whether batch or continuous mode, to achieve a desirable CSD and produce consistent results. If growth is too
fast, inclusions and occlusions can occur, hindering purity. Inclusions are impurities that are chemically incorporated into
the crystal structure. Occlusions are those that are physically trapped inside a crystal.
If supersaturation is too high and operations shift into the labile region, nucleation can be induced such that significant
fines can “crash out” of solution. This leads to various problems and ultimately negatively impacts purity, yield, and
capacity. Seeding can be essential for controlling supersaturation during startup so that acceptable process performance
is attained. Otherwise, high supersaturation could be required to initiate crystal formation, potentially causing one of these
uncontrolled nucleation events in which high numbers of fines are produced.
In batch processes, seeding is often performed at the start of each batch to manage supersaturation. After startup, the
cooling or evaporation rate dictates the level of supersaturation. In addition, residence time also influences growth and
particle size. Adequate residence time for growth is essential. In continuous processes, when conditions are constant, super-
saturation is controlled by many variables including slurry density, mixing, and residence time. Since growth occurs at only
2mm/min, residence times commonly range from 1 to 10h (American Institute of Chemical Engineers, 2019). The impor-
tance of slurry density with respect to growth relates to the ability of molecules to find an existing particle upon which to
solidify and grow versus producing a new particle.
Agitation and circulation are factors that have both chemical and mechanical effects on particle size. Proper agitation
and circulation must be used to disperse supersaturation and maintain solids in suspension, but excessive mixing can cause
mechanical abrasion and crystal breakage and can potentially induce secondary nucleation. Primary nucleation refers to
crystal production that occurs due to supersaturation, while secondary nucleation occurs due to crystal-crystal collisions or
crystal-impeller collisions.
Even if reasonable design consideration has been given to the above factors, sometimes further improvements to the
CSD are needed. Product classification enhances the CSD by removing the product slurry from a location in which settling
occurs. A fines destruction system is often associated with a classifying crystallizer. Fines destruction dissolves small par-
ticles by heating (or dilution) and returns supersaturated solution to the system where it encounters existing particles that
grow larger to absorb the supersaturation. Fig. 1.3 shows a DTB crystallizer that has a classification zone and incorporates
fines destruction. In this design, larger particles are taken as product, and fine particles are removed, redissolved in the heat
exchanger, and recirculated back to the crystallizer. An elutriation leg is another type of classifier that is used in some
crystallizers to increase its classifying capability. An elutriation leg, sometimes referred to as a “salt leg,” can be integrated
into various crystallizer designs, as in Fig. 1.3. It operates based on fluidization of particles to achieve separation due to
particle size (i.e., settling velocity).
Production capacity requirements help establish the foundation for the crystallizer design. It is important to have a well-
defined material balance to develop a system with the desired capacity. Once equipment has been specified, changes to any
of the system variables can lower the system’s processing capacity. Changes to feed rate or feed concentration will directly
impact production capacity. Practical upper limits must be imposed for feed rate to ensure that the system performs well and
meets the performance targets for particle size, purity, and yield. Though most systems can manage a decrease in volumetric

throughput, there can be unintended consequences. In a continuous system, if feed rates become too low, residence times
could increase to the point that crystal breakage becomes detrimental to the CSD and can be problematic for downstream
solid-liquid separation. Inadequate heat transfer can also limit throughput, creating bottlenecks that affect downstream pro-
cessing. Feed concentration not only has a direct bearing on the material balance, but it will also affect slurry density, in
turn, affecting secondary nucleation and crystal growth and having implications on solid-liquid separation. In batch
systems, residence time is an important factor needed for establishing batch cycle time and determining capacity of the unit.
Often a crystallizer system will comprise more than one crystallizer stage. Multiple recovery stages might be necessary
to achieve the target yields while limiting the slurry density to a manageable level. Thus, additional cooling or evaporation
will be conducted in subsequent stages to recover additional cuts (crops) of crystals. This case usually offers an opportunity
for heat integration among stages. Similarly, crystal purity requirements can necessitate additional purification stages,
known as recrystallization. Recrystallization consists of dissolving crystals back into solution so they can be crystallized
again to further reject impurities and improve quality.
Recycle is common in crystallization and is primarily used to enable higher yields. It is often a good solution when one
or more of the following situations apply:
l The target for product recovery is very high.
l Solubility of the desired component in solution is high enough that it cannot be recovered at practical operating con-
ditions (temperature or pressure).
l The available utilities restrict the operating conditions such that vacuum is not low enough or the cooling water tem-
perature is not low enough to reach the temperatures needed to recover the target amount of product, or steam limitations
result in inadequate evaporation.
l The solution is subject to degradation at the temperatures needed for further concentration (when evaporation is used).
The drawbacks of recycle include the need for larger equipment (to process the higher feed rates while maintaining ade-
quate residence time) and the potential negative effect on quality since the impurity concentrations will increase. A purge is
necessary to prevent impurities from continuing to build up in solution.
At the laboratory scale, recycle can be difficult to study. At the lab stage of process development, tests are frequently
conducted with a step-by-step approach in which the various stages are run independently. Moreover, true continuous oper-
ation is often impractical at very small scales because of the difficulty in controlling very small flow rates and maintaining a
HEATER
FEED
FINES REMOVAL
CONDENSER
BOILING SURFACE
SETTLING ZONE
BAFFLE
DRAFT TUBE
ELUTRIATION LEG
RETURNED MOTHER LIQUOR
PRODUCT DISCHARGE
FIG. 1.3 DTB crystallizer with an elutriation leg.

