© 2024 IWA Publishing

This is an Open Access book distributed under the terms of the Creative Commons
Attribution-Non Commercial-No Derivatives Licence (CC BY-NC-ND 4.0), which
permits copying and redistribution in the original format for non-commercial
purposes, provided the original work is properly cited.
(http://creativecommons.org/licenses/by-nc-nd/4.0/). This does not affect the rights
licensed or assigned from any third party in this book.

This title was made available Open Access through a

partnership with Knowledge Unlatched.

IWA Publishing would like to thank all the libraries for

pledging to support the transition of this title to Open
Access through the 2024 KU Partner Package program.
Experimental Methods for
Membrane Applications
in Desalination and Water Treatment
Sergio G. Salinas-Rodríguez
Loreen O. Villacorte
Experimental Methods for Membrane Applications in

Desalination and Water Treatment

 Experimental Methods for
Membrane Applications in
Desalination and Water Treatment

SERGIO G. SALINAS-RODRÍGUEZ

LOREEN O. VILLACORTE
Published by: IWA Publishing
Unit 104 – 105, Export Building
1 Clove Crescent
London E14 2BA, UK
Telephone: +44 (0)20 7654 5500
Fax: +44 (0)20 7654 5555
Email: [email protected]
Web: www.iwapublishing.com

First published 2024

© 2024 IWA Publishing

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as
permitted under the UK Copyright, Designs and Patents Act (1998), no part of this publication may be
reproduced, stored or transmitted in any form or by any means, without the prior permission in writing
of the publisher, or, in the case of photographic reproduction, in accordance with the terms of licences
issued by the Copyright Licensing Agency in the UK, or in accordance with the terms of licenses issued
by the appropriate reproduction rights organization outside the UK. Enquiries concerning reproduction
outside the terms stated here should be sent to IWA Publishing at the address printed above.

The publisher makes no representation, express or implied, with regard to the accuracy of the
information contained in this book and cannot accept any legal responsibility or liability for errors or
omissions that may be made.

Disclaimer
The information provided and the opinions given in this publication are not necessarily those of IWA
and IWA Publishing and should not be acted upon without independent consideration and professional
advice. IWA and IWA Publishing will not accept responsibility for any loss or damage suffered by any
person acting or refraining from acting upon any material contained in this publication.

British Library Cataloguing in Publication Data

A CIP catalogue record for this book is available from the British Library

Library of Congress Cataloguing in Publication Data

A catalogue record for this book is available from the Library of Congress

Reference:
Salinas Rodriguez SG, Villacorte LO (2024) Experimental Methods for Membrane Applications in
Desalination and Water Treatment, 1st edn IWA Publishing, London. doi: 10.2166/9781789062977

Cover design: Hans Emeis

Graphic design: Hans Emeis

ISBN 9781789062960 (Hardback)

ISBN 9781789062977 (eBook)
ISBN 9781789062984 (ePub)

This is an Open Access book distributed under the terms of the Creative Commons Attribution Licence
(CC BY-NC-ND 4.0), which permits copying and redistribution for non-commercial purposes with
no derivatives, provided the original work is properly cited (https:// creativecommons.org/licenses/
by-nc-nd/4.0/). This does not affect the rights licensed or assigned from any third party in this book.
 Foreword

Experimental Methods for Membrane Applications in Desalination and Water Treatment

Water

Few other substances are so abundant on our beautiful planet that they would be able to
cover it in a layer more than three kilometers thick. Seen from space, our planet is blue and
white from water.

Still, water scarcity is a grave and global issue, because water is often not in the right place, at
the right time, and in the right quality. This book is dedicated to the last issue, water quality.
More specifically, the focus is on experimental membrane processes in water treatment
(this, of course, you will have picked up from the book title). Membrane processes in water
treatment are literally as old as life itself, but still a vibrant experimental field, as will be clear
when you enjoy the book.

As a technology, membrane filtration is highly effective, proven to be able to mitigate the

increasingly global challenges of water scarcity and limited access to clean water. Depending
on membrane type, the filtration process can remove a wide range of water contaminants,
making it uniquely suitable for purifying unconventional but abundant water sources such as
seawater, highly polluted surface or groundwater, and various types of wastewater. As water
scarcity impacts billions of people globally, thousands of membrane-based purification
plants have been planned or installed in both developed and developing regions. This
means that plant engineers and operators who have process and analytical knowledge of
membrane technology are urgently needed. Researchers are also needed to further improve
the sustainability and economic feasibility of the technology.

Unfortunately, knowledge on membrane

processes is currently fragmented in various
academic publications, most of which are
not freely available to operators, engineers
and researchers, particularly in developing
countries. This book aims to address this
critical issue by bringing it all together in
a series of chapters written by some of the
foremost experts in the field.

The Grundfos Foundation is proud to

co-sponsor this book.

Poul Toft Frederiksen

Head of Programme, Research and Learning,
The Grundfos Foundation

V
VI
 Contributors

Chapter(s)

Aamer Ali, PhD, MSc, Assistant Professor 5

Center for Membrane Technology, Department of Chemistry and Bioscience,
Aalborg University, Denmark

Adam C. Hambly, PhD, Senior Researcher Water Technology & Processes 12

Dep. of Environmental and Resource Engineering, Technical University of Denmark, Denmark

Alberto Tiraferri, PhD, MSc, Full Professor of Applied Environmental Engineering 4

Department of Environment, Land and Infrastructure Engineering, Politecnico di Torino, Italy

Almotasembellah Abushaban, PhD, MSc, Assistant Professor in Water Desalination 1, 7, 15

The Applied Chemistry and Engineering Research Center of Excellence,
Mohammed VI Polytechnic University, Morocco

Barun Lal Karna, ME (Research), Assistant Chief Engineer, Sofitel Sydney Darling Harbour 11
Water Research Centre, School of Civil and Environmental Engineering,
University of New South Wales, Australia

Cejna Anna Quist-Jensen, PhD, MSc, Associate Professor of Membrane Technology 5

Center for Membrane Technology, Department of Chemistry and Bioscience,
Aalborg University, Denmark

Claus Hélix-Nielsen, PhD, MSc, Professor DTU Sustain 4

& President Danish Natural Sciences Academy
Head of Dep. of Environmental and Resource Engineering, Technical University of Denmark, Denmark

Francisco Javier García Picazo, MEng, Assistant Chemist 19

Environmental Chemistry Services, City of San Diego, United States of America

Guillem Gilabert-Oriol, PhD, MSc, Research and Development Leader 2, 3

DuPont Water Solutions, Spain, Adjunct Professor in the Universitat Rovira i Virgili, Spain

Gustavo A. Fimbres Weihs, PhD, Lead Research Fellow 19

School of Chemical and Biomolecular Engineering, The University of Sydney, Australia

Helen Rutlidge, PhD, BSc, Lecturer 11

School of Chemical Engineering, University of New South Wales, Australia

Helga Calix Ponce, MSc, Researcher 13

Denmark

Irena Petrinic, PhD, MSc, Associate Professor 4

University of Maribor, Slovenia

Jan Frauholz, MSc, Process Engineer 4

Aquaporin A/S, Denmark & RWTH Aachen, Germany

Javier Rodriguez Gómez, Laboratory supervisor 18

Genesys and PWT brands, H2O innovation, Spain

VII
Jia Xin Tan, BEng 19
Faculty of Chemical and Process Engineering Technology,
Universiti Malaysia Pahang Al-Sultan Abdullah, Malaysia

Johannes S. Vrouwenvelder, PhD, MSc, Professor of Environmental Science and Engineering, 17

Director of Water Desalination and Reuse Center (WDRC)
Biological & Environmental Science & Engineering Division (BESE), King Abdullah University
of Science and Technology (KAUST), Saudi Arabia
Delft University of Technology , Faculty of Applied Sciences, Department of Biotechnology ,
The Netherlands

Karima Bakkali, BSc, Research Engineer 15

Mohammed VI Polytechnic University, Morocco

Kathleen Foo, MSc, PhD Research Fellow 19

Faculty of Chemical and Process Engineering Technology,
Universiti Malaysia Pahang Al-Sultan Abdullah, Malaysia

Léonie Le Bouille, MSc, Research Fellow 15

IHE Delft Institute for Water Education / CIRSEE Suez / Delft University, The Netherlands / France

Loreen O. Villacorte, PhD, MSc, Lead Water Treatment Specialist 1, 13

Global Technology and Innovation, Grundfos Holding A/S, Denmark

Luca Fortunato, PhD, MSc, Research Scientist 17

Water Desalination and Reuse Center (WDRC), Biological & Environmental Science & Engineering
Division (BESE), King Abdullah University of Science and Technology (KAUST), Saudi Arabia

Lucia Ruiz Haddad, MSc, PhD Candidate 14

Environmental Science and Engineering Program, Biological and Environmental Science and
Engineering (BESE) Division, King Abdullah University of Science and Technology (KAUST), Thuwal,
Kingdom of Saudi Arabia

Maria Salud Camilleri-Rumbau, PhD, MSc, Researcher, R&D Project Manager 4

Technology Centre of Catalonia - Fundació Eurecat, Spain / Aquaporin A/S, Denmark

Mohamed Chaker Necibi, PhD, Associate Professor in Circular Economy 15

International Water Research Institute (IWRI), Mohammed VI Polytechnic University, Morocco

Mohamed Fauzi Haroon, PhD, Associate Director Analytical Science & Technology 14
Moderna, United States of America

Mohammad Mahdi A. Shirazi, PhD, MSc, Senior Postdoctoral Fellow 5

Center for Membrane Technology, Department of Chemistry and Bioscience,
Aalborg University, Denmark

Mohaned Sousi, PhD, MSc 16

Water Supply, Sanitation and Environmental Engineering Department,
IHE Delft Institute for Water Education, The Netherlands

Morten Lykkegaard Christensen, PhD, MSc, Associate Professor of Wastewater Treatment 2

and Membrane Technology Department of Chemistry and Bioscience,
Aalborg University, Denmark

VIII
Muhammad Ali, PhD, Martin Naughton Assistant Professor in Environmental Microbiology 14
Department of Civil, Structural & Environmental Engineering, Trinity College Dublin,
The University of Dublin, Ireland

Muhammad Nasir Mangal, PhD, MSc 10

Membrane specialist, Berghof Membranes, The Netherlands

Nuria Peña García, PhD, MSc, Research Director 9, 18

Director of Scientific Global Services for Genesys and PWT brands, H2O innovation, Spain

Pascal E. Saikaly, PhD, MSc, Professor of Environmental Science and Engineering, 14

Chair of Environmental Science and Engineering Program, Biological and Environmental Science
and Engineering (BESE) Division, King Abdullah University of Science and Technology (KAUST),
Thuwal, Kingdom of Saudi Arabia

Pierre Le-Clech, PhD, MSc, Associate Professor of Water and Wastewater treatment 11
School of Chemical Engineering, University of New South Wales, Australia

Poul Toft Frederiksen, PhD, MSc, Head of Programme Research and Learning F
The Grundfos Foundation, Denmark

Pouyan Mirzaei Vishkaei, MSc 1

Researcher, Water Supply, Sanitation and Environmental Engineering Department,
IHE Delft Institute for Water Education, The Netherlands

Rita Kay Henderson, PhD, MSc, Professor of Water Quality and Treatment 11
School of Chemical Engineering, University of New South Wales, Australia

Sergio G. Salinas-Rodriguez, PhD, MSc, Associate Professor of Water Supply Engineering 1, 7, 8, 15

Water Supply, Sanitation and Environmental Engineering Department,
IHE Delft Institute for Water Education, The Netherlands

Steven J. Duranceau, PhD, PE, Professor and Director 6

Environmental Systems Engineering Institute Department of Civil, Environmental & Construction
Engineering, College of Engineering and Computer Science,
University of Central Florida, United States of America

Urban J. Wünsch, PhD, Postdoctoral Researcher 12

National Institute of Aquatic Resources, Technical University of Denmark, Denmark

Vanida A. Salgado-Ismodes, MSc, PhD research fellow 7

Water Supply, Sanitation and Environmental Engineering Department,
IHE Delft Institute for Water Education, The Netherlands

Victor Augusto Yangali Quintanilla, PhD, MSc, Lead Water Treatment Specialist 4
Global Technology and Innovation, Grundfos Holding A/S, Denmark

Victoria Sanahuja-Embuena, PhD, MSc, Chemical Engineer, Scientist 4

Aquaporin A/S, Denmark

Wen Yew Lam, BEng 19

Faculty of Chemical and Process Engineering Technology,
Universiti Malaysia Pahang Al-Sultan Abdullah, Malaysia

IX
Weng Fung Twong, BEng 19
Faculty of Chemical and Process Engineering Technology,
Universiti Malaysia Pahang Al-Sultan Abdullah, Malaysia

Xuan Tung Nguyen, BSc, Project Manager 4

Aquaporin Asia, Singapore

Yie Kai Chong, BEng 19

Faculty of Chemical and Process Engineering Technology,
Universiti Malaysia Pahang Al-Sultan Abdullah, Malaysia

Yuli Ekowati, PhD, MSc, Postdoctoral Researcher 1, 13

Global Technology and Innovation, Grundfos Holding A/S, Denmark

Yong Yeow Liang, PhD, Senior Lecturer 19

Faculty of Chemical and Process Engineering Technology, Centre for Research in Advanced Fluid
and Processes, Universiti Malaysia Pahang Al-Sultan Abdullah, Kuantan, Pahang, Malaysia

X
About the editors

Sergio G. Salinas-Rodriguez is Associate Professor and

desalination and water treatment technology professional
at IHE Delft Institute for Water Education. He has a PhD
in Desalination and Water Treatment from the Technical
University of Delft, an MSc in Water Supply Engineering
from UNESCO-IHE Institute for Water Education, a Master’s
in Irrigation and Drainage and a BSc in Civil Engineering
from San Simon Major University. He also obtained the
University Teaching Qualification in the Netherlands.

He has over 75 publications in books, chapters, international

peer-reviewed journals and conference proceedings in the
areas of seawater and brackish water desalination, water
treatment, water reuse, and natural organic matter characterization.

Sergio is involved in teaching and curriculum development of the MSc Programme in Water
and Sustainable Development at IHE Delft. His projects comprise capacity building, research
and innovation (e.g., EU-MEDINA, EU-MIDES, EU-MAR2PROTECT). He has mentored more
than 50 MSc students, co-promoted 4 PhD students, and currently supervises 2 PhD students.
Sergio lectures and coordinates several courses on Desalination and membrane technology.

Loreen Ople Villacorte is Lead Water Treatment Specialist at Global Technology and
Innovation in Grundfos Denmark. He has broad experience in research, conceptualization,
development and validation of water treatment technologies including applications of
traditional and emerging membrane technologies.

For the last 18 years, he has held various roles in the academia
and the industry across three countries (Philippines,
Netherlands and Denmark), primarily driving research and
technology development projects to tackle water challenges
in drinking water production and transport, wastewater
treatment or reuse, oil-water separation and industrial
cooling systems. Most of these projects were implemented
through cross-functional collaborations and involves
understanding the physics, biology and chemistry of water
to enable development of effective treatment solutions. He
has published >25 scientific articles and filed numerous
patents in water treatment and desalination applications.

He is a civil engineer with a master’s degree in water supply engineering at IHE-Delft and a
doctoral degree in desalination and water treatment from the Technical University of Delft,
IHE-Delft and Wetsus.