balanced system inventory. For crystallization in particular, small continuous lab units using small tubing diameter and low
slurry flow rates are prone to plugging. Thus, recycle is best developed in the pilot plant, where operations are more rep-
resentative of a commercial process.
Another important aspect of designing a crystallizer system is solid-liquid separations. Filtration or centrifugation are
typically used, and the choice of filter or crystallizer design will depend on many factors. Crystal size and shape, density,
slurry thickness, and required dryness will all be considered. Good separation efficiency is needed to provide adequate
product purity, dryness, and yield.
Laboratory scale vacuum filters, pressure filters, and basket centrifuges are available and are good choices for prelim-
inary testing. However, these are not representative of commercial equipment. They typically only operate batchwise, and
the G-force, pressure, or vacuum conditions commonly used in a lab setting are often impractical at large scale.
Typically, 5–10wt% of mother liquor is retained on the crystals, and some solids retain significantly more. Since sep-
aration efficiency is never 100%, washing the crystals is typically carried out to help remove impurities present in the
residual mother liquor. Most often, the wash medium will be the same solvent as contained in the feed. Extensive washing
can result in decreased recovery, particularly if the product is highly soluble in the wash solvent. Therefore, the quantity of
wash water should be limited to the minimum amount needed for producing acceptable quality crystals. To mitigate this
loss, used wash water (or other wash solutions) can be recycled back to the crystallizers. Sometimes another liquid in which
the substance is relatively insoluble is used for washing to prevent dissolution, but this typically requires a solvent recovery
unit which adds cost and complexity. Solid-liquid separation and washing operations are also best studied in the pilot plant
when the equipment designs are similar to that of a commercial plant and the material balance has been defined.
1.1.3 Upstream variables affecting crystallization
Most unit operations are designed to operate properly within a limited range of conditions and when system variables drift
outside that range, it can be detrimental to performance. Crystallization is no exception. The potential for upstream var-
iations should be understood and considered in the design phase. Some flexibility in design may be feasible, but operating
constraints will need to be established to ensure acceptable system performance. Any change in the feed that enters a crys-
tallizer can potentially impact the crystallizer operation. Feed properties like feed rate, solute concentration, impurity con-
centration, viscosity, and temperature can compromise operation and affect product yield, capacity, and quality.
Generally, a batch system is more flexible and can better adjust to changes in the feed. A continuous system, however, is
intended to operate at steady state over a narrow range of conditions and is not able to tolerate significant variations in its
feed composition or other properties. Many systems are very sensitive to these changes and deviating from the standard
operating window can have a severe impact. Special accommodations should be made to either minimize changes or design
the system to handle variability. Process control strategies (Zhang et al., 2014) or additional processing steps can be imple-
mented to assure the system operates smoothly.
In the laboratory setting, range-finding experiments can be conducted to gain insight regarding acceptable ranges of
operation. Later, when multiple unit operations are integrated in the pilot plant, it becomes feasible to assess the natural
extent of variation, evaluate mitigation methods, and ultimately demonstrate that consistent performance can be achieved.
At a minimum, the ranges of feed rates, composition, and temperature of the incoming feed must be known and accounted
forinthecrystallizerdesign.Asnotedabove,achange infeedrate(throughput) caneasily upsetacrystallization system. Feed
rate changes can occur when there are interruptions or bottlenecks upstream and must be planned for in advance. If feed rate
increases and the system does not adjust, the residence time will be reduced, and crystal growth can be hindered. Supersat-
uration will rise, and although the growth rate will increase, the increase in nucleation rate is expected to be greater (Nývlt,
1992). Evaporation or cooling capacity might be inadequate, and this will be reflected in the slurry thickness and product
yield. Insufficient slurry thickness can impact crystal growth, favoring nucleation rather than growth.
Conversely, a reduction in feed rate is typically easier to manage but will result in higher residence time if no adjustment
is made, which may or may not be acceptable. Operating with significant turndown can be difficult. Lowering the liquid
level to maintain an appropriate residence time might be a good way to manage a reduced feed rate, depending on the
crystallizer type. Shutting down a continuous crystallizer should be avoided whenever possible, since it takes significant
time (10 residence times) to reach steady operation upon restart. Surge tanks might be recommended to provide more
flexibility in moderating swings in feed rates.
One can envision many scenarios in which feed concentration (i.e., solute concentration) might vary, such as when there
is a disturbance that causes lower selectivity in an upstream reaction step. Feed concentration can affect operation in several
ways. Incremental increases in feed concentration can potentially increase product yield. However, as concentration rise
becomes greater and exceeds the limits of the crystallizer design, it is possible for many problems to occur. Rising con-
centration will increase supersaturation, affecting particle number and size. As noted above, as supersaturation increases,

growth rate will increase but is unlikely to compensate for the increased nucleation rate so that more, smaller particles are
produced. Incrustations (hard deposits which form on the crystallizer internals) can occur due to higher levels of super-
saturation. Slurry thickness will increase and could create mixing or plugging issues.
In general, the slurry thickness should be limited to around 20–35wt% of solids. However, DTB crystallizers can handle
somewhat higher slurry thickness of up to 25%–50% (Genck, 2011). This restriction frequently requires that the product is
recovered in multiple stages. The number of stages of recovery is determined primarily based on the slurry thickness that
can be practically handled and, to a lesser extent, on the amount of recycle that is acceptable. Recycling saturated mother
liquor can help manage feed concentration and slurry thickness to some degree but will drive up the crystallizer size
requirement and increase the level of impurities.
In other instances, changes in impurity concentrations could occur. Switching to feed from a new supplier, processing a
different batch of fermentation broth, charging new catalysts to upstream reactors, or cycling of reaction conditions are all
plausible scenarios for causing changes in impurities. Changes in impurity profiles can affect the size, shape, and most
notably, the purity of the final crystals.
The primary method by which impurities are incorporated during crystallization is via adsorption of residual mother
liquor. In this case, increased washing is usually an effective remediation strategy. In contrast, impurities caused by occlu-
sions or inclusions cannot typically be removed via washing. Occlusions typically trap only 5wt% of mother liquor, but
when impurity concentrations rise, this can have a significant impact on final purity (Urwin et al., 2020). Inclusions are less
common due to the limitation of fitting a foreign molecule into a crystal lattice but can occur when impurities have similar
structures and charges as the primary solid product. Both occlusions and inclusions cause surface defects that can alter
crystal size and morphology. Reducing supersaturation and slowing down the growth rate can help minimize these types
of impurities.
It is difficult to anticipate the effect of various impurities on the crystal quality and morphology without conducting
careful studies. Partition coefficients of key impurities can be determined experimentally and can help specify the
acceptable ranges that can be tolerated in the incoming feed.
Temperature is another critical condition that must be defined for the incoming feed. Feed liquor must be held at several
degrees above its saturation temperature to prevent crystallization in the feedline. Insulation of the feed line is usually
recommended, and heat tracing may be required in some cases.
Maintaining the energy balance of the system is essential as it impacts the production rate and supersaturation. Crys-
tallizer contents heating or cooling must be done in such a way as to minimize supersaturation. For surface cooling and
indirect cooling crystallizers, heat exchangers must limit the DT to only a few degrees to prevent high supersaturation and
incrustations. For evaporative crystallizers, the temperature increase should be kept low for similar reasons; high super-
saturation at a boiling surface can cause flashing and entrainment and will contribute to scaling.
Moreover, viscosity is a function of temperature, as well as composition, concentration, and slurry density. The vis-
cosity of the crystallizer contents influences the hydrodynamics and mass transfer and has consequences for the growth
kinetics. High viscosity can also interfere with nucleation. Growth will be slower as viscosity increases, so residence times
need to be longer and particles are generally smaller. Forced-circulation crystallizers and scraped-surface crystallizers can
be good choices for processing high viscosity slurries.
In addition to designing a robust crystallizer system that can adjust to changes, various upstream control strategies can
be implemented if significant excursions in feed conditions are anticipated. Surge tanks can help manage short-term inter-
ruptions and smooth out fluctuations in feed composition and concentration. Preconcentration or dilution might be nec-
essary to provide a more consistent feed concentration and ensure steady state operations in the crystallizer. Finally, if
high levels of impurities are a concern, distillation might be necessary to produce a feed that is amenable to crystallization.
1.1.4 Impact on downstream operations
The performance of the crystallizer sets the requirements for downstream operations. Solid-liquid separation, drying, solids
transport, and dissolution are all directly affected by the CSD and other crystal properties.
Solid-liquid separation processes are quite sensitive to the properties of the suspension. In particular, the particle size,
size distribution, and morphology of the crystals are of great consequence to the performance of centrifuges and filters,
common choices for crystallization processes. The ability for the mother liquor to drain or separate from the solids depends
on the size distribution and the tendency for packing of the particles, the viscosity of the liquid, the density difference
between solid and liquid phases, the particles’ surface properties and interactions with surrounding fluid, and the method
of separation. Solids packing density (compressibility) dictates porosity of the cake and is a key concern in solid-liquid
separations. With regular-shaped particles, and particles with a CSD that allows tight packing, permeability becomes
low such that mother liquor removal is impeded. These points regarding dewatering also apply to the washing step, which

is typically carried out in the same equipment. Cake cracking is another common issue that arises for both mother liquor
removal and washing steps during pressure filtration, vacuum filtration, or centrifugal filtration. This is a particular problem
for washing because the wash fluid will flow out through the cracked bed of solids, circumventing the bulk of the cake and
rendering washing ineffective.
The choice of solid-liquid separation method will depend of the slurry thickness, crystal size, the tendency for moisture
retention, the required level of dryness, and the acceptable loss of yield. Some types of centrifuges require higher slurry
density such that a prethickening (preconcentration) step may be required. More discussion of solid-liquid separation
methods and guidance for selection of an appropriate method can be found in other sections of this book.
Drying will also certainly be affected by the performance of the crystallizer, and subsequently, the solid-liquid sepa-
ration step. It is important to determine the level of residual moisture that will be present in the crystals before designing the
dryer system in order to select the appropriate drying equipment and to define the energy requirements. The crystal size and
shape established in the crystallizer not only affect the quantity of moisture, but also influence caking and agglomeration,
which are also important considerations for drying.
Conveying of solids and transport of solids through bins, hoppers, and chutes can be challenging even in the best of cir-
cumstances. Even when solids properties have been thoroughly studied and the system is well-defined, problems frequently
ariserelatedtocaking,plugging,and otherissuesthatresultinblockedflow.Therefore,upstreamchangescan beproblematic
to solids transport processes. Flowability is essential for transferring the product from solid-liquid separation to downstream
operations. Particle morphology, CSD, and dryness will have a major impact on transport manageability and handling.
Moreover, packaging of the final product will be subject to whatever conditions exist upstream that affect the physical
properties of the solids. In addition to potentially creating solids handling issues, changes to these properties can result in a
product that does not meet specifications, can affect the utility of the final consumer product, and even create safety
problems. For example, a reduction in particle size can lead to problems with dusting which could present a health hazard
as well as create a combustible environment. To prevent these hazards consult a company with expertise in solids handling
should be consulted when materials are prone to dusting, breakage, and small particle size. Further, clumping can present
other types of challenges and could have consequences on product quality and usability for end consumers.
Dissolution of the purified solids is necessary in some applications. The time required for dissolution is a function of the
CSD and may vary greatly if changes in particle size occur. Fortunately, upstream upsets most often result in a decrease in
average particle size, which is favorable for dissolution. However, agglomeration can lengthen dissolution times. Some-
times crystallization is an intermediate step to recover a dissolved component that will undergo reaction downstream. The
level of mother liquor carryover will impact purity for subsequent operations. Recrystallization also involves dissolution.
Again in this scenario, the amount of residual mother liquor can affect the overall purity achieved.
Problems that originate either upstream or within the crystallizer system can proliferate far downstream, creating chal-
lenges for other processing steps and potentially compromising final product quality and operating capacity. Solid-liquid
separations, drying, packaging, dissolution, and downstream reaction steps all rely on steady upstream operations and
achieving acceptable crystallizer performance. The best designs will account for the crystallizer within the broader context
of upstream and downstream operations. A holistic approach to process development will give the optimum operational
performance.
1.2 Pilot scale crystallization studies
1.2.1 Objectives for a pilot plant
The pilot plant is important for demonstrating the technology in equipment similar in design to a commercial unit and
essential for gathering scale-up data. Some aspects of crystallizer design are best suited for development at the pilot plant
scale. This includes:
l Running multiple stages continuously
l Studying recycle
l Verifying performance of solid-liquid separation equipment
l Implementing vapor recompression or other energy recovery techniques
l Finalizing material and energy balances
l Integration with other unit operations
l Developing a control strategy
For practical reasons, the scale at which a pilot plant operates is amenable to continuous operation more than laboratory
scale. A lower limit for continuous operation exists due to the issue of removing slurry through pipes of practical