XI
XII
 Contents

Foreword V
Contributors VII
About the editors XI

Chapter 1
Feedwater Quality Guidelines and Assessment Methods
for Membrane-based Desalination 1
1.1 Introduction 1
1.2 Particulate fouling potential 7
1.3 Inorganic fouling and scaling potential 9
1.4 Organic fouling potential 10
1.4.1 Organic carbon 11
1.4.2 UV absorbance and fluorescence 11
1.4.3 LC-OCD 11
1.4.4 TEP 12
1.4.5 Oil and grease 12
1.5 Biofouling potential 13
1.5.1 Bacterial growth potential 13
1.5.2 Assimilable organic carbon 14
1.5.3 Biodegradable dissolved organic carbon 15
1.5.4 Phosphate 15
1.6 Outlook and opportunities 16
1.7 Abbreviations and symbols 17
1.8 References 18

Part 1
Membrane processes 25

Chapter 2
Microfiltration and ultrafiltration 27
2.1 Introduction 27
2.1.1 Advantages of ultrafiltration compared to conventional treatment 28
2.2 Design and optimize membrane processes 29
2.3 Objective of the filtration process 30
2.4 Membrane types 32
2.5 Basic equations 33
2.6 Normalization 35
2.7 Membrane fouling 36
2.8 Sustainable flux 38
2.9 Membrane design and module 40
2.10 Pretreatment 41
2.11 Cleanings 41
2.11.1 Optimization of hydraulic cleaning 42

XIII
2.12 Membrane cascades 43
2.13 Summary 44
2.14 References 45

Chapter 3
Reverse Osmosis and Nanofiltration 47
3.1 The rise of reverse osmosis 47
3.2 Sustainaiblity of reverse osmosis 48
3.3 Understanding the osmosis process 49
3.3.1 Semi-permeable membranes 49
3.3.2 The reverse osmosis process 50
3.4 Equations 52
3.4.1 Fundamental equations 53
3.4.1.1 Osmotic pressure 53
3.4.1.2 Water flux 54
3.4.1.3 Salt transport 55
3.4.1.4 The difference between convective and concentration driven flows 55
3.4.2 System equations 56
3.4.3 Factors affecting membrane performance 58
3.4.3.1 Feed pressure 58
3.4.3.2 Feed concentration 58
3.4.3.3 Feed temperature 59
3.4.3.4 Concentration polarization 59
3.5 Reverse osmosis membranes 59
3.5.1 The significance of desalination 62
3.6 Performance monitoring 62
3.7 Normalization 63
3.7.1 Why normalization matters 64
3.7.2 Equations 64
3.7.2.1 Normalized permeate flow 64
3.7.2.2 Normalized salt rejection 66
3.7.2.3 Normalized pressure drop 66
3.8 Fouling 66
3.8.1 Biofouling 66
3.8.2 Organic fouling 67
3.8.3 Particulate fouling 67
3.8.4 Scaling 68
3.8.5 Integrity failure 68
3.9 References 70

Chapter 4
Forward Osmosis 71
4.1 Introduction: principles of forward osmosis 71
4.2 Materials and experimental set-up 73
4.2.1 Membrane configurations 73
4.2.2 Experimental modes 73
4.2.3 Draw solutions: properties, regeneration, types and selection criteria 76

XIV
4.3 Experimental methods 80
4.3.1 Typical parameters and phenomena 80
4.3.2 FO process design constraints and considerations 82
4.3.3 Best practices 86
4.4 Data analysis: Basic FO process design 87
4.4.1 FO fundamental equations 87
4.4.2 FO module mass balance 89
4.4.3 FO design considerations 91
4.5 Application examples 92
4.6 Outlook 95
4.7 References 96

Chapter 5
Membrane Distillation 97
5.1 Introduction 97
5.2 Materials, experimental set-up 100
5.2.1 MD membranes 100
5.2.1.1 Membrane properties 100
5.2.1.2 Membrane materials 101
5.2.2 Experimental set-up 104
5.2.2.1 MD configurations 104
5.2.3 Process 106
5.2.3.1 MD system 106
5.2.3.2 Operating parameters 107
5.2.4 MD modules 107
5.3 Methods 112
5.3.1 Process measurements and calculations 112
5.3.1.1 Permeate flux 112
5.3.1.2 Solute rejection 112
5.3.1.3 Logarithmic temperature difference 112
5.3.2 Membrane characterization 112
5.3.2.1 Physical and morphology properties 112
5.3.2.2 Chemical properties 120
5.3.2.3 Thermal properties 121
5.4 Applications and examples 122
5.5 Outlook 123
5.6 References 125

Part 2
Particulate fouling 137

Chapter 6
Silt Density Index 139
6.0 Abstract 139
6.1 Development of the fouling index 140
6.2 Silt as a component of membrane fouling 140
6.3 Standardizaton of the silt density index 141

XV
6.4 Methods and procedures 142
6.5 Limitations of the SDI 146
6.6 Alternatives to the SDI 149
6.7 Summary 151
6.8 References 152

Chapter 7
Modified Fouling Index (MFI-0.45) 155
7.1 Introduction 155
7.2 Theory particulate fouling 156
7.3 Measuring MFI-0.45 159
7.3.1 Filtration set-up and materials 159
7.3.1.1 Membrane filters 160
7.3.1.2 Filter holder 160
7.3.1.3 Feedwater reservoir 161
7.3.1.4 Electronic mass balance 161
7.3.1.5 Software for data acquisition 161
7.3.1.6 Pressure regulator and gauge 162
7.3.1.7 Pressure transducer 162
7.3.1.8 Non-plugging water 162
7.3.2 MFI-0.45 testing procedure 163
7.3.3 MFI-0.45 calculation procedure 163
7.4 Membrane properties of commercial membranes 165
7.5 Effect of filter material on MFI-0.45 values 166
7.5.1 Effect of membrane support holder in MFI-0.45 167
7.6 Application: water quality monitoring of North Sea water 168
7.7 Monitoring of MFI-0.45 in a full-scale desalination plant 169
7.8 References 171

Chapter 8
Modified Fouling Index Ultrafiltration (MFI-UF) Constant Flux 173
8.1 Introduction 173
8.2 Theory Particulate fouling 176
8.2.1 Deposition factor 180
8.2.2 The particulate fouling prediction model 181
8.3 Measuring MFI-UF constant flux 181
8.3.1 Filtration set-up and materials 181
8.3.1.1 Membrane filters 182
8.3.1.2 Constant flow pump 183
8.3.1.3 Pressure transducer 184
8.3.1.4 Membrane filter holder 185
8.3.1.5 Syringe 185
8.3.1.6 Ultra-pure water 185
8.3.1.7 Tubing 186
8.3.1.8 Software 186
8.3.2 Membrane cleaning and conditioning 186
8.3.3 MFI-UF testing procedure 187

XVI
 8.3.3.1 Selection of filtration flux rate 187
8.3.4 Calculation procedure 188
8.3.4.1 Example of membrane resistance calculation of UPW 189
8.3.4.2 Example of MFI-UF calculation 190
8.3.5 Reproducibility 191
8.3.6 Blank and limit of detection 191
8.3.7 Sample storage 192
8.3.8 Concentration of particles 193
8.3.9 Membrane material 194
8.4 Variables and applications of the MFI-UF 195
8.4.1 Plant profiling and water quality monitoring 195
8.4.2 Flux rate 196
8.4.3 Predicting rate of fouling of seawater RO systems 197
8.4.4 Comparing fouling indices 198
8.5 References 200

Part 3
Inorganic fouling and scaling 205

Chapter 9
Inorganic Fouling Characterization Tools and Mitigation 207
9.1 Introduction 207
9.2 Main components of inorganic fouling 210
9.2.1 Colloidal matter/particulate 210
9.2.2 Metals 214
9.2.3 Scaling 218
9.2.4 Other components 221
9.3 Methods for inorganic fouling identification 222
9.4 Methods for inorganic fouling removal 225
9.5 References 228

Chapter 10
Assessing Scaling Potential with Induction Time and a
Once-through Laboratory Scale RO System 229
10.1 Introduction 229
10.2 Induction time measurements 231
10.2.1 Experimental setup 232
10.2.1.1 Glass reactor 232
10.2.1.2 Stirrer device 233
10.2.1.3 pH meter 233
10.2.1.4 Peristaltic pump 233
10.2.1.5 Thermostat 233
10.2.2 Experimental procedure 233
10.2.2.1 Preparation of artificial brackish water 233
10.2.2.2 Induction time measurement 234
10.2.3 Calculation of induction time 234
10.2.4 Cleaning of the reactor 235

XVII
 10.2.5 Example of application of induction time 235
10.3 Once through lab-scale RO system 238
10.3.1 Experimental set-up 238
10.3.2 Experimental protocol 240
10.3.3 Example of application 240
10.4 Outlook and final comments 243
10.5 References 245

Part 4
Organic fouling 247

Chapter 11
Practical Considerations of Using LC-OCD for Organic Matter 249
Analysis in Seawater
11.1 Introduction 249
11.2 LC-OCD analysis 252
11.2.1 Instrumentation and chromatogram integration 252
11.2.2 Effect of salinity on organic characterization and calibration 254
11.2.3 Level of detection 257
11.2.4 Reproducibility of LC-OCD 257
11.2.5 Characterisation of organic mixtures 259
11.2.6 Applications 259
11.2.6.1 OM composition in seawater 260
11.2.6.2 Fouling behaviour of organic matter 261
11.2.6.3 Effectiveness of pretreatment methods 261
11.3 Conclusions 261
11.4 References 262

Chapter 12
Fluorescence Excitation Emission Matrix (EEM) Spectroscopy 265
12.1 Introduction 265
12.2 Sampling & storage 267
12.3 Benchtop instrumentation 268
12.4 Quality assurance 270
12.5 Interferences 271
12.6 Data processing 272
12.7 Data analysis 273
12.7.1 PARAFAC 275
12.8 Application in membrane systems 275
12.9 References 281

Chapter 13
Transparent Exopolymer Particles 287
13.1 Introduction 287
13.2 Quantification methods 290
13.2.1 Alcian blue dye preparation 292
13.2.2 TEP0.4μm measurement 293

XVIII
 13.2.3 TEP10kDa measurement 297
13.2.4 Method calibration 300
13.2.4.1 Xanthan gum standard preparation 300
13.2.4.2 TEP0.4μm calibration 1 301
13.2.4.3 TEP0.4µm calibration 2 301
13.2.4.4 TEP10kDa calibration 302
13.2.5 Other considerations 302
13.2.5.1 Limit of detection 302
13.2.5.2 Impact of storage on TEP concentration 304
13.2.6 Application and interpretation 304
13.3 Summary and outlook 309
13.4 References 310

Part 5
Biological fouling 313

Chapter 14
Genomics Tools to Study Membrane-Based Systems 315
14.1 Introduction 315
14.2 Experimental design and sample preparation 318
14.2.1 Experimental design in a metagenomics 318
14.2.2 Sample collection and preservation 319
14.2.3 DNA extraction 319
14.2.4 Library preparation 320
14.2.5 Sequencing platforms 320
14.3 Bioinformatics analysis 321
14.3.1 Data pre-treatment 321
14.3.2 Amplicon-based approach 321
14.3.3 Metagenomics, read-based approach 322
14.3.4 Metagenomics, assembly-based approach 322
14.3.5 Metagenome-assembled genome (MAG) binning 322
14.3.6 Supervised and unsupervised binning 323
14.3.7 Functional annotation 323
14.3.8 Genome-resolved metatranscriptomics 323
14.4 Data sharing and storage 324
14.5 Bioinformatics analysis workflow examples 324
14.5.1 Amplicon sequences processing workflow 324
14.5.2 Genome-resolved metagenomics 325
14.5.3 Genome-resolved metatranscriptomics 327
14.6 Applications of genomics in membrane filtration research 329
14.7 Outlook 330
14.8 Data availability 331
14.9 References 332

Chapter 15
Biofouling Potentia Measurement using Bacterial Growth Potential (BGP) 337
15.1 Introduction 337

XIX
15.2 Materials 338
15.2.1 Laboratory equipment 338
15.2.2 Chemicals 339
15.2.3 Instrumental equipment 339
15.3 Methods and experimental procedure 340
15.3.1 Sample collection and storage 340
15.3.2 Cleaning glassware 341
15.3.3 Preparation of artificial seawater 341
15.3.4 Intact cell count by flow cytometry 342
15.3.5 Measurement of bacterial growth potential 342
15.3.6 Bacterial yield and calibration line 343
15.4 Applications 345
15.4.1 Example A: BGP monitoring of an SWRO pre-treatment 345
15.4.2 Example B: BGP in the intake and SWRO feed water 345
15.5 Data discussion and interpretation 346
15.6 ATP measurement 347
15.6.1 Introduction 347
15.6.2 Material and methods 348
15.7 References 353

Chapter 16
Assessing Biological Stability of Ultra-low Nutrient Water by Measuring Bacterial
Growth Potential 355
16.1 Introduction 355
16.2 Materials and experimental set-up 357
16.2.1 Equipment 357
16.2.2 Materials and methods 361
16.2.3 Method 361
16.3 Examples of application 364
16.4 Additional considerations 370
16.5 References 372

Chapter 17
Optical Coherence Tomography (OCT) as a Tool for (Bio)-fouling Assessment
in Desalination Systems 375
17.1 Introduction 375
17.2 Materials, experimental set-up 377
17.3 Methods 377
17.3.1 Imaging with optical coherence tomography 377
17.4 Data Analysis 378
17.4.1 Biovolume calculation 378
17.4.2 Image processing 381
17.5 Data discussion and interpretation 382
17.5.1 Biomass quantification 382
17.5.2 Membrane performance 383
17.6 Applications, examples 384
17.6.1 Biomass distribution 384

XX
 17.6.2 Biomass and performance decline 385
17.6.3 Biomass thickness map 386
17.7 Additional considerations 387
17.7.1 OCT image analysis 387
17.7.2 Biomass accumulation and membrane performance 388
17.7.3 Biomass location in the flow channel 389
17.7.4 Use of OCT in biofouling studies 390
17.7.5 Mapping the biofouling 390
17.8 Summary 391
17.9 References 392

Part 6
General applications 397

Chapter 18
Membrane Autopsy 399
18.1 Introduction 399
18.2 Materials, experimental set-up 401
18.3 Membrane autopsy protocol 402
18.4 Methods 403
18.4.1 Visual inspection 403
18.4.2 Analytical methods for foulant and damage identification 408
18.4.3 Membrane performance 416
18.4.4 Cleaning tests 418
18.5 References 420

Chapter 19
CFD as a Tool for Modelling Membrane Systems 421
19.1 Introduction 421
19.1.1 What is not modelled 423
19.1.2 How modelling can assist membrane systems 423
19.2 Methods 424
19.2.1 Geometry 425
19.2.2 Flow types 427
19.2.2.1 1D, 2D and 3D 427
19.2.2.2 Laminar, transient, turbulent 427
19.2.2.3 Single phase 428
19.2.2.4 Multiphase 428
19.2.3 Boundary conditions 431
19.2.3.1 Steady-state and transient-state 432
19.2.4 Initial conditions 432
19.2.5 Meshing and algorithms 433
19.2.6 Convergence 434
19.3 Data Analysis 434
19.3.1 Verification 434
19.3.2 Validation 436
19.4 Data discussion and interpretation 437

XXI
 19.4.1.1 Data processing and assessment 437
19.5 Applications, examples 438
19.5.1 Flow stability 438
19.5.1.1 Laminar steady 438
19.5.1.2 Laminar unsteady – oscillating vs. vortex shedding 438
19.5.1.3 Quasiperiodic flow 439
19.5.1.4 Turbulent flow 441
19.5.2 Mass transfer and vortex shedding 441
19.5.3 Spacer design 442
19.5.3.1 Two-dimensional feed spacer 442
19.5.3.2 Three-dimensional feed spacer 442
19.5.4 Flow perturbation 443
19.5.4.1 Electro-osmosis 444
19.5.4.2 Modelling electro-osmosis in CFD 445
19.5.4.3 Significant learnings of EOF slip velocity in CFD studies 446
19.5.4.4 Oscillating flow 447
19.5.4.5 Vibrations 449
19.5.5 Fouling modelling 449
19.5.5.1 Particulate fouling 449
19.5.5.2 Tracer test 452
19.5.5.3 Biofouling 452
19.6 Additional considerations 455
19.6.1 Multi-scale modelling 455
19.6.1.1 Techno-economics 456
19.7 Outlook 456
19.8 References 457

XXII
 doi: 10.2166/9781789062977_0001

Chapter 1

Feedwater Quality Guidelines

and Assessment Methods for
Membrane-based Desalination
Loreen O. Villacorte, Grundfos, Denmark

Yuli Ekowati, Grundfos, Denmark

Almotasembellah Abushaban, UM6P, Morocco

Pouyan Mirzaei Vishkaei, IHE Delft, The Netherlands

Sergio G. Salinas-Rodriguez, IHE Delft, The Netherlands

The learning objectives of this chapter are the following:

• To review the existing feedwater quality guidelines for membrane-based

desalination

• To present and discuss the existing and proposed methods for assessing fouling and
scaling potential of feedwater.

1.1 INTRODUCTION

Amidst the global problem of dwindling freshwater water resources, desalination of

unconventional but abundant water resources such as seawater and brackish water has
grown rapidly over the last three decades. From a global operational capacity of ~7.5 million
m3/day in 1990 to ~115 million m3/day in 2023, water desalination technologies have
been the leading solution to address the growing municipal, agricultural and industrial
demand for clean freshwater (Figure 1a and 1b). Furthermore, desalination technologies
are generally applied for triple barrier wastewater reuse applications, which currently has a
global installed capacity of >60 million m3/year (Birch et al., 2023).

© 2024 The Authors. This is an Open Access book chapter distributed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0),
(https://creativecommons.org/licenses/by-nc-nd/4.0/).
The chapter is from the book Experimental Methods for Membrane Applications in Desalination and
Water Treatment, Sergio G. Salinas-Rodriguez, Loreen O. Villacorte (Eds).
Experimental Methods for Membrane Applications

Seawater is the main water source for desalination globally, with the exception of North
America, where the majority of applications is based on brackish water desalination. Japan,
South Korea, Taiwan, and China desalinate seawater, brackish water, and wastewater
effluent at relatively similar rates.

120
Total
Membrame based (RO) 100
Thermal (MSF + MED)
80

60

40

20

Capacity (million m3/d) 0

1970 1980 1980 2000 2010 2020
year

Figure 1a The growth of global desalination application in terms of online production capacity since
1970 (top) and current online seawater desalination by technology, region and end-user.
Produced with information from (DesalData, 2023). MSF = Multi-stage flash distillation,
MED = Multi-effect distillation, RO = Reverse osmosis

In terms of technology, membrane-based desalination using reverse osmosis (RO) dominates

the application (~74% of global capacity). This is mainly driven by the significantly lower
investment cost and energy requirements today are lower than thermal processes (e.g., MSF,
MED). A large majority (>75%) of the desalinated water are used for supplying drinking
water supply while about 20% are used in industries. Most of the Middle East countries rely
on desalination for municipal use, while countries such as China, India, South Korea, Brazil,
Taiwan, Chile, Indonesia use desalination to satisfy industrial demand.

2
 Chapter 1

Global online desalination 2023 (~115 Mm3/d)

Wastewater 9% Brackish water 19%

Seawater 60% Pure water 4%

Fresh water 8%

North America 9% Southern Asia 4%

Sub-Saharan Africa 2%
Western Europe 7%

Middle East / East Asia / Pacific 18%

North Africa 52%
Eastern Europe / Central Asia 42%

Latin America / Caribbean 6%

Irrigation 4% Drinking water 46%

Industry 50%

Seawater desalination 2023 (~70 Mm3/d)

MED 9% MSF 21%

RO 70%

North America 1% Southern Asia 3%

Sub-Saharan Africa 1%
Western Europe 7%
East Asia / Pacific 13%

Middle East / Eastern Europe / Central Asia 2%

North Africa 52% Latin America / Caribbean 4%

Industry (<10ppm) 21%

Irrigation (<1000ppm) 1%
Drinking water 78%
(10 ppm - 1000 ppm)

Figure 1b Current online global desalination (top 3) and online seawater desalination (bottom 3)
by technology, region and end-user. Produced with information from (DesalData, 2023)

3
Experimental Methods for Membrane Applications

Western Europe
Eastern Europe
7.7 Mm3/d
North America 2.3 Mm3/d
9.5 Mm3/d China
10.3 Mm3/d Japan, Korea,
Taiwan 2.0 Mm3/d

Caribbean Rest East Asia / Pacific

1.4 Mm3/d Middle East 2.0 Mm3/d
North Africa 46.7 Mm3/d
India
7.2 Mm3/d
3.4 Mm3/d
Singapore
2.1 Mm3/d
Latin America Sub-Saharan Africa
Seawater 4.5 Mm3/d 1.8 Mm3/d
Brackish water Australia
2.9 Mm3/d
Wastewater

Figure 2 Feedwater sources and (online, in construction) production capacities of desalination

plants in different geographic locations in 2023 (updated from Salinas Rodriguez and
Schippers (2021) with information from DesalData (2023)).

The price per cubic metre of desalinated water has reduced significantly over the years due
to more efficient membrane production, implementation of energy recovery devices, cost
of engineering, etc, and a more competitive market. The specific energy consumption has
already been reduced by at least 50% over the last 20 years and the overall carbon footprint
of desalination could be reduced down further by switching to renewable energy sources
(Birch et al., 2023). On the downside, membrane fouling and scaling are the main ‘Achilles
heel’ for the sustainable application of RO (Voutchkov, 2010, Salinas Rodriguez, 2021).

Fouling and scaling in membranes can lead to a variety of problems, such as the need for
(frequent) chemical cleaning, reduction of production capacity, higher energy consumption,
decrease in produced water quality, that it makes RO production facilities less reliable, and
require more frequent membrane replacement (Dhakal et al., 2020, Salinas Rodriguez et
al., 2021b). Fouling and scaling are broadly categorized into i) particulate/colloidal fouling
due to suspended and colloidal matter, ii) inorganic fouling due to iron and manganese, iii)
organic fouling due to organic compounds e.g., polymers, iv) biofouling due to growth of
bacteria, and v) scaling due to deposition of sparingly soluble compounds.

During RO operation, membrane fouling and scaling may manifest in three ways, namely:
i) increasing the differential pressure across the spacer in spiral wound elements due to
‘clogging’, resulting in potential membrane damage (such as telescoping, channelling, or
squeezing); ii) increasing membrane resistance (or decreasing the normalized permeability)
due to deposition and/or adsorption of materials on the membrane surface, resulting in
higher required feed pressure to maintain capacity; and iii) increasing in normalized salt
passage due to concentration polarization in the fouling layer, resulting in higher salinity in
the product water.

4
 Chapter 1

Particulate and colloidal fouling are mostly well controlled by the pre-treatment systems
(mostly media filtration or membrane filtration), but the occurrence of organic fouling and
biofouling is still a major issue in RO membranes, and is the main reason for the need for
frequent cleaning of the reverse osmosis membranes (Peña et al., 2022).

To minimize the occurrence of membrane fouling/scaling in RO, pre-treatment of the

feedwater is essential. Additionally, methods and tools can help significantly, by monitoring
the performance of the pre-treatment with regards to fouling/scaling control and process
optimization. Pre-treatment can take place in the form of media filters with or without
coagulation, membrane filtration with or without inline coagulation (e.g., ultrafiltration),
and dissolved air floatation in combination with the previous mentioned two options.

Along with the increase in the number of desalination plants (>22,800 plants in 2023),
the capacity of newly installed plants has also increased significantly over time. A growing
preference for extra-large (XL) plants (capacity >50,000 m3/d) has been reported in recent
years (Birch et al. (2023); Kurihara and Ito (2020). More XL seawater RO (SWRO) plants are
expected in the future. This means reliable pre-treatment systems and monitoring tools will
be essential for these XL plants, as Cleaning-in-Place (CIP) of membrane modules more than
once per year is rather challenging. The design and operational settings of such pre-treatment
systems will depend on the water quality and their temporal variations of the water source
alongside the feedwater quality guidelines provided by the membrane supplier.

For a long time, the silt density index (SDI) has served as a sum ‘king/ultimate’ parameter
for assessing RO feed water. DuPont (2020) for the first time introduced the MFI-0.45
in their RO feedwater guidelines. This is a major step forward due to the limitations of
the SDI in assessing fouling in RO (Schippers et al., 2014). In addition, the inclusion of
parameters like AOC and BFR bring relevance for the monitoring of the biofouling potential
of RO feedwater as several types of fouling take place simultaneously. Table 1 presents the
recommended guideline values for RO feedwater by RO manufacturers and literature. The
majority of RO membrane manufacturers are in agreement with the recommendations by
DuPont although they main guideline is the SDI value less than 4-5 and preferable less than
3 for RO feedwater.

5
Experimental Methods for Membrane Applications

Table 1 Parameters and recommended guideline values for RO feed water

Parameter Unit DuPont (2023) Other sources Standard Methods
(a) Particulate fouling indicators
SDI15 %/min <5 (target <3) <5 (target <3) (Wilf and Klinko, (ASTM D4189 - 07)
2016)
< 3 (Badruzzaman et al., 2019)
< 4 (Voutchkov, 2010)
MFI-0.45 s/L2 4 (target <1) (ASTM D8002 - 15)
MFI-UF s/L2 - <490 at 15 lmh (safe MFI*)
(Salinas Rodríguez, 2011)
Turbidity NTU <1 < 0.5 (Badruzzaman et al., 2019) (ASTM D1889-00)
< 0.1 (Voutchkov, 2010)
(b) Organic fouling indicators
Oil and grease mg/L 0.1 < 0.1 (Badruzzaman et al., 2019) (ASTM D7575-11)
< 0.02 (Voutchkov, 2010)
TOC mg-C/L 3 < 2 (Badruzzaman et al., 2019) (ASTM D2579-93e1)
< 2 (target <0.5)
(Voutchkov, 2010)
SUVA L/mg-m < 4 (USEPA, 2005)
COD mg/L 10 (ASTM D1252-
06(2020))
(c) Biological fouling indicators
AOC µg/L Ac-C 10 (target <5) <10 µg-C acetate/L (threshold for (NEN 6271:1995 nl)
biofouling in freshwater) (van der
Kooij et al., 1982)
BGP µg-C/L - - <70 (Abushaban, 2019)
BFR pg-ATP/ 5 (target <1) < 1 (Vrouwenvelder and van der
cm2 Kooij, 2001)
PO4-P µg/L 0.3 µg P/L (Vrouwenvelder et al.,
2010)
(d) Inorganic fouling and scaling indicators
Ferrous iron mg/L 4 < 2 (Badruzzaman et al., 2019) (ASTM D1068-15)
< 2 (Voutchkov, 2010)
Ferric iron mg/L 0.05 < 0.1 (Badruzzaman et al., 2019) (ASTM D1068-15)
<0.05 (Voutchkov, 2010)
Manganese mg/L 0.05 0.05 (Badruzzaman et al., 2019) (ASTM D858-17)
0.02 (Voutchkov, 2010)
Aluminium mg/L 0.05 (ASTM D857-17)
Silica mg/L 20 (Badruzzaman et al., 2019) (ASTM D859-
16(2021)e1)
pH - 4-11 (Voutchkov, 2010) (ASTM D1293-12)
LSI - Concentrate (ASTM D3739-19)
(freshwater) LSI < 0 (if no
antiscalant is
added)

6
 Chapter 1

Parameter Unit DuPont (2023) Other sources Standard Methods

S&DSI - Concentrate (ASTM D4582-
(seawater) S&DSI < 0 (if 91(2001))
no antiscalant is
added)
(e) Membrane material limits

Temperature °C < 35 (Voutchkov, 2010)

Free chlorine mg/L <0.1 < 0.1 (Badruzzaman et al., 2019) (ASTM D1253-
< 0.01 (Voutchkov, 2010) 14(2021)e1)
ORP mV <175-200 (ASTM D1498-
14(2022)e11)
Parameter Unit DuPont (2023) Other sources Standard Methods
* Safe MFI is a value for RO feedwater that will yield a 1 bar pressure increase in a 6 months period.

In addition to the established feedwater quality parameters in Table 1, tens of thousands of

scientific articles and patents were published over the past 30 years describing or applying
new assessment tools/indices for evaluating the fouling/scaling potential of RO feedwater
as well as to characterize the impact of specific feedwater components to RO operation
(Figure 3). Some of these tools were also applied to optimize the design and operation of
RO pre-treatment system, including MF/UF processes (see Chapter 2). The succeeding
sections review the advantages as well as the challenges of applying these assessment tools
in membrane-based desalination systems.

10 100

Cumulative publications (x1,000)

Yearly publications
9 90
Publications (year x1,000)

8 Cummulative publications 80
7 70
6 60
5 50
4 40
3 30
2 20
1 10
0 0
19 0
19 1
19 2
19 3
19 4
19 5
19 6
19 7
19 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
23
9
9
9
9
9
9
9
9
9
9
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
2
2
2
19

Figure 3 Number of scientific and patent publications related to fouling and scaling assessment
in reverse osmosis process from 1990 and 2023 (November). Data generated through
Google Scholar using the search string: “(fouling OR scaling) AND (characterization OR
assessment OR potential OR indicator OR index) AND (reverse osmosis)”.

1.2 PARTICULATE FOULING POTENTIAL

Fouling indices to measure the particulate fouling potential of RO feedwater have been
in development since the 1960’s (Figure 4). The oldest and most widely used index, the
silt density index (SDI) has been standardised by ASTM D4189 - 14 (2014), is applied
worldwide as it is simple to perform and with low-cost consumables (see Chapter 6).

7
Experimental Methods for Membrane Applications

However, increasingly the value of this test to predict the rate of fouling in RO systems due to
particle deposition is being questioned. The limitations of the SDI test are well documented
(Schippers and Verdouw, 1980, Nahrstedt and Camargo Schmale, 2008, Alhadidi et al.,
2011a, Alhadidi et al., 2011b, Rachman et al., 2013, Salinas Rodriguez et al., 2019) and
include no correction for feedwater temperature variation (higher SDI values at higher
temperatures); the result is heavily dependent on the permeability of the test membrane
filter; not applicable for testing high fouling feed water e.g., raw water – ASTM recommends
that turbidity should be < 1 NTU; not applicable for testing UF system permeate, which
is increasingly being used in desalination pre-treatment; no linear relation with colloidal/
suspended matter; fouling potential of particles smaller than 0.45 μm are not taken into
account; it is not based on any filtration mechanism.

MFI-UF CF
with flat MFI-0.45
MFI-UF membranes ASTM
constant pressure (10, 30, 50, 100) D8002-15

MFI-UF Flux effect, Improved fouling

constant Flux deposition prediction
with HF factor, (5 kDa membrane,
MFI-0.45 MFI-0.05 membrane safe MFI porosity effect)

1970 1980 1990 2000 2010 2020

FI SDI ASTM SDI ASTM SDI ASTM SDI ASTM

(SDI) D4189-95 D4189-02 D4189-07 D4189-14

Figure 4 Historical development of fouling indices for particulate fouling assessment (adapted from
Salinas Rodriguez et al. (2021a)). Legend: FI = fouling index, SDI = silt density index, MFI
= modified fouling index, MFI-UF = modified fouling index ultrafiltration, CF = constant
flux, ASTM = American society for testing and materials.

The modified fouling index (MFI-0.45) test, also standardized by ASTM D8002 - 15 (2015),
uses the same materials and equipment as the SDI test. It is based in the cake filtration
fouling mechanism. The obtained MFI value is corrected for temperature and pressure and
shows a linear relation with colloidal/suspended matter concentration (see Chapter 7). The
predicted rate of fouling turns out to be very low at a level of MFI = 1 s/L2, which is in the
range of SDI 1 to 3.

Based on the low sensitivity of MFI-0.45, the MFI-UF test with ultra-filtration membranes
was developed to capture these smaller particles (see Chapter 8). The strong dependence of
MFI on flux, means that to be able to predict accurately the potential of particulate fouling
in RO systems, the MFI should preferably be measured at a flux similar to a RO system
(15- 20 L/m2/h) or extrapolated from higher fluxes. A theoretical ‘safe MFI’ was proposed
assuming e.g., an allowable increase in NDP of 1 bar in 6 months (Salinas Rodríguez, 2011,
Salinas Rodriguez et al., 2019). The safe MFI calculated for a deposition factor Ω = 1 (worst
case) at a flux of 15 L/m2/h (average flux in RO), has been reported about 490 s/L2. And
could be used as a threshold value for assessing RO feed water quality. Good correlation with
RO membrane fouling development was observed when applying the MFI-UF prediction
model (Abunada et al., 2023).

8
 Chapter 1

1.3 INORGANIC FOULING AND SCALING POTENTIAL

Inorganic fouling occurs in RO processes due to the deposition of insoluble or sparingly

soluble inorganic compounds from the feedwater to the RO membrane. In many
literatures, inorganic fouling also includes scaling, however, deposition of metal ions, such
as iron, manganese and aluminium is also called as inorganic fouling. Iron, manganese and
aluminium ions are present naturally in surface water and groundwater. Fouling due to these
ions can be reduced by pre-treatment, such as aeration, oxidation, and filtration. Membrane
manufacturers provide guidelines for feedwater quality compatible with the membrane
which typically includes limits on iron, manganese, and aluminium concentrations
(Table 1). Analytical methods to quantify critical inorganic components in the feedwater
or to identify accumulation on the membrane surface are well established, which usually
includes ICP and SEM-EDX. There is no single analytical method than can completely
identify all inorganic foulants so it is important to get familiarized with these methods (see
full list of tools in Chapter 9) when studying inorganic fouling.

Scaling on RO membranes is specifically caused by precipitation and accumulation of

sparingly soluble salts usually as a consequence of up-concentrating these salts in the RO
process, eventually exceeding their solubility limit. The type of scaling depends on the ion
composition, temperature and pH of the feedwater. Scaling issues typically observed in RO
processes are calcium carbonate, calcium sulphate, calcium phosphate, barium sulphate,
and silica/metal silicates. Scaling can be avoided by limiting the RO recovery, acid dosing,
and antiscalant dosing. The latter is the most preferable approach because it does not
compromise the RO production and it is effective for different types of scaling with only
low dose requirement. Pretreatment of feedwater by ion exchange or lime softening can also
be applied to control scaling.