dimensions while simultaneously maintaining the velocity required to keep solids suspended. Running the crystallizers
with continuous feed and product removal and running solid-liquid separation continuously is something that can often
be accomplished in the pilot plant, even if some special accommodations are necessary. To overcome scale limitations
and achieve the velocities needed, keeping slurry density lower can be a useful strategy. Intermittent slurry removal per-
formed at a high rate on a semicontinuous basis is another option that is frequently implemented at relatively small scales.
In addition, since the pilot plant typically incorporates the use of a more advanced control system for automating oper-
ation, and because larger volumes are processed, it is easier to maintain inventory in the system and operate with a stable
material balance. Thus, it is also more convenient for studying recycle, which is critical to closing the material balance and
providing representative scale-up data needed for designing the commercial system. Note that for continuous crystalli-
zation, it takes 10 residence times to begin to approach steady state. Therefore, demonstrating performance and gathering
design data should be based on operations after this condition has been met.
Verifying the performance of the solid-liquid separation equipment will be essential for ensuring effective solid-liquid
separation at the commercial scale. Using the same basic equipment design and operating conditions as the commercial
plant, pilot plant testing can be used to determine the percent removal of mother liquor, understand the purity that can be
achieved, and develop an effective washing protocol.
Since multiple stages can be tied together and operated continuously, it is possible to demonstrate heat integration and
other energy recovery techniques at the pilot plant scale. Performing this at a pilot scale can be beneficial for verifying the
energy balance and refining the economics (OPEX) of the commercial plant. Energy efficiency is a particularly important
aspect of evaporative crystallizer design. Using multiple-effect evaporation can offer significant energy savings compared
to using single effect/stage systems. Thermal vapor recompression (TVR) or mechanical vapor recompression (MVR) can
be used for further reduction in steam usage in conjunction with multieffect evaporation. MVR is more commonly practiced
in crystallization, but economics limit it to use in large-scale systems with low boiling point elevations and high energy
demand (Genck, 2019).
Finally, integrating crystallization with other unit operations at the pilot plant scale is valuable for considering how it fits
upstream and downstream processing. It enables a more comprehensive understanding of the consequences of making
changes and the cycling or swings in conditions that can be expected. Further, it gives engineers an opportunity to develop
a practical control scheme for ensuring good performance and maintaining stable operating conditions. Liquid level control,
pressure/temperature control, slurry density control, steam or cooling water flow control, and feed or product flowrate
control are commonly employed in crystallizers. A process control scheme can be designed to hold one critical condition
steady and allow another less critical one to vary in order to maintain operations. For example, when feed rate is high, the
level in the crystallizer might be controlled at a higher point to maintain an acceptable residence time. Likewise, if feed
solute concentration increases, the operating temperature could be adjusted to ensure slurry density does not exceed a prac-
tical upper limit.
1.2.2 Scale-up criteria
Scaling up based on geometric similarity is not recommended for crystallizers. There are many design criteria that should be
considered, and scaling up one variable often interferes with the scale-up for other variables. Ideally, supersaturation, slurry
density, residence times, and various mixing parameters would all be identical, across all scales of operation. However, all
parameters do not scale proportionally, so achieving equivalency for all design parameters upon scale-up is an impossi-
bility. Therefore, compromises must be made to ensure acceptable performance of scaled-up processes.
As has already been discussed, supersaturation, the driving force for crystallization, is the predominant factor con-
trolling nucleation and growth. Growth and nucleation processes compete for relieving supersaturation. An acceptable
balance must be achieved to enable good crystallizer performance. Changing the level of supersaturation can result in
any number of issues. It can alter the CSD, cause incrustations, influence yield and purity, and even lead to a new particle
morphology. Thus, supersaturation needs to be controlled by ensuring desirable slurry density and proper mixing.
Mixing is a critical factor in crystallizer scale-up, as it directly impacts supersaturation, secondary nucleation, and the
distribution of slurry. The role of mixing must be considered on various levels for crystallizer scale-up. It is sometimes
described as being divided into three categories: macromixing, micromixing, and mesomixing. Macromixing refers to
the overall recirculation of slurry in the vessel. This impacts residence time distribution and suspension and circulation
of the slurry. It involves blending of the crystallizer contents to reduce gradients in temperature, concentration, supersat-
uration, and slurry density within the vessel. Micromixing occurs at the molecular level and controls nucleation and growth.
Mesomixing refers to mixing that occurs at the feed inlet, but this relates mainly to precipitation or antisolvent crystalli-
zation processes (Genck, 2003).