The scaling potential of feedwater in the RO system can be assessed based on scaling indices.
The commonly used indices for RO application are saturation index (SI) and supersaturation
ratio (Sr), Langelier saturation index (LSI), and Stiff-Davis stability index (S&DSI). SI and Sr
are applicable for all types of scaling species while LSI and S&DSI are specifically used for
calcium carbonate scaling. A positive value of the indices generally indicates supersaturation
condition where precipitation can occur. Calculation of saturation indices are available in
standardized ASTM methods (ASTM D4692-01, D3739-19, D4582-23) and in literatures
(e.g., Mangal et al. (2021)). In practice, the scaling potential of RO feedwater are usually
determined using the design software developed by antiscalant chemical suppliers and
membrane manufacturers, such as WAVE (DuPont), IMSDesign (Hydranautics) and
MembraneMaster MM5 (Genesys). A geochemical modelling program, PHREEQC, can also
be used to calculate SI.

An emerging parameter known as induction time has been increasingly used to predict or
study scaling in RO. Scaling in RO occurs when the lattice ions in supersaturated solution
start to agglomerate and form nuclei or clusters. If the size of the cluster is below the critical
size, then the formed crystal will re-dissolve into the solution and if the cluster size is above
the critical size, then the crystal become stable and grow bigger. Induction time is defined
as the time required for the supersaturated solution to form nuclei in the critical cluster size
dimensions just before it becomes stable and starts to accelerate growth (He et al., 1995,

9
Experimental Methods for Membrane Applications

Boerlage et al., 2000, Boerlage, 2001). Induction time can be determined by measuring the
change in pH, turbidity, conductivity, and specific ion concentration in the water over a
period in a controlled lab environment (Waly, 2011, Mangal, 2023). Measuring induction
time is potentially useful tool in developing new design, pretreatment and operational
strategies to control scaling in RO but better understanding is needed to measure impact
of other inorganics as well as organic components in the water on the induction time of
specific salts.

1.4 ORGANIC FOULING POTENTIAL

The accumulation of organic matter in membrane systems can directly cause substantial
decline in operational performance (e.g., permeability). Organic foulants are typically
abundant in surface waters but are also present in ground water sources. They can originate
from natural sources from human activities (e.g., sewage discharge) or from chemicals
used in the pre-treatment processes. Identifying organic foulants and understanding their
characteristics provides valuable insights on how to prevent organic fouling, by choosing
the appropriate pre-treatment and operational and cleaning strategies. Currently available
assessments tools for organic fouling can range from simple spectrophotometric methods
to more advanced chromatographic techniques (see Figure 5).

Total Organic Carbon (TOC)

Dissolved Organic Particulate Organic
Carbon (<0.45µm) Carbon, POC (>0.45µm)

Low molecular weight (LMW) organics Biopolymers (>> 1 kDa)

Building blocks Humic substances

(0.3 - 0.5 kDa) (0.5 - 1.2 kDa) Proteins (PT) Polysaccharides (PS)

LMW acids LMW neutrals Other Glyco- Acidic Other

(<0.35 kDa) (<0.35 kDa) PT proteins PS PS

TOC/POC UV254 TEP + precursors

FEEM TEP0.4µm/TEP10kDa

LC-OCD-UVD-OND
Analytical methods

Figure 5 Overview of the different fractions of organic matter in the water and the applicable
analytical methods to identify or quantify them (Villacorte et al., 2021). LC-OCD-UVD-
OND is liquid chromatography (LC) with inline detectors for organic carbon (OCD),
UV absorbance at 254 nm (UVD) and organic nitrogen (OND); FEEM is fluorescence
excitation-emission matrices; TEP refers to transparent exopolymer particles measured
with a 0.4 μm or 10 kDa membrane.

10
 Chapter 1

1.4.1 Organic carbon

Traditionally, the presence of organic matter in RO feedwater is assessed by measuring
the total organic carbon (TOC) or dissolved organic carbon (DOC). Routine TOC/
DOC measurements have been used for monitoring the bulk organic fouling potential
of the feedwater. However, not all organic carbon in the water can be directly associated
with organic fouling. So, in many cases, TOC/ DOC as such is not sufficient to assess the
variations in organic fouling potential of the feedwater. According to Voutchkov (2010),
if TOC concentration is below 0.5 mg/L then biofouling is unlikely; and above 2 mg/L,
biofouling is very likely. Nevertheless, as the TOC is a bulk concentration value, it is very
important to identify the fraction of the TOC responsible for bacterial growth. Cationic
organic polymers have also been reported with negative fouling effects on RO membranes,
as they may coprecipitate with negatively charged antiscalants and foul the membrane
irreversibly (Ekowati et al., 2014, Peña et al., 2015, DuPont, 2023).

1.4.2 UV absorbance and fluorescence

Hydrophobic and aromatic organic compounds such as humic substances can be
abundant in surface water sources. UV absorbance at 254 nm (UVA254) is typically used
as an indicator of their relative abundance. Specific UVA254 (SUVA), defined as the ratio
between UVA254 and DOC, is a parameter shown to correlate with the aromatic contents
and hydrophobicity of organic matter (Baghoth, 2012). Such measurement is simple,
fast and can be measured routinely. UV absorbing aromatic and hydrophobic organics are
typically removed by coagulation pretreatment more efficiently than the non-UV absorbing
hydrophilic components (Matilainen et al., 2011). Hence, a low SUVA after pretreatment
does not necessarily mean low organic fouling potential because hydrophilic compounds
may still remain in the RO feedwater. Moreover, humic substances were reported to be less
problematic foulant than hydrophilic organic substances (Amy et al., 2011). Therefore, it
is recommended that UVA254 or SUVA is supplemented with other measurements when
assessing organic fouling potential of RO feedwater.

An emerging technique to characterize organic foulants is by measuring differences

in fluorescence spectra associated with specific organic compounds in a sample using
spectrofluorometer (see Chapter 12). The technique can be applied on both liquid (e.g.,
feedwater) and solid (e.g., membrane surface) samples, and generates a three-dimensional
excitation-emission matrix (EEM). The location of EEM peaks provides a qualitative
indication of types of organic molecules present in water samples (Westerhoff et al., 2001).
For example, humic-like fluorescence peak could be clearly discriminated from protein-like
peak in the EEM spectra. In general, fluorescence spectroscopy can be used as a rapid and
sensitive method to characterize dissolved organic matter, but analysis is limited to organic
components in the water that contains fluorophores. For instance, other organic matter
components such as polysaccharides do not fluoresce and could not be analyzed using the
EEM spectra. Therefore, FEEM is typically not a standalone method for organic fouling
assessment.

1.4.3 LC-OCD
Liquid chromatography-organic carbon detection (LC-OCD) is an advanced method for
fractionation and measurement of organic carbon by size-exclusion chromatography

11
Experimental Methods for Membrane Applications

followed by inline analyses through multiple detectors (i.e., organic carbon, UV254 and
organic nitrogen). While DOC measures organic carbon in bulk, LC-OCD fractionates DOC
based on their molecular weight and hydrophobicity (see Chapter 11). LC-OCD measures
the concentration of organic carbon fractions as biopolymers, humic substances, building
blocks, low molecular weight acids, and low molecular weight neutrals (Huber et al., 2011).
The limit of detection of the method for each fraction is at ppb level and the method has
been adapted for high salinity water (Amy et al., 2011). LC-OCD method has been applied
in many studies to characterize organic matter in surface water for assessment of fouling
potential of the different fractions and their removal through the water treatment processes
(Lozier et al., 2009, Salinas Rodriguez et al., 2009, Villacorte et al., 2009, Simon et al., 2013,
Ho et al., 2015, Shanmuganathan et al., 2015, Jeong et al., 2016, Yin et al., 2019, Altmann
et al., 2023).

1.4.4 TEP
Transparent exopolymer particles (TEP) and their precursors can have a major role in organic
and biological fouling in membrane filtration processes. Berman and Holenberg (2005) first
proposed the potential role of TEP as a major initiator of biofilm leading to biofouling in
reverse osmosis (RO) membranes. Consequently, various experimental studies investigated
the role of TEP, on biofouling (Bar-Zeev et al., 2012) and organic fouling (Villacorte et al.,
2021) in membrane systems. Several methods have been developed, adopted, modified, and
demonstrated to quantify TEP and elucidate their role to membrane fouling (see Chapter 13).
The TEP0.4μm and TEP10kDa methods has been successfully used to semi-quantitatively
demonstrate the role of TEP on the operational performance of membrane processes
(Villacorte et al., 2015a). They have been applied as an indicator of fouling potential of RO
feedwater and showed significant correlation with MFI-UF (Villacorte et al., 2015b). So far,
no study has successfully determined the threshold level of TEP in the feedwater at which
membrane fouling will likely not occur. TEP methods still have some inherent limitations
(Discart et al., 2015, Bittar et al., 2018, Li et al., 2018), so it should be implemented with
proper attention to the protocol used and by someone who is experienced with laboratory
analytical techniques.

The quantification of algae in the RO feedwater source can act as an indication of the
occurrence of algal blooms, which can generate organic foulants like TEP. Algae can be
quantified directly through microscopic counting as cell density or indirectly through
chlorophyll-a measurement. Standard chlorophyll-a methods are widely available (Arar and
Collins, 1997; ASTM D3731-20, 2020; Lipps et al., 2023c). A spike in algae concentrations
can coincide with an increase of organic fouling mainly due to extracellular substances
released by algae. However, a spike in algae density or chlorophyll-a concentration in the
water does not necessarily result in high organic fouling because bloom-forming algal
species can vary in shapes/sizes, specific chlorophyll-a concentration, TEP/EPS production
and their characteristics that affects their removal in the pre-treatment process and their
organic fouling potential to RO.

1.4.5 Oil and grease

One of the most detrimental types of organic foulants are oily compounds which can impact
both the operation and integrity of the membrane units. Ideally, oil and grease should not
exceed 0.1 mg/L in the feed water of a RO system because it can attach and accumulate

12
 Chapter 1

on the membrane surface which may lead to irreversible organic fouling. Pre-treatment is
necessary in the case of treating produced water from oil and gas extraction and industrial
wastewater. The method to determine oil and grease in water consists of extraction by
liquid/liquid extraction, solid phase extraction, or microwave extraction and measurement
by gravimetric and infrared analysis. The standard method for determination of oil and
grease in water through these various extraction methods and analyses can be found in Lipps
et al. (2023b), USEPA (2010), ASTM D7066-04 (2010) and ASTM D7575-11 (2017).
Generally, the infrared methods are more sensitive compared to gravimetric methods
with detection limit of approximately 1 mg/L and even down to 0.1 mg/L when using
tetrachloroethylene as the extraction solvent (Farmaki et al., 2007).

1.5 BIOFOULING POTENTIAL

Biofouling occurs due to the growth of microorganisms on the membrane and feed spacer
of the RO system. Biofouling is a common issue in most RO desalination plants (Peña et
al., 2022) and is often inevitable even when bacteria/microorganisms are completely
removed through the pre-treatment system (i.e., using microfiltration or ultrafiltration
system). If a single bacteria/ microorganism finds their way to reach the RO system, it can
rapidly grow and form a biofilm layer on the membrane and/or feed spacer of the RO when
nutrient concentrations in the feedwater are limited. Biofouling occurs often in plants with
open water sources (e.g., sea, river, lake) as they typically contain higher concentrations of
organics and other nutrients. Thus, biofouling of brackish water RO is less frequent than
that of seawater RO system.

1.5.1 Bacterial growth potential

Given that RO biofouling is mainly due to bacterial growth on the membrane surface and
feed spacer, a method to measure the bacterial growth potential (BGP) in the RO feed water
and through the pre-treatment process was developed by Abushaban (2019). The full
description of the protocol and optimization of each step are discussed in Chapter 15. The
basic concept of the method is to measure the growth of constant number of indigenous
bacteria due to the presence of any nutrients (C: N: P) in a seawater sample. The limit of
detection (LoD) of the method is 10 μg-C (as glucose)/L which is low enough to measure
the BGP in the BWRO feedwater. However, this LoD might not be low enough in some
BWRO system where BGP can be much lower, especially after the pre-treatment process.
The correlation between BGP in the SWRO feed water and biofouling in selected SWRO
membrane systems in Australia, Europe, and the Middle East was investigated (Abushaban
et al., 2019b, Abushaban et al., 2020, Abushaban et al., 2021). Results show that a higher
BGP in the SWRO feedwater (100 - 950 μg-C/L) corresponds to a higher normalised
pressure drop or higher CIP cleaning frequency in the RO system, a demonstration of
the applicability of BGP as a biofouling indicator in RO systems. It was estimated that
a BGP value of 70 μg-C/L in the SWRO feedwater requires once per year CIP frequency
(Abushaban, 2019, Abushaban et al., 2019a, Dhakal et al., 2020, Salinas Rodriguez et al.,
2021b). Consequently, a safe level of BGP (below 70 μg-C/L as glucose equivalent) was
preliminarily proposed to control biofouling in SWRO desalination plants.

13
Experimental Methods for Membrane Applications

The advantage of the BGP method over other methods is that it measures the growth of
indigenous bacteria until when the biodegradable nutrients present in seawater sample is
depleted. Moreover, the duration of the test is around 4-5 days which is relatively short
compared to conventional assimilable organic carbon (12-14 days) and biomass production
potential (15-28 days) methods. A test duration of that takes days can still be a practical
limitation of the method, particularly, when the concentration of BGP varies significantly
(hourly or daily) in the water source. However, it should be noted that biofilm formation
usually takes couple of weeks to be grow on the membrane system. Thus, getting the results
within a few days can still be considered as an early detection of biofouling and a corrective
action can still be made to remove or control growth of microorganisms.

There are other limitations in the application of the BGP method. Firstly, the protocol itself
is quite complex. The complexity is due to the requirement of carbon-free glassware of each
step. Any introduction of carbon or organic matter during handling and measuring BGP will
negatively affect the results (Abushaban, 2019). Secondly, the cost of measuring BGP is high
as it needs a qualified and skilled technician to measure the sample for around two weeks
(including preparation and measurements) and the reagents used to measure microbial
adenosine triphosphate (ATP) are also expensive. The frequency of measuring the BGP in
SWRO feed water and/or along the pre-treatment processes, depends on the water source
quality and the expected variation in the quality from season/month/week/day to another.
Finally, another limitation of BGP method is that the results are influenced by the salinity
of the water which is normally not the case in many desalination plants where the salinity
of the water source is somehow constant. If the salinity of the water source is changed, it
means new calibration curves for both ATP and BGP should be established. The higher the
salinity the lower ATP signal is expected and thus lower slope of BGP.

In general, BGP method is a promising assessment method to control biofouling in SWRO

system. However, the method is rather complex and thus it is currently applied mainly for
research studies and not yet on a routine basis.

1.5.2 Assimilable organic carbon

Assimilable organic carbon (AOC) method has been developed 4 decades ago, mainly to
monitor growth potential in drinking water distribution systems. In the recent years, the
method has been adapted for seawater application. The difference between BGP and AOC
methods is that AOC uses a single strain of bacteria while BGP uses a mix of indigenous
bacteria (Abushaban et al., 2022). The use of single strain of bacteria enables standardization
the method (use one conversion value for samples from different location). Further
optimization of the inoculum used in the test led to substantial reduction the duration of
the test to 1-2 days.

The AOC method is that the method is simpler to implement than the BGP method.
However, it is considered less accurate than BGP as it represents bacterial growth of a single
strain of bacteria. It was also reported that the difference in bacterial growth of a single strain
of bacteria is typically at least 20% lower than the growth of indigenous bacteria in fresh
water (Ross et al., 2013).

14
 Chapter 1

Various methods of AOC have been developed over the years using different bacterial strains
such as Vibrio fischeri and Vibrio harveyi (Weinrich et al., 2011, Jeong et al., 2013). The LoD
of these two methods are 10 and 0.1 μg-C/L, respectively. However, the extremely low LoD
(0.1 μg-C/L) reported is questionable as it was calculated after subtracting the AOC of the
blank, which was >50 μg-C/L.

Weinrich et al. (2015) reported a good correlation between AOC and differential pressure
increase and specific flux decline at the Tampa Bay pilot seawater desalination plant
where the feedwater AOC concentrations measured were between 22 and 161 μg-C/L.
Consequently, a preliminary threshold concentration of AOC (50 μg-C/L) was proposed
using Vibrio harveyi bacteria in seawater (Weinrich et al., 2015). So far, the reported AOC
concentrations in SWRO feed water varies between 10 and 220 μg-C/L.

Overall, the AOC method is applicable to monitor biofouling potential along the SWRO
pre-treatment process and in the SWRO feed water. However, the accuracy of the reported
bacterial growth may not represent actual conditions in the RO system as only one single
strain is used to mimic the bacterial growth.