A general goal for mixing includes dispersing supersaturation such that primary nucleation is avoided. Instead, sec-
ondary nucleation should be the main source of new nuclei. This can occur due to crystal-impeller contact or crystal-crystal
contact. In either case, mixing has a major effect. It is also important to have available crystal surface area at locations of
high supersaturation, such as at the boiling surface in an evaporative crystallizer or at the heat transfer surface in an indirect
cooling crystallizer. Again, mixing is critical to achieving this goal.
Clearly, the agitator plays a significant role in mixing. Defining the agitator parameters is essential for effective scale-
up. Common scale-up strategies for the agitator include using constant tip speed or constant input of power per unit volume
(P/V). It can be useful to consider the consequences of each approach to determine which is more practical or to find a
compromise between the two methods. Usually, the larger scale crystallizer will end up with slower rates of blending
and circulation, higher tip speed, and lower rotational speed. The average shear rate is often lower, which can lead to larger
crystal sizes.
While crystallizer scale-up is complex, a lot of information can be gathered in the pilot plant to help guide the scale-up
process. Carefully evaluating the mixing requirements will be necessary. Sometimes, it can be useful to change the dimen-
sions of the pilot unit to approach those of the intended commercial unit so that scale-up is more straightforward. For
example, per Genck (2003), a smaller D/T ratio (where D is agitator diameter and T is tank diameter) is usually more prac-
tical for large scale operation. Using a smaller D/T to make the pilot unit more geometrically similar to a commercial unit
can facilitate development of a more practical mixing scheme.
1.3 Commercialization of crystallization processes
By this stage the technology is well-defined. However, there can be a number of surprises that arise when implementing a
commercial crystallization system. Since crystallization can be highly empirical, it is not unusual to observe some changes
in performance upon scale-up.
In one plant in the United States, a new, undesirable polymorph appeared when the process was run at commercial scale.
As a result, the crystals displayed a needle morphology, and solid-liquid separation became much more challenging. After
significant troubleshooting, oxygen exposure was determined responsible for the new crystal form. This was difficult to
entirely eliminate in the full-scale process. Thus, the plant continued to struggle with this for many years.
Working with an established vendor with scale-up experience can reduce risks related to scale-up and can be vital to
successful commercial operation. These vendors can perform tests in equipment that has undergone careful scale-up
studies. Sometimes vendors will provide a performance guarantee for their equipment, so long as the process variables
are maintained within a specified window. This provides a higher level of comfort when purchasing expensive equipment,
but the guarantee is often subject to constraints.
GEA and Swenson are considered industry leaders in industrial crystallization, each offering a variety of crystallization
equipment as well as supporting testing services. GEA’s product line includes traditional FC, DTB, and OSLO-type crys-
tallizers, along with crystallizers suited for melt crystallization and freeze concentration. Swenson designs batch vacuum
crystallizers in addition to its more common FC and DTB designs.
Other reputable vendors include Sulzer and Armstrong Chemtech Group, both having experience with fractional melt
crystallization. Sulzer has several melt crystallizer designs, including static, falling film, suspension, and freeze concen-
tration. Armstrong Chemtech offers a scraped surface crystallizer that can be used for solution or melt crystallization appli-
cations. For melt crystallization, they can use the scraped surface heat exchanger in combination with a hydraulic wash
column to provide ultra-high purity molten product. Many of these companies provide relevant literature on their websites
that is useful when pursuing a crystallization project.
Improvement of existing commercial crystallization processes is often desired. Increasing the yield, capacity, or quality
is a common request for engineering companies working in this field. Reduction of energy usage or environmental impact is
also sometimes of interest. In one plant, the desire for increased production capacity of a specific product line prompted
reevaluation of an old crystallization process. The desire to increase throughput in existing equipment led to a transitioning
from batch operation to continuous. Studying the system in the laboratory was an effective way to relearn the critical aspects
of the dated process technology and determine how to make required system modifications.
Maintaining smaller scale operating capability is useful for troubleshooting issues that arise in the plant. Either labo-
ratory or pilot facilities can provide helpful process support to address unforeseen issues during commercial startup and
beyond. These systems are easier to study since they can be isolated from other operations and modified more easily with
the ability to control and change variables. A cost savings can be realized by leveraging the speed and flexibility of small-
scale testing to correct issues at the commercial scale. Reduction in downtime or fewer off-spec batches can quickly com-
pensate for the cost of small-scale investigations.

References
American Institute of Chemical Engineers. (2019). AIChE academy. CH110: Crystallization operations.
Bamforth, A. W. (1965). Industrial crystallization. Leonard Hill.
Genck, W. (2003). Optimizing crystallizer scaleup. Chemical Engineering Progress, 36–44. June.
Genck, W. (2004). Guidelines for crystallizer selection and operation. Chemical Engineering Progress, 100(10), 26–32.
Genck, W. (2011). A clearer view of crystallizers. Chemical Engineering Magazine, 7, 28–32.
Genck, W. J. (2019). Liquid-solid operations and equipment. In D. W. Green, M. Z. Southard (Eds.), Perry’s chemical engineers’ handbook (9th ed., pp.
18-41–18-47). New York: McGraw Hill.
Nývlt, J. (1992). Design of crystallizers. CRC Press.
Urwin, S. J., Levilain, G., Marziano, I., Merritt, J. M., Houson, I., Ter Horst, J. H. (2020). A structured approach to cope with impurities during industrial
crystallization development. Organic Process Research Development, 24(8), 1443–1456. https://doi.org/10.1021/acs.oprd.0c00166.
Zhang, H., Lakerveld, R., Heider, P. L., Tao, M., Su, M., Testa, C. J., et al. (2014). Application of continuous crystallization in an integrated continuous
pharmaceutical pilot plant. Crystal Growth Design, 14(5), 2148–2157. https://doi.org/10.1021/cg401571h.