1.5.3 Biodegradable dissolved organic carbon

Not all DOC is bioavailable or can be directly utilized by microorganisms. Biodegradable
dissolved organic carbon (BDOC) represents the fraction of DOC that can be utilized by
microorganisms. It is calculated by subtracting the initial DOC of the water sample from the
final DOC at the end of incubation period. The incubation period in BDOC measurement
can be varied, depending on the time required to reach a stable DOC. Servais et al. (1987)
suggested 4 weeks of incubation period while shorter incubation periods were introduced
in various application of BDOC method (Joret et al., 1989, Frías et al., 1992, Kadjeski et al.,
2020). BDOC includes a larger fraction of total organic carbon (10-30%) with LoD of 0.1-
0.2 mg/L which is about 10 folds higher than LoD of BGP and AOC. In general, the BDOC
method is time consuming for routine monitoring and less sensitive compared to AOC and
BGP methods. Additionally, in membrane filtration, it is likely that a large portion of BDOC
is retained on the membrane while still allowing the majority of AOC to pass through
(Escobar et al., 2000, Escobar and Randall, 2001).

1.5.4 Phosphate
Biofilm formation in membranes can be largely influenced by the C:N:P nutrient mass ratio
in the water which is ideally 100:23:4.3. Based on this ratio, the requirement for phosphorus
(P) is lower than other substrates (carbon and nitrogen), so a small change in P can lead to
a significant change in microbial growth. Limiting P down to a low level can disrupt the
nutrient balance, restricting bacterial growth in the water and reducing biofouling (van
der Kooij et al., 2007, Galjaard et al., 2008, Jacobson et al., 2009, Vrouwenvelder et al.,
2010, Kim et al., 2014). Reliable analytical methods to measure phosphate down to sub-
ppb level is critical to this strategy. Standard phosphate analytical methods are available
such as the widely used phosphomolybdenum blue method (Lipps et al., 2023a). The
method has been applied and adapted in biofouling studies of water containing very low
phosphate concentrations (Abushaban et al., 2020, Javier et al., 2020). Ultra-low phosphate

15
Experimental Methods for Membrane Applications

concentration down to below 0.3 μg PO4-P/L was reported to limit biofouling even in water
with high concentration of organic carbon (Vrouwenvelder et al., 2010). Some antiscalants
(e.g., phosphonates) which are added to the feedwater to prevent scaling in RO, contain
phosphate which has been found to cause biofouling (Vrouwenvelder et al., 2000, Sweity
et al., 2013, Sweity et al., 2015, Hasanin et al., 2023). It is therefore recommended that the
dosage and type of antiscalants be taken into consideration when applied as pretreatment
for RO.

1.6 OUTLOOK AND OPPORTUNITIES

RO desalination of brackish and saline water sources is increasingly applied globally to

solve water scarcity challenges in the utility and industry sector. Assessing the fouling and
scaling potential of the feedwater source is highly critical when designing and operating the
RO desalination plant including its pre-treatment units. For many years, RO membrane
suppliers have provided specific guidelines of the ideal RO feedwater quality to minimize
possible particulate, inorganic, organic and biological fouling, and scaling issues in the
plant. However, some of these standard water quality parameters and indices have some
limitations so alternative/supplemental assessment and characterization tools has been
developed over the years for feedwater monitoring and experimental investigations.

Measuring the individual concentration of all potential foulants present in the RO feed
water is sort of a mission impossible due to costs, duration and specialized laboratory
facilities needed. Particulate fouling indices like the SDI or MFI-0.45 can be used in a daily
basis for monitoring of water quality with onsite measurements. Other parameters like the
MFI-UF constant flux are promising due to its sensitivity and ability to predict accurately the
rate of RO membrane fouling.

Standardized inorganic fouling and scaling assessment methods are widely available for lab-
based analyses. Online monitoring is currently a challenge and would be an important area
for further development. The induction time concept for predicting when scaling can occur
is a promising tool to developing RO plant design and control strategies to minimize scaling
issues (e.g., Mangal et al. (2022)).

The presence of organic foulants can be routinely measured with offline/online TOC
measurements. Routine chlorophyll-a measurement can be beneficial for feedwater sources
that are prone to seasonal algal blooms. For in-depth investigations of organic fouling,
more advanced or complex assessment methods (e.g., LC-OCD, FEEM, TEP) can be further
considered.

Biological fouling is currently the leading cause of operational challenges in RO applications.

Genomic tools (see Chapter 14) have been applied to identify bacterial communities often
associated with biofilm development in RO. Promising methods for assessing biofouling
potential of RO feedwater are either based on the bacterial growth capacity (see Chapters
15 and 16) or based on the concentration of specific limiting nutrient (phosphate) in the
feedwater. However, these methods either requires days to weeks of incubation to generate
results or have high sensitivity to contamination. Future developments should focus on
overcoming either one of these limitations.

16
 Chapter 1

Ideally, assessment methods and indices should be made practical for onsite monitoring
and low cost for consumables, enabling it for wide use. Testing conditions should be
standardized/calibrated and should specify the method LOD and their applicability in
fresh and saline water matrices. The need and opportunities for real time online monitoring
should be explored further. In some cases, a few analyses per day is sufficient and in other
cases once a day or once a week are considered sufficient.

Chemicals used in the treatment process may contribute to the membrane fouling
development due to introduction of nutrients or organic foulants. Hence, it is recommended
that operators also assess the fouling potential of such chemicals before applying it to the
RO system.

Current RO feedwater quality guideline values recommended by membrane manufacturers

may need to be verified under local conditions and modified/adjusted accordingly based
on latest developments. On the other hand, feedwater guidelines for emerging desalination
technologies such as forward osmosis (see Chapter 4) and membrane distillation (see
Chapter 5) are currently non-existent. Future developments of these new technologies
should also include understanding their propensity to fouling or scaling by applying current
or new feedwater quality assessment methods.

1.7 ABBREVIATIONS AND SYMBOLS

AOC Assimilable organic carbon

ATP Adenosine tri-phosphate
BGP Bacterial growth potential
BFR Biofilm formation rate
BWRO Brackish water reverse osmosis
CIP Cleaning in place
COD Chemical oxygen demand
DOC Dissolved organic carbon
EEM Excitation-emission matrix
LC-OCD Liquid chromatography-organic carbon detection
LoD Limit of detection
MED Multi-effect distillation
MFI-0.45 Modified fouling index, constant pressure, 0.45 μm filter
MFI-UF Modified fouling index, constant flux, 10 kDa or 100 kDa membrane filter
MSF Multi stage flash distillation
RO Reverse osmosis
SDI Silt density index, %/min
S&DSI Stiff and Davis saturation index
SWRO Seawater reverse osmosis
TEP Transparent exopolymer particles
TOC Total organic carbon

NB. This chapter is currently being prepared for submission as a scientific article in a journal.

17
Experimental Methods for Membrane Applications

1.8 REFERENCES

Abunada M (2023). Prediction of Particulate Fouling in Reverse Osmosis Systems: MFI-UF Method
Development and Application. PhD Dissertation, IHE Delft
Abunada M, Dhakal N, Gulrez R, Ajok P, Li Y, Abushaban A, Smit H, Moed D, Ghaffour N, Schippers
JC, Kennedy MD (2023) Prediction of particulate fouling in full-scale reverse osmosis plants
using the modified fouling index – ultrafiltration (MFI-UF) method. Desalination 553: 116478
DOI https://doi.org/10.1016/j.desal.2023.116478
Abushaban A (2019) Assessing Bacterial Growth Potential in Seawater Reverse Osmosis Pretreatment:
Method Development and Applications CRC Press9781000034707.
Abushaban A, Salinas-Rodriguez SG, Dhakal N, Schippers JC, Kennedy MD (2019a) Assessing
pretreatment and seawater reverse osmosis performance using an ATP-based bacterial
growth potential method. Desalination 467: 210-218 DOI https://doi.org/10.1016/j.
desal.2019.06.001
Abushaban A, Salinas-Rodriguez SG, Kapala M, Pastorelli D, Schippers JC, Mondal S, Goueli S, Kennedy
MD (2020) Monitoring Biofouling Potential Using ATP-Based Bacterial Growth Potential in
SWRO Pre-Treatment of a Full-Scale Plant. Membranes 10: 360
Abushaban A, Salinas-Rodriguez SG, Pastorelli D, Schippers JC, Mondal S, Goueli S, Kennedy MD
(2021) Assessing Pretreatment Effectiveness for Particulate, Organic and Biological Fouling in a
Full-Scale SWRO Desalination Plant. Membranes 11: 167
Abushaban A, Salinas-Rodriguez SG, Philibert M, Le Bouille L, Necibi MC, Chehbouni A (2022) Biofouling
potential indicators to assess pretreatment and mitigate biofouling in SWRO membranes: A short
review. Desalination 527: 115543 DOI https://doi.org/10.1016/j.desal.2021.115543
Abushaban A, Salinas-Rodriguez SG, Schippers JC, Kennedy MD (2019b) Assessing pretreatment and
seawater reverse osmosis performance using an ATP-based bacterial growth potential method.
Desalination 467: 210-218 DOI https://doi.org/10.1016/j.desal.2019.06.001
Alhadidi A, Blankert B, Kemperman AJB, Schippers JC, Wessling M, van der Meer WGJ (2011a) Effect of
testing conditions and filtration mechanisms on SDI. Journal of Membrane Science 381: 142-151
Alhadidi A, Kemperman AJB, Schippers JC, Wessling M, van der Meer WGJ (2011b) The influence of
membrane properties on the Silt Density Index. Journal of Membrane Science 384: 205-218
Altmann T, Rousseva A, Vrouwenvelder J, Shaw M, Das R (2023) Effectiveness of ceramic ultrafiltration
as pretreatment for seawater reverse osmosis. Desalination 564: 116781 DOI https://doi.
org/10.1016/j.desal.2023.116781
Amy GL, Salinas Rodríguez SG, Kennedy MD, Schippers JC, Rapenne S, Remize P-J, Barbe C, Manes
C-LdO, West NJ, Lebaron P, Kooij Dvd, Veenendaal H, Schaule G, Petrowski K, Huber S, Sim
LN, Ye Y, Chen V, Fane AG (2011) Water quality assessment tools. In: Drioli E, Criscuoli A,
Macedonio F (eds) Membrane-Based Desalination - An Integrated Approach (MEDINA): 192
10.2166/9781780400914
Arar EJ, Collins GB (1997) Method 445.0: In vitro determination of chlorophyll a and pheophytin a in
marine and freshwater algae by fluorescence United States Environmental Protection Agency, Office
of Research and Development, National Exposure Research Laboratory, Cincinnati, pp. 1-22
ASTM D857-17 (2017) Standard Test Method for Aluminum in Water ASTM International, West
Conshohocken, PA, pp. 5 https://www.astm.org/d0857-17.html.
ASTM D858-17 (2017) Standard Test Methods for Manganese in Water ASTM International, West
Conshohocken, PA, pp. 9 https://www.astm.org/d0858-17.html.
ASTM D859-16(2021)e1 (2021) Standard Test Method for Silica in Water ASTM International, West
Conshohocken, PA, pp. 5 https://www.astm.org/d0859-16r21e01.html.

18
 Chapter 1

ASTM D1068-15 (2016) Standard Test Methods for Iron in Water ASTM International, West
Conshohocken, PA, pp. 13 https://www.astm.org/d1068-15.html.
ASTM D1252-06(2020) (2020) Standard Test Methods for Chemical Oxygen Demand (Dichromate
Oxygen Demand) of Water ASTM International, West Conshohocken, PA https://www.astm.
org/d1252-06r20.html.
ASTM D1253-14(2021)e1 (2021) Standard Test Method for Residual Chlorine in Water ASTM
International, West Conshohocken, PA, pp. 6 https://www.astm.org/d1253-14r21e01.html.
ASTM D1293-12 (2018) Standard Test Methods for pH of Water ASTM International, West
Conshohocken, PA, pp. 10 https://www.astm.org/d1293-12.html.
ASTM D1889-00 (2017) Standard Test Method for Turbidity of Water (Withdrawn 2007) ASTM
International, West Conshohocken, PA https://www.astm.org/d1889-00.html.
ASTM D2579-93e1 (2021) Standard Test Method for Total Organic Carbon in Water (Withdrawn 2002)
ASTM International, West Conshohocken, PA https://www.astm.org/d2579-93e01.html.
ASTM D3731-20 (2020) Standard Practices for Measurement of Chlorophyll Content of Algae in
Surface Waters ASTM International, West Conshohocken, PA, pp. 5 www.astm.org.
ASTM D3739-19 (2019) Standard Practice for Calculation and Adjustment of the Langelier Saturation
Index for Reverse Osmosis ASTM International, West Conshohocken, PA, pp. 4 https://www.
astm.org/d3739-19.html.
ASTM D4189 - 07 (2007) Standard Test Method for Silt Density Index (SDI) of Water ASTM
International, West Conshohocken, PA
ASTM D4189 - 14 (2014) Standard Test Method for Silt Density Index (SDI) of Water ASTM
International, West Conshohocken, PA
ASTM D4582-91(2001) (2017) Standard Practice for Calculation and Adjustment of the Stiff and
Davis Stability Index for Reverse Osmosis ASTM International, West Conshohocken, PA, pp. 4
https://www.astm.org/d4582-91r01.html.
ASTM D7066-04 (2010) Standard Test Method for dimer/trimer of chlorotrifluoroethylene (S-
316) Recoverable Oil and Grease and Nonpolar Material by Infrared Determination ASTM
International, West Conshohocken, PA, pp. 9 www.astm.org.
ASTM D7575-11 (2017) Standard Test Method for Solvent-Free Membrane Recoverable Oil and
Grease by Infrared Determination ASTM International, West Conshohocken, PA, pp. 12 www.
astm.org.
ASTM D8002 - 15 (2015) Standard Test Method for Modified Fouling Index (MFI-0.45) of Water
ASTM International, West Conshohocken, PA www.astm.org.
Badruzzaman M, Voutchkov N, Weinrich L, Jacangelo JG (2019) Selection of pretreatment technologies
for seawater reverse osmosis plants: A review. Desalination 449: 78-91 DOI https://doi.
org/10.1016/j.desal.2018.10.006
Baghoth SA (2012) Characterizing natural organic matter in drinking water treatment processes and
trains CRC Press/Balkema, Delft. 9781138000261.
Bar-Zeev E, Berman-Frank I, Girshevitz O, Berman T (2012) Revised paradigm of aquatic biofilm
formation facilitated by microgel transparent exopolymer particles. Proceedings of the National
Academy of Sciences 109: 9119-9124 DOI doi:10.1073/pnas.1203708109
Bassa FGM (2021). Evaluating surface water pre-treatment with the MFI-UF for assessing fouling in
membrane systems. MSc, IHE Delft UWS-WSE/21.25.
Belila A, El-Chakhtoura J, Otaibi N, Muyzer G, Gonzalez-Gil G, Saikaly PE, van Loosdrecht MCM,
Vrouwenvelder JS (2016) Bacterial community structure and variation in a full-scale seawater
desalination plant for drinking water production. Water Research 94: 62-72 DOI https://doi.
org/10.1016/j.watres.2016.02.039