Chapter 2
Fermentation and downstream processing:
Part 1
Gabelman Process Solutions, LLC, West Chester, OH, United States
2.1 Introduction
Fermentation is the chemical transformation of one substance into another by the action of enzymes, which are produced
by microorganisms. The microorganism employed is analogous to an inorganic catalyst in a traditional chemical
reaction, although the analogy certainly is not perfect. Strictly speaking, the term fermentation applies only to anaerobic
transformations, while those that use oxygen are called respiration. However, in common usage of the term, fermentation
refers to both types of biotransformations. Such processes have been practiced since the dawn of time. Bioconversion of
grains and fruits to beer and wine has been documented dating from around 7000 BC in China. Cheese, prepared from
fermented milk, was consumed during ancient times in Eastern Europe and Central Asia, and the use of yeast in the
preparation of leavened bread also dates back to prehistoric times (Benvenuto, 2019). Numerous other fermented foods
followed, including various East Asian consumables, yogurt and other fermented milk products, pickles, sauerkraut, and
vinegar, to name a few.
Starting in the early twentieth century, people began to understand the utility of fermentation for applications
beyond alcoholic beverages and fermented foods. The well-known acetone-butanol-ethanol (ABE) fermentation,
developed by Weizmann (1919), was the main source of the acetone used in explosives during World War
I. For several decades, the process was a viable commercial source of these three solvents, but it was gradually
replaced by more economical chemical routes in the 1950s and 1960s. The advent of penicillin and other
fermentation-derived antibiotics after World War II changed medicine forever, and undoubtedly saved countless
lives. Other fermentation products developed in relatively recent years include amino acids, enzymes, organic acids,
vitamins, and food gums. The global fermentation market was $58.68 billion in 2018 (Grand View Research, 2019),
and is expected to grow at a compounded annual growth rate (CAGR) of more than 5.25% until 2026 (Market
Watch, 2020). A partial listing of products made by fermentation is given in Table 2.1. Fermentation-derived
enzymes are listed separately in Table 2.2.
Fermentation encompasses an exceedingly wide range of topics and information, too much to cover here. Indeed,
entire books have been written on the subject, and several of these are cited in the discussion that follows. The focus
of this chapter and Chapter 3 is primarily on the biochemical engineering aspects of fermentation, with some discussion
of microbiology and biochemistry. The scope is limited to microbial fermentation (bacteria, yeasts, and filamentous
fungi), with only a passing mention of plant, mammalian, and insect cultures. While these chapters cover anaerobic fer-
mentation to some extent, the main emphasis is on aerobic processes, which present considerably more biochemical
engineering challenges.
2.2 Microbiology and biochemistry basics
Microbiology is the study of microorganisms, which are organisms that are too small to be seen by the naked eye. About
half of their dry weight consists of proteins (although the percentage varies widely), with polysaccharides, nucleic acids,
and lipids making up the rest. Microorganisms are broadly classified as procaryotes or eucaryotes, characterized by the
absence or presence, respectively, of a membrane-bound nucleus where genetic information is stored. A schematic drawing
of each type of microbial cell is shown in Fig. 2.1.
Integration and Optimization of Unit Operations. https://doi.org/10.1016/B978-0-12-823502-7.00015-3
Copyright © 2022 Elsevier Inc. All rights reserved. 13

Characteristics of bacteria, yeasts and molds, traditionally the industrially important types of microorganisms, are
summarized in Table 2.3. Bacteria are smaller, faster growing, and less complex than yeasts or molds. They reproduce
asexually via binary fission, with a doubling time of approximately 20–30min. Bacteria exist in three general morphol-
ogies: bacilli (rods), cocci (spheres), and spirilla (spirals) (see Fig. 2.2). The particular morphology of a fermentation
culture is readily discernable under the microscope, and the presence of cells with unexpected morphology indicates
TABLE 2.1 Partial list of products made by fermentation (enzymes excluded;
see Table 2.2).
Product Examples, comments
Ethanol Alcoholic beverages, biofuel
Organic acids Lactic (preservative, curing or flavoring agent in foods; monomer for
polylactic acid, a biodegradable plastic), citric (acidulant in
beverages), acetic (vinegar)
Antibiotics and other
pharmaceuticals
Penicillin, erythromycin, streptomycin, chloramphenicol,
tetracycline, human insulin, human growth hormone, antibodies,
nucleic acid products, vaccines, human serum albumin, Taxol
Amino acids Monosodium glutamate (flavor enhancer), lysine (animal feed)
Vitamins Vitamins C (ascorbic acid), B2 (riboflavin), B12 (cobalamin)
Food gums These are polysaccharides that are used as texturing agents and
emulsifiers in salad dressings, sauces and beverages. Examples:
xanthan gum, gellan gum
Fermented foods Bread, cheese, yogurt, soy sauce, pickles, sauerkraut, vinegar and
more; Wikipedia lists over 150 fermented foods
Single cell yeast protein Used as animal feed, or as a precursor to autolyzed yeast extract, a
food flavoring
Biopesticides These are microorganisms, or chemicals derived from
microorganisms, that are used in the same manner as chemical
pesticides but are more environmentally friendly. An example is the
bacterium Bacillus thuringiensis, whose toxin has been integrated into
the genome of corn and other crops to render them resistant to pests
TABLE 2.2 Partial list of fermentation-derived enzymes.
Product Function Applications
Amylase Starch hydrolysis Food manufacturing, detergents, textiles
Cellulase Cellulose hydrolysis Fruit juice clarification, textiles, detergents, drying
of coffee beans
Invertase Hydrolyzes sucrose to glucose
and fructose
Candy
Glucose
isomerase
Converts glucose to fructose High fructose corn syrup (HFCS)
Pectinase Hydrolyzes pectin Fruit juice clarification
Proteinase Hydrolyzes proteins to amino
acids
Chilled beer (haze removal), textiles, autolyzed
yeast, detergents