19
Experimental Methods for Membrane Applications

Berman T, Holenberg M (2005) Don’t fall foul of biofilm through high TEP levels. Filtration & Separation
42: 30-32 DOI https://doi.org/10.1016/S0015-1882(05)70517-6
Birch H, Pasture LdL, Gall M, Gasson C, Pankratz T, Quaresma M, Qureshi Z, Walker C (2023) Market focus
deck Desalination and reuse October 2023 Media Analytics Ltd., Oxford, UK. 978-1-907467-67-7.
Bittar TB, Passow U, Hamaraty L, Bidle KD, Harvey EL (2018) An updated method for the calibration of
transparent exopolymer particle measurements. Limnology and Oceanography: Methods 16: 621-
628 DOI https://doi.org/10.1002/lom3.10268
Boerlage SFE (2001) Scaling and Particulate Fouling in Membrane Filtration Systems Swets&Zeitlinger
Publishers, Lisse. 90 5809 242 9.
Boerlage SFE, Kennedy MD, Aniye MP, Abogrean EM, El-Hodali DEY, Tarawneh ZS, Schippers JC
(2000a) Modified Fouling Index ultrafiltration to compare pretreatment processes of reverse
osmosis feedwater. Desalination 131: 201-214
Boerlage SFE, Kennedy MD, Bremere I, Witkamp GJ, van der Hoek JP, Schippers JC (2000b) Stable
barium sulphate supersaturation in reverse osmosis. Journal of Membrane Science 179: 53-68
DOI 10.1016/s0376-7388(00)00504-4
DesalData (2023) 36th Desalination Plants Inventory In: Global Water Intelligence (ed)
Dhakal N, Abushaban A, Mangal MN, Abunada M, Schippers JC, Kennedy M (2020) Membrane Fouling
and Scaling in Reverse Osmosis. In: Sapalidis A (ed) Membrane Desalination: 325-344
Dhakal N, Salinas Rodriguez SG, Schippers JC, Kennedy MD (2015) Induction time measurements in
two brackish water reverse osmosis plants for calcium carbonate precipitation. Desalination and
Water Treatment 53: 285-293 DOI 10.1080/19443994.2014.903870
Discart V, Bilad MR, Vankelecom IFJ (2015) Critical Evaluation of the Determination Methods
for Transparent Exopolymer Particles, Agents of Membrane Fouling. Critical Reviews in
Environmental Science and Technology 45: 167-192 DOI 10.1080/10643389.2013.829982
DuPont (2020) FILMTEC™ Reverse Osmosis Membranes Technical Manual, Form No. 45-D01504-en,
Rev. 3Water solutions, pp. 207
DuPont (2023) FILMTEC™ Reverse Osmosis Membranes Technical Manual, Form No. 45-D01504-en,
Rev. 16Water solutions, pp. 188 https://www.dupont.com/content/dam/dupont/amer/us/
en/water-solutions/public/documents/en/RO-NF-FilmTec-Manual-45-D01504-en.pdf.
Ekowati Y, Msuya M, Salinas Rodriguez SG, Veenendaal G, Schippers JC, Kennedy MD (2014) Synthetic
organic polymer fouling in municipal wastewater reuse reverse osmosis. Journal of Water Reuse
and Desalination 4: 125-136 DOI doi:10.2166/wrd.2014.046
Escobar IC, Hong S, Randall AA (2000) Removal of assimilable organic carbon and biodegradable
dissolved organic carbon by reverse osmosis and nanofiltration membranes. Journal of Membrane
Science 175: 1-17 DOI https://doi.org/10.1016/S0376-7388(00)00398-7
Escobar IC, Randall AA (2001) Assimilable organic carbon (AOC) and biodegradable dissolved organic
carbon (BDOC):: complementary measurements. Water Research 35: 4444-4454 DOI https://
doi.org/10.1016/S0043-1354(01)00173-7
Farmaki E, Kaloudis T, Dimitrou K, Thanasoulias N, Kousouris L, Tzoumerkas F (2007) Validation of a
FT-IR method for the determination of oils and grease in water using tetrachloroethylene as the
extraction solvent. Desalination 210: 52-60 DOI https://doi.org/10.1016/j.desal.2006.05.032
Frías J, Ribas F, Lucena F (1992) A method for the measurement of biodegradable organic carbon in waters.
Water Research 26: 255-258 DOI https://doi.org/10.1016/0043-1354(92)90226-T
Galjaard G, Lampe M, Kamp PC (2008) (8) years of RO-experience at WTP Heemskerk biofouling aspects.
Paper presented at the Membrane Technology Conference, Naples, Florida2008

20
 Chapter 1

Hasanin G, Mosquera AM, Emwas A-H, Altmann T, Das R, Buijs PJ, Vrouwenvelder JS, Gonzalez-Gil
G (2023) The microbial growth potential of antiscalants used in seawater desalination. Water
Research 233: 119802 DOI https://doi.org/10.1016/j.watres.2023.119802
He S, Oddo JE, Tomson MB (1995) The Nucleation Kinetics of Barium Sulfate in NaCl Solutions up to 6 m
and 90°C. Journal of Colloid and Interface Science 174: 319-326 DOI https://doi.org/10.1006/
jcis.1995.1397
Ho JS, Ma Z, Qin J, Sim SH, Toh C-S (2015) Inline coagulation–ultrafiltration as the pretreatment for
reverse osmosis brine treatment and recovery. Desalination 365: 242-249 DOI https://doi.
org/10.1016/j.desal.2015.03.018
Huber SA, Balz A, Abert M, Pronk W (2011) Characterisation of aquatic humic and non-humic matter with
size-exclusion chromatography – organic carbon detection – organic nitrogen detection (LC-OCD-
OND). Water Research 45: 879-885 DOI http://dx.doi.org/10.1016/j.watres.2010.09.023
Jacobson JD, Kennedy MD, Amy G, Schippers JC (2009) Phosphate limitation in reverse osmosis: An
option to control biofouling? Desalination and Water Treatment 5: 198-206 DOI 10.5004/
dwt.2009.578
Javier L, Farhat NM, Desmond P, Linares RV, Bucs S, Kruithof JC, Vrouwenvelder JS (2020) Biofouling
control by phosphorus limitation strongly depends on the assimilable organic carbon concentration.
Water Research 183: 116051 DOI https://doi.org/10.1016/j.watres.2020.116051
Jeong S, Naidu G, Vigneswaran S, Ma CH, Rice SA (2013) A rapid bioluminescence-based test of
assimilable organic carbon for seawater. Desalination 317: 160-165 DOI http://dx.doi.
org/10.1016/j.desal.2013.03.005
Jeong S, Naidu G, Vollprecht R, Leiknes T, Vigneswaran S (2016) In-depth analyses of organic matters in a
full-scale seawater desalination plant and an autopsy of reverse osmosis membrane. Separation and
Purification Technology 162: 171-179 DOI https://doi.org/10.1016/j.seppur.2016.02.029
Joret C, Levi Y, Gilbert M (1989) The measurement of biodegradable organic carbon (BDOC): a tool in
water treatment. Water Supply 7: 41-45
Kadjeski M, Fasching C, Xenopoulos MA (2020) Synchronous Biodegradability and Production of
Dissolved Organic Matter in Two Streams of Varying Land Use. Frontiers in Microbiology 11 DOI
10.3389/fmicb.2020.568629
Khan MT, Busch M, Molina VG, Emwas A-H, Aubry C, Croue J-P (2014) How different is the composition
of the fouling layer of wastewater reuse and seawater desalination RO membranes? Water Research
59: 271-282 DOI https://doi.org/10.1016/j.watres.2014.04.020
Kim C-M, Kim S-J, Kim LH, Shin MS, Yu H-W, Kim IS (2014) Effects of phosphate limitation in feed
water on biofouling in forward osmosis (FO) process. Desalination 349: 51-59 DOI https://doi.
org/10.1016/j.desal.2014.06.013
Kurihara M, Ito Y (2020) Sustainable Seawater Reverse Osmosis Desalination as Green Desalination
in the 21st Century. Journal of Membrane Science and Research 6: 20-29 DOI 10.22079/
jmsr.2019.109807.1272
Li X, Skillman L, Li D, Ela WP (2018) Comparison of Alcian blue and total carbohydrate assays for
quantitation of transparent exopolymer particles (TEP) in biofouling studies. Water Research 133:
60-68 DOI https://doi.org/10.1016/j.watres.2017.12.021
Lipps WC, Braun-Howland EB, Baxter TE (2023a) 4500-P Phosphorus. In: American Public Health
Association, American Water Works Association, Water Environment Federation, (ed) Standard
Methods for the Examination of Water and Wastewater 24th Edition, 24th edn:
Lipps WC, Braun-Howland EB, Baxter TE (2023b) 5520 Oil and grease. In: American Public Health
Association, American Water Works Association, Water Environment Federation, (ed) Standard
Methods for the Examination of Water and Wastewater 24th Edition, 24th edn:

21
Experimental Methods for Membrane Applications

Lipps WC, Braun-Howland EB, Baxter TE (2023c) 10150 Chloropyll a. In: American Public Health
Association, American Water Works Association, Water Environment Federation, (ed) Standard
Methods for the Examination of Water and Wastewater 24th Edition, 24th edn:
Lozier JC, Bankston A, Beaty J, Garcia-Aleman J, Scharf R, Amy G, Salinas Rodríguez SG (2009) Use
of advanced NOM characterization methods to trace the fate of organic contaminants from a
membrane backwash recycle scheme. Paper presented at the Membrane Technology Conference,
Memphis, Tennessee United States2009
Mangal MN (2023) Controlling scaling in groundwater reverse osmosis: minimizing antiscalant
consumption Veenman, The Netherlands. 978-90-365-5588-3. 10.3990/1.9789036555890
Mangal N, Salinas Rodriguez SG, Yangali Quintanilla VA, Schippers JC, Kennedy MD (2021) Ch 08 -
Scaling. In: Salinas Rodriguez SG, Schippers JC, Amy GL, Kim IS, Kennedy MD (eds) Seawater
Reverse Osmosis Desalination: Assessment and Pre-treatment of Fouling and Scaling, 1st edn:
207-242 10.2166/9781780409863_0207
Matilainen A, Gjessing ET, Lahtinen T, Hed L, Bhatnagar A, Sillanpää M (2011) An overview of the methods
used in the characterisation of natural organic matter (NOM) in relation to drinking water treatment.
Chemosphere 83: 1431-1442 DOI https://doi.org/10.1016/j.chemosphere.2011.01.018
Nahrstedt A, Camargo Schmale J (2008) New insights into SDI and MFI measurements. Water Science
and Technology: Water Supply 8: 401-412 DOI https://doi.org/10.2166/ws.2008.087
NEN 6271:1995 nl (1995) Bacteriological examination of water - Determination of the easily
assimilable organic carbon (AOC) content NEN, Netherlands https://www.nen.nl/en/nen-
6271-1995-nl-16389.
Peña N, del Vigo F, Salmerón O, Rodriguez J, Borrel A, Chesters SP (2015) Pre-coating of outside-
inside capillary UF membranes with iron hydroxide particles to limit non-backwashable
fouling during seawater algal blooms. Desalination and Water Treatment 55: 2973–2987 DOI
10.1080/19443994.2014.957950
Peña N, Rodriguez J, del Vigo F, Chesters SP (2022) 20 years of data from 500 seawater membrane
autopsies. Paper presented at the IDA World Desalination Congress, Sydney, Australia, 9-13
October 2022
Rachman RM, Ghaffour N, Waly F, Amy GL (2013) Assessment of Silt Density Index (SDI) as fouling
propensity parameter in Reverse Osmosis (RO) desalination systems. Desalination and Water
Treatment 51: 1091-1103 DOI doi:10.1080/19443994.2012.699448
Ross PS, Hammes F, Dignum M, Magic-Knezev A, Hambsch B, Rietveld LC (2013) A comparative
study of three different assimilable organic carbon (AOC) methods: results of a round-robin test.
Water Supply 13: 1024-1033 DOI 10.2166/ws.2013.079
Salinas Rodriguez SG (2021). Fouling and scaling in seawater reverse osmosis desalination. The Source:
4https://www.thesourcemagazine.org/fouling-and-scaling-in-seawater-reverse-osmosis-
desalination/
Salinas Rodríguez SG (2011) Particulate and organic matter fouling of SWRO systems: Characterization,
modelling and applications CRC Press/Balkema, Delft. http://dx.doi.org/10.1201/b11609
Salinas Rodriguez SG, Boerlage SFE, Kenedy MD, Schippers JC (2021a) Ch 04 - Particulate fouling.
In: Salinas Rodriguez SG, Schippers JC, Amy GL, Kim IS, Kennedy MD (eds) Seawater Reverse
Osmosis Desalination: Assessment and Pre-treatment of Fouling and Scaling, 1st edn: 85-124
10.2166/9781780409863_0085
Salinas Rodriguez SG, Kennedy MD, Schippers JC, Amy GL (2009) Organic foulants in estuarine and bay
sources for seawater reverse osmosis – Comparing pre-treatment processes with respect to foulant
reductions. Desalination and Water Treatment 9: 155-164 DOI 10.5004/dwt.2009.766

22
 Chapter 1

Salinas Rodriguez SG, Schippers JC (2021) Ch 01 - Introduction to desalination. In: Salinas Rodriguez
SG, Schippers JC, Amy GL, Kim IS, Kennedy MD (eds) Seawater Reverse Osmosis Desalination:
Assessment and Pre-treatment of Fouling and Scaling, 1st edn: 26 10.2166/9781780409863_0001
Salinas Rodriguez SG, Schippers JC, Amy GL, Kim IS, Kennedy MD (2021b) Seawater Reverse Osmosis
Desalination: Assessment and Pre-treatment of Fouling and Scaling, 1st edn IWA Publishing,
London. 9781780409856. 10.2166/9781780409863
Salinas Rodriguez SG, Sithole N, Dhakal N, Olive M, Schippers JC, Kennedy MD (2019) Monitoring
particulate fouling of North Sea water with SDI and new ASTM MFI0.45 test. Desalination 454:
10-19 DOI https://doi.org/10.1016/j.desal.2018.12.006
Schippers JC, Salinas Rodriguez SG, Boerlage SFE, Kennedy MD (2014). Why MFI is edging SDI as a
fouling index. Desalination & Water Reuse: 28-32
Schippers JC, Verdouw J (1980) The modified fouling index, a method of determining the fouling
characteristics of water. Desalination 32: 137-148
Servais P, Billen G, Hascoët M-C (1987) Determination of the biodegradable fraction of dissolved
organic matter in waters. Water Research 21: 445-450 DOI https://doi.org/10.1016/0043-
1354(87)90192-8
Shanmuganathan S, Nguyen TV, Jeong S, Kandasamy J, Vigneswaran S (2015) Submerged membrane
– (GAC) adsorption hybrid system in reverse osmosis concentrate treatment. Separation and
Purification Technology 146: 8-14 DOI https://doi.org/10.1016/j.seppur.2015.03.017
Simon FX, Penru Y, Guastalli AR, Esplugas S, Llorens J, Baig S (2013) NOM characterization by LC-
OCD in a SWRO desalination line. Desalination and Water Treatment 51: 1776-1780 DOI
10.1080/19443994.2012.704693
Sweity A, Oren Y, Ronen Z, Herzberg M (2013) The influence of antiscalants on biofouling of RO
membranes in seawater desalination. Water Research 47: 3389-3398 DOI https://doi.
org/10.1016/j.watres.2013.03.042
Sweity A, Zere TR, David I, Bason S, Oren Y, Ronen Z, Herzberg M (2015) Side effects of antiscalants
on biofouling of reverse osmosis membranes in brackish water desalination. Journal of Membrane
Science 481: 172-187 DOI https://doi.org/10.1016/j.memsci.2015.02.003
USEPA (2005) Membrane Filtration Guidance Manual. EPA 815-R-06-009 USEPA, Online, pp. 35
https://nepis.epa.gov/Exe/ZyPDF.cgi/901V0500.PDF?Dockey=901V0500.PDF.
USEPA (2010) Method 1664 Revision B: n-hexane extractable material (HEM; oil and grease) and
silica gel treated n-hexane extractable material (SGT-HEM; non-polar material) by extraction
and gravimetry USEPA, Online, pp. 35 https://www.epa.gov/sites/default/files/2015-08/
documents/method_1664b_2010.pdf.
van der Kooij D (1992) Assimilable organic carbon as an indicator of bacterial regrowth. J American
water works association 84: Van der Kooij, D. 1992. Assimilable organic carbon as an indicator of
bacterial regrowth. J. AWWA 1984(1992):1957-1965.
van der Kooij D, Hijnen W, Cornelissen E, Van Agtmaal S, Baas K, Galjaard G (2007) Elucidation of
membrane biofouling processes using bioassays for assessing the microbial growth potential of
feed water. Paper presented at the Membrane Technology Conference, Tampa Bay, Florida2007
van der Kooij D, Visser A, Hijnen WAM (1982) Determining the concentration of easily
assimilable organic carbon in drinking water. Journal AWWA 74: 540-545 DOI https://doi.
org/10.1002/j.1551-8833.1982.tb05000.x
Villacorte LO, Boerlage SFE, Dixon MB (2021) Ch 06 - Algal Blooms and RO desalination. In: Salinas
Rodriguez SG, Schippers JC, Amy GL, Kim IS, Kennedy MD (eds) Seawater Reverse Osmosis
Desalination: Assessment and Pre-treatment of Fouling and Scaling:

23
Experimental Methods for Membrane Applications

Villacorte LO, Ekowati Y, Calix-Ponce HN, Schippers JC, Amy GL, Kennedy MD (2015a) Improved
method for measuring transparent exopolymer particles (TEP) and their precursors in fresh and
saline water. Water Research 70: 300-312 DOI http://dx.doi.org/10.1016/j.watres.2014.12.012
Villacorte LO, Ekowati Y, Winters H, Amy G, Schippers JC, Kennedy MD (2015b) MF/UF rejection
and fouling potential of algal organic matter from bloom-forming marine and freshwater algae.
Desalination 367: 1-10 DOI https://doi.org/10.1016/j.desal.2015.03.027
Villacorte LO, Kennedy MD, Amy GL, Schippers JC (2009) The fate of Transparent Exopolymer
Particles (TEP) in integrated membrane systems: Removal through pre-treatment processes and
deposition on reverse osmosis membranes. Water Research 43: 5039-5052 DOI 10.1016/j.
watres.2009.08.030
Voutchkov N (2010) Considerations for selection of seawater filtration pretreatment system. Desalination
261: 354-364 DOI https://doi.org/10.1016/j.desal.2010.07.002
Vrouwenvelder JS, Beyer F, Dahmani K, Hasan N, Galjaard G, Kruithof JC, Van Loosdrecht MCM (2010)
Phosphate limitation to control biofouling. Water Research 44: 3454-3466 DOI https://doi.
org/10.1016/j.watres.2010.03.026
Vrouwenvelder JS, Manolarakis SA, Veenendaal HR, van der Kooij D (2000) Biofouling potential of
chemicals used for scale control in RO and NF membranes. Desalination 132: 1-10 DOI https://
doi.org/10.1016/S0011-9164(00)00129-6
Vrouwenvelder JS, van der Kooij D (2001) Diagnosis, prediction and prevention of biofouling of
NF and RO membranes. Desalination 139: 65-71 DOI https://doi.org/10.1016/S0011-
9164(01)00295-8
Waly TKA (2011) Minimizing the use of chemicals to control scaling in SWRO - Improved prediction
of the scaling potential of calcium carbonate CRC Press / Balkema, Leiden. 978-0-415-61578-5.
Waly TKA, Kennedy MD, Witkamp G-J, Amy G, Schippers JC (2012) The role of inorganic ions in the
calcium carbonate scaling of seawater reverse osmosis systems. Desalination 284: 279-287 DOI
http://dx.doi.org/10.1016/j.desal.2011.09.012
Weinrich L, LeChevallier M, Haas C (2015) Application of the bioluminescent saltwater assimilable
organic carbon test as a tool for identifying and reducing reverse osmosis membrane fouling
in desalination Water Reuse Research Foundation, Virginia, USA https://www.waterrf.org/
research/projects/application-bioluminescent-saltwater-assimilable-organic-carbon-test-tool
Weinrich LA, Schneider OD, LeChevallier MW (2011) Bioluminescence-based method for measuring
assimilable organic carbon in pretreatment water for reverse osmosis membrane desalination. Appl
Environ Microbiol 77: 1148-1150 DOI 10.1128/aem.01829-10
Westerhoff P, Chen W, Esparza M (2001) Fluorescence Analysis of a Standard Fulvic Acid and Tertiary
Treated Wastewater. Journal of Environmental Quality 30: 2037-2046 DOI https://doi.
org/10.2134/jeq2001.2037
Wilf M, Klinko K (2016) Effect of new pretreatment methods and improved membrane performance
on design of RO seawater systems Hydranautics, Online, pp. 1-17 https://membranes.com/
wp-content/uploads/Documents/Technical-Papers/Application/SWRO/Effect-Of-New-
Pretreatment-Methods-And-Improved-Membrane-Performance-On-Design-Of-RO-Seawater-
Systems.pdf.
Yin W, Li X, Suwarno SR, Cornelissen ER, Chong TH (2019) Fouling behavior of isolated dissolved
organic fractions from seawater in reverse osmosis (RO) desalination process. Water Research 159:
385-396 DOI https://doi.org/10.1016/j.watres.2019.05.038
Zhang M, Jiang S, Tanuwidjaja D, Voutchkov N, Hoek EMV, Cai B (2011) Composition and Variability of
Biofouling Organisms in Seawater Reverse Osmosis Desalination Plants. Appl Environ Microbiol
77: 4390-4398 DOI doi:10.1128/AEM.00122-11

24
 Part 1
Membrane processes
 doi: 10.2166/9781789062977_0027

Chapter 2

Microfiltration and
ultrafiltration
Morten Lykkegaard Christensen, Aalborg University, Denmark

Guillem Gilabert-Oriol, DuPont Water Solutions, Spain

The learning objectives of this chapter are the following:

• give an understanding of how microfiltration and ultrafiltration is used for water

and wastewater treatment

• give knowledge of how membranes system is best operated to reduce membrane

fouling and ensure high flux

• give an introduction to different cleaning options and how hydraulic cleaning

methods are optimized

• give knowledge of how microfiltration and ultrafiltration system are designed and
proper membrane is selected.

2.1 INTRODUCTION

Microfiltration and ultrafiltration are pressure-driven membrane filtration processes, where

a transmembrane pressure (TMP) is used to press water through the membrane. They are
operated at relatively low TMP compared with the other pressure-driven membrane
processes i.e., nanofiltration and reverse osmosis (Figure 1). Micro- and ultrafiltration
membranes are porous membranes and the separation based on sieving effects (size
exclusion). This means that large particles or macromolecules are rejected by the membrane,
whereas small molecules pass the membrane.

Microfiltration membrane typically have pores larger than 0.1 μm whereby bacteria and
particles can be removed by the membrane (Figure 1). Often a nominal pore size is given
from the manufacturer, but the performance of the membrane will also depend on pore size

© 2024 The Authors. This is an Open Access book chapter distributed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0),
(https://creativecommons.org/licenses/by-nc-nd/4.0/). The chapter is from the book Experimental
Methods for Membrane Applications in Desalination and Water Treatment, Sergio G. Salinas-Rodriguez,
Loreen O. Villacorte (Eds).
Experimental Methods for Membrane Applications

distribution and membrane materials. Microfiltration is often used as pretreatment prior to

other membrane filtration processes such as RO, electrodialysis, and membrane distillation.
Ultrafiltration membranes have pores sizes below 0.1 μm and retains macromolecules
(Figure 1). Ultrafiltration membranes are used for removal of colloids, proteins, virus.
Molecular weight cut-off (MWCO) is used to describe the membrane, instead of pore size.
MWCO is the lowest molecular weight where more than 90% of the macromolecules are
rejected by the membrane.

Size Molecular weight Examples Process Pressure

100 µm Pollen

Starch
10 µm
Microfiltration 0.2 – 2 bar
Blood cells

Bacteria
1 µm

Latex emulsions

1000 Å

100,000 Albumin
100 Å
Ultrafiltration 1 – 10 bar
10,000 Pepsin

1,000 Vitamin B-12

10 Å

Glucose Nanofiltration 5 – 20 bar

Water
1Å RO 10 – 150 bar
Na+ Cl-

Figure 1 Pressure driven filtration processes

2.1.1 Advantages of ultrafiltration compared to conventional treatment

Micro- and ultrafiltration offers several advantages compared with conventional filtration
methods. As an example, ultrafiltration used as a pretreatment step to treat water has
experienced an impressive increase as a result of the continuous search for cost-effective
technologies which enable a sustainable production of water (Chu et al., 2009). Key benefits
associated to the ultrafiltration technology versus conventional pretreatment are a low
footprint, the ability to remove virus and bacteria and to significantly reduce colloids,
suspended particles, turbidity and some total organic carbon. Even more importantly, the
ability to reliably provide good quality filtrate water to the downstream reverse osmosis are
the most remarkable benefits associated with this technology (Mourato et al., 2003).

28
 Chapter 2

As another example Milwaukee, on the United States of America suffered back in 1993 on
of the worst modern water-born epidemies due to Cryptosporidium microscopic parasite
causing diarrhea. Cryptosporidium is known for being highly resistant to chlorination.
This can pose a risk when conventional water treatment schemes based on sand filter and
chlorination are used (Morris et al., 2005).

Since then, ultrafiltration has emerged as a preferred technology to treat drinking water, thus
replacing old water treatment schemes based on sand filters and chlorination. Ultrafiltration,
as it is based on an absolute pore-size filtration, can eliminate Cryptosporidium or other
chlorine-resistant organisms, as they cannot pass through the membrane (Pressdee, 2005).
Additional benefits of ultrafiltration versus sand filters are summarized in the table 1.

Table 1 Summary of the benefits of ultrafiltration over sand filters (DuPont, 2022)
Sand Filtration Ultrafiltration
Pathogenic bacteria removal (coliforms) ≤2 log ≥ 4 log
Pathogenic virus removal (enteric) ≤ 3 log ≥ 4 log
Water effluent quality (Turbidity) 0.1 NTU 0.01 NTU
Organics removal (TOC) 5% 34%
Lower Silt Density Index 3.2 2.6
Footprint reduction No 99%
Improve plant availably (reduce downtime) 95% > 98%
Operate reverse osmosis at higher flux < 14 L/m²h > 14 L/m²h

Most of the advantages of ultrafiltration rely on the small pores for water transport, typically
in the range of 20-30 nm. This allows the system to be significantly more compact, reducing
the required footprint by up to 99% compared to a conventional sand filter. Additionally,
thanks to this smaller pore size, it is very challenging for bacteria, parasites, protozoa and
even viruses to pass through the membrane. This greatly improves the water effluent
quality produced. When ultrafiltration is used as a pretreatment to the reverse osmosis, this
means that the reverse osmosis system can operate at a higher flux, as particulate fouling risk
is eliminated.

2.2 DESIGN AND OPTIMIZE MEMBRANE PROCESSES

Before setting up a micro- and ultrafiltration process, the following five points must be
considered (Raghunath et al., 2012)
1) Objective of the process and the success criteria
2) Pre-treatment of feed
3) Membrane selection
4) Module selection
5) Operation parameter optimization

29
Experimental Methods for Membrane Applications

It is usually not possible to follow the points chronologically, as e.g., a proper pre-treatment
depends on choice of membrane, membrane module and operational parameters. In the text
the objectives of the membrane will be discussed first, then membrane selection, operation,
membrane module and finally pre-treatment.

2.3 OBJECTIVE OF THE FILTRATION PROCESS

Ultra- and microfiltration are used in many industries, at water facilities and wastewater
plants. The objective of the process varies, and the relevant stream is not the same for all
filtration processes but can be
1. Concentrate e.g., for recovery of macromolecules,
2. Filtrate e.g., for removal of pathogenic microorganism from drinking water
3. Both concentrate and filtrate e.g., for harvesting valuable product from waste stream by
fractionation of macromolecules.

Feed Concentrate
Qf ,Cf Qr ,Cr

Filtrate Qp ,Cp

Figure 2 Membrane filtration process and mass balance.

This is important because it affect the optimal choice of membrane, modules, and operation
parameters. In order to define the success criteria different key equations and parameters
will be defined.

The optimal design of the membrane process depends on the flow and concentrations
of the inlet, the wanted flow and concentration of the outlets, and the operation time of
the membrane system (Figure 2). Q is the volumetric flow and C is the concentration of
the compounds of interest. A key factor for membrane operations, is most particles or
macromolecules (rejected materials) are concentrated in the concentrate stream, which can
be found by defining a concentration factor.

Qf Eq. 1
CF =
Qr

The concentration factor should be high and concentration factors up to 10 is typically set as
the goal. The concentration factor gives a maximum value for how much the product can be
concentrated and if the particles are fully rejected by the membrane, Cr = CF×Cf.
Another key parameter is the yield (Y)

Qr Cr Eq. 2
Y=
Qf C f

30
 Chapter 2

i.e., fraction of the wanted product that are left in the concentrate. The yield should be high
and ideally 1.

Filtrate may be the main product stream i.e., if the membrane is used for water treatment. CF
must then be high as well because as much water as possible have to be recovered, but often
the recovery is calculated instead, as it represents the water that is being produced compared
to the water that is being fed into the ultrafiltration membrane

Qp 1 Eq. 3
%Recovery = =1
Qf CF

Another important parameter is availability. Availability represents the amount of time, in

percentage, that the ultrafiltration is operating. Availability is defined as

toperation Eq. 4
% Availability =
toperation + tstopped

This is an important factor, since as an example, if the membrane needs a lot of time to be
properly cleaned, the overall ultrafiltration efficiency will be low.

%Efficiency = % Recovery % Availability Eq. 5

Filtrate flux (Jw) across a membrane can be described as the filtrate flow (Qp) normalized by
the membrane’s active area (A). Its units are typically expressed as liters/(m2·h) or LMH, or
as US gallons/(ft2·day) or GFD.

Op Eq. 6
Jw =
A

One useful parameter to calculate is the net flux (Jnet). This represents the net flux a micro-
and ultrafiltration installation is producing in a whole year. This takes into consideration
the time installation is operating after discounting the time the plant is stopped, and the net
water it produced after discounting the water that was produced but later consumed during,
for example, cleanings. It can be calculated by multiplying the design flux by the efficiency

Jnet = J w %Efficiency Eq. 7

This parameter is important since sometimes, in order to optimize the plant throughput
production, it might be, for example, more beneficial to increase the design flux while
decreasing a bit the recovery and availability, if this can help getting an overall higher net
flux. The net flux can be used to calculate the required membrane area

31
Experimental Methods for Membrane Applications

2.4 MEMBRANE TYPES

The selection of the membrane is important for the filtration process, and different thing
must be addressed. It is important to have a high flux through the membrane and a high
rejection of particles and macromolecules. However, high rejection usually results in
lower flux or higher energy demand, so often these two key parameters must be balanced
depending on the propose of the filtration process. It is important to have a membrane with
minimal risk of fouling or clogging, and a membrane with high lifetime, which depends on
the composition of the membrane, surface properties and pore size.

Different types of membrane modules exist, such as flat sheet membranes (e.g., 1×1 m and
200 μm thick), spiral wound membranes, hollow fiber membranes and tubular membranes.
Module length is typically around 1 m and the membrane area for one module can for large
installation be up to 40-80 m2.

Ultrafiltration and microfiltration can be operated either outside-in, or inside-out,

depending on the type of features that is mainly desired. Another classification criteria
depends on whether all the water that goes into the membrane is being treated, this is
called dead-end filtration; or if a portion of the water is not being treated, it is called cross-
flow filtration. Cross-flow filtration has the advantage of better managing fouling and
lower energy consumption, but at the expense of a reduction on water recovery. Dead-end
filtration has the advantage of having a higher water recovery and a smaller footprint, but
fouling is more difficult to control.

Outside-in membranes are hollow fibers characterized by having the water to be treated
outside the fiber, and the filtrated water collected inside the hollow fiber. These are typically
made of Polyvinylidene fluoride (PVDF). They are usually operated in dead-end mode. Key
features of this technology involve a higher resistant to chemicals, and the ability to perform
air scouring during cleanings.

Inside-out membranes are hollow fibers characterized by having the water to be treated
outside the fiber, and the filtrated water collected inside the hollow fiber. These are typically
made of Poly(ether-Sulfone) (PES). They are usually operated in dead-end mode. Key
feature of this technology is a higher membrane permeability.

Additionally, ceramic membranes are also used in the industry. Ceramic membranes are
usually made of Aluminum, Silica, Titanium, and Zirconium oxides. They are typically
operated in cross-flow filtration (Gruskevica and Mezule, 2021). One of the key challenges
of ceramic membranes are its cost, and managing fouling properly. Key features of this
technology are its thermal and chemical stability, as well as being able to operate at higher
fluxes (Sondhi et al., 2003).

Membranes system can also be classified as submerged and pressurized system. Submerged
membranes have the advantage that they can be visually observed, and they usually operate
at lower energy. Pressurized membranes have the advantage of operating at higher fluxes, a
lower footprint (Nick, 2019) , but might suffer from higher fouling rates, as a possible result
of higher compaction of the fouling layer. (Kim et al., 2015)

32
 Chapter 2

The use of coagulation in microfiltration and ultrafiltration can be used as a way to reduce
fouling rates and increase overall flux (Konieczny et al., 2009). Coagulants are typically iron
or aluminum based.

2.5 BASIC EQUATIONS

The filtrate flux (Jp) must be high and is a function of permeability and transmembrane
pressure (TMP)

TMP
J p = K TMP = Eq. 8
Rm

where K is the permeability (or more correctly the permeance) in m/(s·Pa), Rm is the
hydraulic resistance of the membrane in m-1, and η is the viscosity in Pa·s.

For filtration of pure water, the permeability is only dependent on membrane properties
and can be calculated from membrane thickness, pore numbers and size.

i
(ni di4 ) Eq. 9
K=
128 ητl

Where n is number of pores per square meter, di is pore diameter, or hydraulic diameter for
non-cylindrical pores (m), τ is tortuosity (-) and l is membrane thickness (m).

The ideal membrane is thin, to ensure high permeability, but thick enough to withstand
the transmembrane pressure. To improve permeability but still ensure the mechanical
properties asymmetric membrane or thin-film composite membrane are usually used
(Figure 3). Composite membrane consists of a thin active layer, where the separation
happens, and a support layer with low hydraulic resistance.

Asymmetric membranes Thin film composite

Figure 3 Anisotropic membranes

Notice that the number of pores per square meter is another important parameter and
implicit given from the porosity of the membrane.
Besides flux, membrane rejection (R) is important and is defined as

Cp
R =1 Eq. 10
Cr

where R is a number between 0 and 1, where all particles are rejected by the membrane if
R = 1 whereas the particles can freely pass the membrane if R = 0

33
Experimental Methods for Membrane Applications

For some applications , it is important that some molecules pass the membrane and other are
retained by the membrane. The selectivity (α) can then be calculated and used to compare
different type of membranes

1 Ri Eq. 11
i/ j
= and i/ j
1
1 Rj

Where i is the particle that should pass the membrane, and j is the particles that should be
rejected by the membrane. Several particles or macromolecules are partly rejected by the
membrane. The reason for this is that pores are not unisized, but a large pore size distribution
are often observed (Figure 4).