contamination by a foreign organism. While there are seemingly countless examples of industrially important bacteria,
one that is particularly notable is Escherichia coli, a rod-shaped bacterium that is not only widely used in fermentation
processes, but is also a resident of the human gut, and the cause of numerous food poisoning incidents. A distinguishing
characteristic of bacteria is their Gram reaction. When stained with the dye crystal violet, followed by treatment with an
iodine solution then washing with alcohol, Gram positive organisms retain the purple color while Gram negative ones do
not. Active bacterial cells exist in what is known as the vegetative state, but some species revert to spores (specifically,
endospores) under adverse environmental conditions. This dormant form cannot reproduce, but it can withstand difficult
Membrane-
enclosed nucleus
Nucleolus
Mitochondrion
Capsule
Flagellum
Cell Wall
Cell Membrane
Ribosomes
Nucleoid
(some prokaryotes)
(in some eukaryotes)
FIG. 2.1 Cell structure of eukaryotes (left) and prokaryotes (right). The simpler prokaryotes contain no internal organelles, and the nucleoid contains the
genetic material. In the more complex eukaryotes, genetic material is housed in the membrane-bound nucleus. The mitochondrion (absent in prokaryotes)
is considered the power plant of the cell because oxidations are carried out there to generate energy. Both types of organism contain ribosomes, where
protein synthesis occurs. (Source: https://commons.wikimedia.org/wiki/File:Celltypes.svg.)
TABLE 2.3 Characteristics of microorganisms used in industrial fermentation processes.a
Bacteria Yeasts Molds
Classification Prokaryote Eukaryote Eukaryote
Approximate size, mm 0.5–3 1–5 Diameter: 5–15
Length: 50–5000
Reproductive method Binary fission Binary fission, budding Growth of hyphae
Approximate doubling
time, minutesb
20–30 90 Not applicable
Morphologies Rods, spheres, spirals Many, including spheres,
ellipses, cylinders
Filaments, pellets
Specific gravity 1.05–1.1 1.05–1.1 1.05–1.1
Weight (g/cell) 1012
1011
Variable
Composition, %
Protein 65–75 45–55 25–55
Nucleic acid 15–25 5–12 5–10
Carbohydrate and lipid 5–30 10–50 10–50
Examples Acetobacter, Bacillus,
Corynebacterium, Pseudomonas
Saccharomyces,
Cryptococcus, Candida
Aspergillus, Penicillium,
Xanthomonas
a
Adapted from Massachusetts Institute of Technology Summer Session. (1984). Fermentation technology. Cambridge MA: Massachusetts Institute of Technology
Summer Session.
b
Values vary widely, depending on medium components, temperature and other environmental conditions.
Fermentation and downstream processing Chapter 2 15

conditions such as extreme dryness, heat, and high levels of some toxins, for periods of years, then germinate and grow
when conditions are more favorable. Spores are harder to kill during sterilization, a fact that must be considered in
equipment and process design.
Yeasts, a subset of the biological kingdom of fungi, are a step up from bacteria on the evolutionary scale. Yeasts
reproduce asexually by binary fission or budding. In the latter process (see Fig. 2.3), a small daughter cell grows on
the side of the parent, and eventually separates into an independent cell. Sexual reproduction is also possible. Saccharo-
myces cerevisiae is arguably the most widely used industrial yeast, with various strains employed in the production of beer
(brewer’s yeast), wine, and bread (baker’s yeast).
Molds, also known as filamentous fungi, are a more advanced subset of the kingdom of fungi. Unlike yeasts and
bacteria, these microorganisms do not grow as discrete cells. Instead, they develop threads or filaments (known as
hyphae) that divide repeatedly along their length to form long, branched chains containing multiple cells, sometimes
of different-but-related types (see Fig. 2.4). These chains become intertwined to form a network, called a mycelium.
The size of these networks is limited by the shear forces exerted by the mixer, which break the mycelia into discrete
entities called mycelial pellets. Unlike bacteria and yeasts, this morphology leads to fermentation broths that are viscous
and non-Newtonian (often pseudoplastic, or shear-thinning (Stanbury, Whitaker, Hall, 2017)), presenting challenges
to achieving uniform mixing and adequate rates of diffusion of nutrients (notably oxygen) and products. The problem is
compounded by the size of the mycelial pellets (see Table 3.3), which can be large enough to result in starvation or
product accumulation near the center. These challenges are addressed primarily by judicious selection and design of
the mixer and associated components. Products made using filamentous fungi include food gums (polysaccharides), anti-
biotics, and organic acids. Examples are Aspergillus niger (citric acid), Penicillium chrysogenum (penicillin), and
Xanthomonas campestris (xanthan gum).
Microbial growth can be aerobic or anaerobic, meaning in the presence or absence of oxygen, respectively. Aerobic
organisms derive energy by cellular respiration, with oxygen acting as the final electron acceptor in the formation of carbon
dioxide and water. With anaerobes, the final electron acceptor is a compound other than oxygen, and energy generation is
less efficient. Obligate aerobes cannot grow without oxygen, while obligate anaerobes are poisoned by it. Facultative
anaerobes can grow with or without oxygen, and follow different biochemical pathways in its presence or absence. For
example, S. cerevisiae and other yeasts generate predominantly ethanol when air is not supplied, and mainly yeast biomass
when aerated (Visser, Scheffers, Batenburg-van der Vegte, van Dijken, 1990). There are also aerotolerant anaerobes,
which are indifferent to the presence of oxygen, and microaerophiles, which need oxygen but only in small amounts.
The latter require only about 1%–10% oxygen, much lower than the 21% present in air (Lumen Microbiology, n.d.).
Aerobic growth on an industrial scale presents numerous challenges, mainly attributable to the exceedingly low solubility
of oxygen in water, and the considerable amount of heat evolution.
FIG. 2.2 Bacterial cell morphologies. (A) bacilli (rods); (B) cocci (spheres); (C) spirilla (spirals). (From Sarles, W. B., Frazier, W. C., Wilson, J. B.,
Knight, S. G. (1956). Microbiology, general and applied.)

FIG. 2.3 Sequence of budding of the yeast Saccharomyces cerevisiae. (From Sarles, W. B., Frazier, W. C., Wilson, J. B., Knight, S. G. (1956). Micro-
biology, general and applied.)

The energy source for all cellular activities is adenosine triphosphate (ATP) (see Fig. 2.5), sometimes referred to
as the energy currency of the cell. ATP undergoes dephosphorylation to adenosine diphosphate (ADP) or monopho-
sphate (AMP) to generate energy, then AMP and ADP are phosphorylated to regenerate ATP. In the absence of
oxygen, this occurs upon conversion of glucose to pyruvate, a process known as glycolysis. The most common gly-
colytic pathway is the Embden–Meyerhof pathway (see Fig. 2.6), which produces two moles of ATP per mole of
glucose consumed. The pyruvate is further metabolized either to lactic acid or ethanol. Lactic acid formation, by Lac-
tobacillus and other lactic acid bacteria, is responsible for the souring of milk and for the acidic taste of sauerkraut
(fermented cabbage). Conversion of pyruvate to lactic acid also occurs in skeletal muscle tissue during vigorous
physical activity. Ethanol production by yeast is important not only in beer and wine production, but also for use
as a gasoline supplement.
If oxygen is available, aerobic organisms derive energy from cell respiration, a highly efficient pathway consisting
of the three stages shown in Fig. 2.7. In the first stage, pyruvate obtained from glycolysis is decarboxylated to form an
acetic acid complex called acetyl coenzyme A (acetyl-CoA). In the second stage, this complex enters a cyclic pathway
known as the tricarboxylic acid (TCA) cycle, also called the citric acid cycle or the Krebs cycle. Other molecules that
are capable of forming acetyl-CoA, such as amino acids and fatty acids, can also be metabolized using the TCA cycle.
Output of the TCA cycle includes carbon dioxide and high-energy hydrogen atoms. In the third stage of respiration, the
hydrogen atoms separate into protons and energy-rich electrons, which then move along a chain of electron-carrying
molecules known as the respiratory chain. This process is called electron transport, and the generation of ATP that
occurs is referred to as oxidative phosphorylation. The final electron acceptor is oxygen, which is reduced upon com-
bination with hydrogen to form water. With 15 moles of ATP generated per mole of pyruvate consumed, respiration is a
much greater energy producer than glycolysis. The Gibbs free energy change upon conversion of glucose to lactate is
only 47.0kcalmol1
, while the free energy change upon complete oxidation of glucose to carbon dioxide and water is
686kcalmol1
(Lehninger, 1982).
FIG. 2.4 Branched growth of hyphae in molds, also
known as filamentous fungi. (From Bailey, J. E.,
Ollis, D. F. (1986). Biochemical engineering funda-
mentals. McGraw Hill.)

FIG. 2.5 The molecular structure of adenosine triphosphate (ATP), the energy source for cellular metabolism. (From Wikipedia. (2022). Adenosine
triphosphate. https://en.wikipedia.org/wiki/Adenosine_triphosphate.
FIG. 2.6 The Embden–Meyerhof pathway.

These are just examples of the multitude of biochemical pathways, consisting of literally thousands of enzyme-
catalyzed reactions, that a microbial cell uses for all of its metabolic activities. Intermediates and products generated
by these pathways are called metabolites. Intracellular products remain inside the cell, while extracellular ones are
excreted into the fermentation broth. The former must be released by rupturing the cells (usually mechanically, using
high shear homogenizers) early in the downstream purification process, a complexity that is avoided with extracellular
products. Some microorganisms contain flexible pathways that can generate a range of products, depending on the pre-
cursor that is supplied. For example, methanol-utilizing yeasts such as Candida boidinii are capable of transforming
methanol to successively more oxidized molecules, namely formaldehyde, formic acid, and finally, carbon dioxide,
deriving energy along the way. The enzyme catalyzing the first step can also oxidize other low molecular weight
alcohols, but the enzymes for the subsequent steps are active only on the one-carbon molecules. Consequently, aldehydes
of commercial interest can be made to accumulate by feeding the corresponding alcohol, e.g., acetaldehyde can be pro-
duced from ethanol (Gabelman Luzio, 1997; Sahm Wagner, 1973). In other cases, the metabolism can be directed to
the desired pathway by changes in the composition of the growth medium, or an environmental condition such as pH or
dissolved oxygen (DO) concentration. An example of medium manipulation is found in the overproduction of glutamate
(a commercially important food flavor enhancer, discussed at some length in Chapter 3), induced in Corynebacterium
glutamicum by limiting the amount of biotin available (Kimura et al., 1999). An example of environmental control, men-
tioned previously, is the switchover from biomass growth to ethanol production in yeast when the supply of oxygen is
removed.
Stage 1
Amino
acids
Pyruvate
2H
Acetyl-CoA
CO2
CO2
CO2
2H 2H 2H 2H
ATP
ATP
ATP
H2O
Fatty
acids
CO2
NH3
Oxaloacetate
Malate
Fumarate
Succinate
GTP
NADH
NADH
dehydrogenase
Ubiquinone
Cytochrome b
Cytochrome c1
Cytochrome aa3
2H+
+ ½02
ADP + Pi
ADP + Pi
ADP + Pi
Cytochrome c
Succinyl-
CoA
D-Ketoglutarate
Isocitrate
cis-Aconitate
Citrate
Stage 2
Stage 3
FIG. 2.7 Cell respiration. (Redrawn from Lehninger, A. L. (1982). Principles of biochemistry. Worth.)

Barry A. Perlmutter - Integration and Optimization of Unit Operations_ Review of Unit Operations from R&D to Production_ Impacts of Upstream and Downstream Process Decisions-Elsevier (2022).pdf

Barry A. Perlmutter - Integration and Optimization of Unit Operations_ Review of Unit Operations from R&D to Production_ Impacts of Upstream and Downstream Process Decisions-Elsevier (2022).pdf

Recommended

Recommended

More Related Content

Similar to Barry A. Perlmutter - Integration and Optimization of Unit Operations_ Review of Unit Operations from R&D to Production_ Impacts of Upstream and Downstream Process Decisions-Elsevier (2022).pdf

Similar to Barry A. Perlmutter - Integration and Optimization of Unit Operations_ Review of Unit Operations from R&D to Production_ Impacts of Upstream and Downstream Process Decisions-Elsevier (2022).pdf (20)

Recently uploaded

Recently uploaded (20)

Barry A. Perlmutter - Integration and Optimization of Unit Operations_ Review of Unit Operations from R&D to Production_ Impacts of Upstream and Downstream Process Decisions-Elsevier (2022).pdf