0.25

0.20

0.15

0.10

0.05

Distribution frequency 0
0.01 0.1 1.0 10 100
Pore size (µm)

Figure 4 Pore size distribution microfiltration membranes (Data obtained from Liu et al., 2018)

For water treatment and sterilization, it is important to ensure that bacteria cannot pass the
membrane. Here the size of the largest pores is critical. A method to determine the size of
the largest pores or ensure that all pores are below a given size is the bubble point method
(breakthrough pressure). An alternative method is the water intrusion procedure, which is
described elsewhere (see ‘water intrusion procedure’).

For wetted membranes, the air pressure must overcome the capillary pressure of the pores,
before liquid can pass the pore (Figure 5).

Experimental setup Individual pore

Water
Wetted membrane

N2

Figure 5 Bubble point method

34
 Chapter 2

The pressure required to overcome the capillary pressure is called breakthrough pressure
(P*)

4γ Eq. 12
P* =
d

where γ is the surface tension of the liquid. The surface tension for water is 0.072 N/m.
For ultrafiltration membrane it is more complicated to measure pores size, instead cut-off
values are often determined instead. MWCO can be determined by filtering mixtures of
macromolecules and measuring rejection of the molecules (Figure 6). It is thereby possible
to determine the size of the molecules where more than 90% of the molecules are retained
by the membrane.

100

80

60

40

20
cut-off
Rejection (%) 0
0 50 100 150 200 250 300 350 400 450
Molecular weight (kDa)

Figure 6 Molecular weight cut-off of PES/PSFNA membrane measured by using a mixture of PEG
molecules (Data obtained from Wu et al., 2018)

The selection of the membrane depends on feed composition and required quality of
permeate or molecules that must be concentrated. The following parameters for the
membrane must be considered

1. Selectivity (pore size and cut-off) so the membrane reject the particle that should be
concentrated or removed
2. Permeability (high flux through the membrane at low pressure)
3. Thermal stability
4. Mechanical properties
5. Chemical stability (resistance for cleaning process)
6. Membrane composition and surface properties (to avoid adsorption of foulant)

2.6 NORMALIZATION

Filtrate flow permeating through a micro- and ultrafiltration membrane is heavily influenced
by the water temperature. This happens as the process of water moving through convection
across the membrane pores is facilitated by a higher temperature, as water then becomes less

35
Experimental Methods for Membrane Applications

viscous, and less energy (TMP) is required for water to pass through the pores. On the other
side, a lower water temperature increases the water viscosity and therefore more energy
(TMP) is required for water to pass through the membrane pores.

Therefore, data normalization is of utmost importance in microfiltration and ultrafiltration

membranes to properly assess the performance of a membrane-based installation. As an
example, if data would not be normalized, one could observe an increase over time of raw
TMP. This could induce the plant engineer to think that there might be irreversible fouling
developing over time on the membrane system. If then this data would be operated, one
could see that this TMP increase over time is due to a decrease in temperature over time
as a result of moving to winter season. After normalizing this data considering water
temperature, it could be observed that TMP evolution over time, thus dismissing any
fouling issue.

The normalized TMP (TMP*) is determined by multiplying the raw TMP by the temperature
correction factor (TCF)

TMP* = TMP TCF Eq. 13

The purpose of the TCF is to take into consideration the effect of the Temperature (T) in
Celsius degrees and its influence on the viscosity of water, and normalizing its value from
the temperature the water has, to an ideal temperature of 25 ºC (Daucik and Dooley, 2008)
247.8
25 + 273.16 140 Eq. 14
TCF = 10 247.8
T + 273.16 140
10
It should be noticed that filtrate flux can also be normalized. This can be especially important
if a system is operated at constant TMP, as then Flux will change over time. It might also be
relevant in order to assess the stress that the ultrafiltration is operated, as the higher the flux,
the more stressed the system will be, and the more difficult it would be to be able to operate
the system under a sustainable way

TMP* = TMP TCF Eq. 15

Special attention should be put in order to not normalize the filtrate flux twice, as normalize
filtrate flux needs to be calculated using a non-normalized TMP, and then applying the TCF
in the equation described above.

2.7 MEMBRANE FOULING

Membrane fouling is an inevitable phenomenon in membrane filtration. Membrane

filtration occurs when particles are deposited on a membrane surface or in membrane pores
in a process. Fouling reduces the permeability of the membrane and thereby increase the
required pressure to keep the flux constant. Further, fouling may change the selectivity

36
 Chapter 2

of the membrane. This may in some cases be beneficial i.e., it may result in better water
treatment. However most often all type of fouling must be minimized, and different
methods exists to reduce fouling build up or remove already formed fouling. Microfiltration
is typically operated at low pressure (up to some bars) or flux. Too high flux increases the
risk of accumulation of materials at the membrane surface. Thus, pressure must be kept
low to avoid too high flux and thereby reduce fouling risk. Ultrafiltration membranes is
operated at higher pressure to obtain same flux through membrane due to the smaller pores
and higher hydraulic resistance. Micro- and ultrafiltration membrane are typically cleaned
hydraulically or chemically. Cleaning reduces the operation time of the membrane whereby
the availability of the membrane is reduced cf., Eq. (4).

Membrane fouling can be due to internal fouling such as adsorption of materials in the pores
or pore blocking, and it can be external such as gel formation, precipitation of salts (scaling)
and biofilm growth. The type of fouling is important for the optimal operational parameter,
cleaning strategy and different method exists to analyses fouling (later chapters). Besides
direct analysis of the fouling materials and the membrane, flux-pressure data can give some
hint of the type of fouling and is useful to optimize operational parameters such as flux,
pressure and cross-flow.

If flux-pressure data is used to quantify fouling, the resistance-in-series model is usually

used to determine individual contributors to resistance (Géasan-Guiziou et al., 1999) . One
method is to separate fouling into reversible fouling (Rrev) or irreversible fouling (Rirrev).
Reversible fouling can be removed by hydraulically cleaning whereas irreversible fouling
cannot.

For fouled membrane, the water flux through the membrane can be described as
P
Jp = Eq. 16
η i
R

Where the resistance is the sum of individual contributors to the hydraulic resistance

∑ R= R
i m
+ Rirrev + Rrev Eq. 17

Membrane resistance can be determined for a pristine membrane by filtering clean water
(i.e., RO treated water)
P
Rm = Eq. 18
ηJ w,0

Determination of membrane resistance is critical to determine the other resistances

Irreversible fouling can be determined by filtering the suspension, clean the membrane and
measure the water flux after cleaning.

37
Experimental Methods for Membrane Applications

P Eq. 19
Rirrev = Rm
ηJ w,cleaned

Irreversible fouling is often ascribed to internal fouling i.e., pore blocking, and adsorption,
and is problematic as it results in a permanent reduction of the membrane permeability.
Further, it is difficult to avoid by changing operation condition of the filtration process. It
depend on the type of membrane chosen i.e., pore size distribution and composition. Thus,
if the irreversible resistance is too high, another membrane may be chosen, or feed must be
pretreated prior to the filtration.

Reversible fouling usually increases continuously during filtration and can be calculated as
P
Rrev = Rm Rir Eq. 20
ηJ w

The reversible fouling strongly depends on the operational conditions and will be discussed
more deeply in next section.

2.8 SUSTAINABLE FLUX

Most filtration processes are operated at constant flux to have a stable output, i.e., a constant
cross-flow (CF). The flux must be set so the energy cost is low, and the required membrane
are is low. Often the system is operated so build-up of reversible fouling is low. Reversible
fouling is a result of external fouling and strongly dependent on the transport of particles
and molecules to the surface of the membrane. High flux increases the transport of material
to the surface and thereby increase fouling build-up. At low pressures, material transport to
the surface is low, and flux increase almost linearly with pressure also called the pressure-
controlled region (Figure 7). At high pressure, the effect of increasing the pressure becomes
less important for the flux. If pressure is increased, flux increase as well but decline to a lower
steady state flux. Thus, performance of the membrane process can no longer be improved
by increasing the pressure (pressure-independent region). Turbulence and high shear at the
membrane surface can remove part of the external fouling layer and thereby reduce fouling
build-up. A common method is to apply a high crossflow (up to 2-3 m/s) to reduce fouling
built up and increase flux. Besides high crossflow shear can be induced by using turbulence
promoters, rotating, or vibrating membrane or air scouring. If the concentration in the feed
is low, dead-end, or semi-dead-end setup can be used to lower the energy consumption.
At higher concentration feeds crossflow are required and for high viscous feed, rotating
membrane are may be the most energy efficient solution.

A critical flux has been defined as the transition from the pressure-controlled region to the
pressure independent region (Cleck et al., 2003, Bacchin et al., 2009). In the ideal case, the
filtrate flux follows that of clean water until the critical flux, but often the filtrate flux is
lower than the pure water flux due to irreversible fouling. The optimal flux is typical lower
than the critical flux; a number around 75% of the critical flux can be used (Raghunath et al
2012).

38
 Chapter 2

Water flux

Increased crossflow
Critical flux

Pressure-controlled region Pressure-independent region

Permeate flux (LMH)
Transmembrane pressure

Figure 7 Flux-pressure curve and definition of critical flux

For long operation, membrane performance will usually decline also if the flux is lower than
the critical flux. An alternative to the critical flux, is therefore the concept of sustainable flux.
The sustainable flux is the maximum flux at which the fouling rate is acceptable and can be
handled by hydraulic and chemical cleaning. At constant flux the fouling rate (FR) can be
defined as

ΔTMP Eq. 21
FR =
Δt

By assuming that the transmembrane pressure increases linearly with time. Notice at high
flux, a TMP jump may be observed where the fouling rate increases with time. At lower flux,
it is usually reasonable to assume a linear relationship between TMP and time at constant
flux. The sustainable flux can be determine using the flux step method, where the flux is
gradually increased, and FR measured. The sustainable flux is then defined as the point
where FR is exceeded a given threshold value for the fouling rate.

Threshold
fouling rate

Flux Fouling rate

Time Flux

Figure 8 Step-flux test for determination of sustainable flux (Modified from Wang et al., 2014)

39
Experimental Methods for Membrane Applications

Critical and sustainable flux is higher increased at higher cross-flow i.e., for microfiltration of
lactalbumin suspension, the critical flux increase a factor of three, when the cross-flow was
increased from 0.5 to 4-5 m/s (Vyas et al., 2002). However, a high crossflow is costly; thus
there is a balance with high flux vs. high cross-flow. Secondary effluent from wastewater
treatment have been treated by membrane filtration, and the membrane have been tested
at different flux and cross-flow. The optimal crossflow has been determined to 1.5 m/s for
polymeric membranes and 4.5 m/s for ceramic membrane as the ceramic membranes was
more expensive than polymeric one (Owen et al., 1994).

2.9 MEMBRANE DESIGN AND MODULE

A typical membrane setup is the feed-and-bleed process (Figure 9). The system consists
of a feed pump that ensure the transmembrane pressure and a constant flow of feed to
the system. The recirculation pump is added to ensure a high crossflow at the membrane
surface. The crossflow can be up to 2-3 m/s. At higher crossflow, the energy consumption
increases rapidly. Due to the high crossflow, the pressure drop along the membrane module
may decline significantly and in worst case, the transmembrane pressure becomes negative
at the end of the module. The permeate flux is usually kept constant and the transmembrane
pressure regulated to ensure a constant flux.

Retentate Valve 1

Qr ,Cr

Qp ,Cp
Membrane

Valve 2 Permeate

Valve 3

Qf ,Cf
Chemicals Backwash tank
Feed
Feed Prefilter Recirculation
pump pump

Figure 9 Feed-and-bleed filtration set-up.

For a feed-and-bleed system, the concentration of particles or macromolecules in the

retentate can then be calculated as

CF Eq. 22
Cr = C f
1− (1− R)(CF −1)

Where Cf and Cr is the concentration of particles in the feed and retentate, respectively. CF is
the concentration factor and given as the ratio between the feed flow (Qf) and the retentate
flow (Qr). If the particles are fully retained by the membrane, then R = 1 and Cr = Cf × CF .

40
 Chapter 2

The yield can be calculated as

1
Y= Eq. 23
1− (1− R)(CF −1)

2.10 PRETREATMENT

Pre-treatments are often required to prevent fouling, and is a necessary step for most
membrane filtration process. Often bigger solids must be removed. In this case,
macrofiltration is used. Its aim is to reduce big solids such as branches and leaves, that
could otherwise potentially clog the ultrafiltration membranes and reduce its effectivity.
Different methods exist, being the most typical screening, the use of hydro-cyclones,
prefiltration using cartridge filters or multimedia filters. Suspended particles and colloidal
particles can be problematic in filtration as such particles are difficult to remove from the
membrane surface. Thus, it may be necessary to flocculate particles before the pre-filtration.
Flocculation is usually done by adding salts (ferric, aluminum or poly-aluminum salts),
polymers or combinations, whereby particles aggregates and can be removed by filters.
As an example, coagulation and prefiltration have been used for treatment of raw water: a
coagulation using iron coagulant (FeCl3) with anionic polyelectrolyte in the first step and
aluminum coagulant in the second one was used before raw water enters the pre-filtration
(Sakola and Konierczny, 2004). The pre-treatment reduces the negative effect of membrane
fouling, but also improved the quality of the treated water (Sakola and Konierczny, 2004)

2.11 CLEANINGS

Microfiltration and ultrafiltration is a typically a semi-batch process, specially for dead-end

filtraiton applications. This is because it mainly handles particulate fouling, it gets fouled
quickly. This means that the system needs to operate for 20 to 90 min depending on the
water type and undergo a cleaning to restore its permeability. Optimizing cleanings remains
thus of outmost importance to maximize the efficiency of the filtration system (Gilabert-
Oriol, 2021). The main type of fouling that micro- and ultrafiltration membrane experience
is detailed below. The most frequent one is particulate fouling. This is the most common
and occurrent one. As the main task of an ultrafiltration membrane is to remove particles
represented by total suspended solids (TSS) that give turbidity to the water. As water is
filtrated through the membrane, it gets clogged by particles. Therefore, is it of utmost
importance to perform a backwash to the membrane, so these particles can get detached
from the membranes and its permeability can be restored.

A backwash (BW) typically consists of multiple steps. These steps can consist of an air scour,
which blows air across the membranes, and shakes them to create abrasion and detach foulants
accumulated on the top of the fibers. Later, a draining can be done to empty the module and
remove the detached foulants. Then, a backwash top can be performed, together with an
Air Scour. The main purpose of the backwash is to removed particles that are blocking the
pores. Water with foulants is removed through the concentrate side of the module located
at the top part of it. A backwash bottom follows the same approach, but water with foulants
is evacuated through the bottom part of the module. It is not practical to perform an air scour

41
Experimental Methods for Membrane Applications

in this step, as air is blown from the bottom of the element, and this would interfere with
the aeration which is also blown from the bottom of the module. Finally, a forward flush
is performed in order to provide a shear force and remove any remaining foulant from the
top of the membrane. Key steps in this backwash process are performing a backwash top
with and air scour, followed by a forward flush. By optimizing the duration of each one of
these sub steps, as well as the frequency of backwashes, the overall efficiency of the whole
ultrafiltration treatment can be greatly optimized (Gilabert-Oriol et al., 2021).

A chemical enhanced backwash (CEB) consists of a typical Backwash sequence, where

sodium hypochlorite (NaOCl) is dosed. As filtration cycles are done, bacteria start to grow
and develop a biofilm. This biological fouling, or biofilm, is noted typically after 1 to 2 days
of operation. It can be assessed as after certain number of backwashes, the TMP cannot be
recovered to its initial values.

A cleaning-in-place (CIP) consists of a tailor-made chemical cleaning. It is typically

performed every 3 to 4 months. As the membranes get cleaned on a regular basis, very
specific foulant specific to the water type being treated starts to slowly build up over time.
Typical CIPs performed are a caustic CIP, used to remove organics being accumulate on the
membrane, where NaOH is used; and acid CIPs, used to remove scaling or metallic-based
foulants, where HCl or oxalic acid is used.

2.11.1 Optimization of hydraulic cleaning

In order to remove fouling, hydraulic cleaning strategies can be used such as backwash,
backflush or relaxation. Relaxation is done by closing the valve 1 and maybe valve 2 at the
permeate side in Figure 9, whereby the transmembrane pressure drops to zero. Backwash
or backflush are done by opening valve 3 and pump permeate through the membrane in
Figure 9. Backwash and relaxation are typically done at regular intervals after 30-90 min
and the duration is typical 1-3 min. Backflush are shorter cycles e.g., 30-60 s flush every 15
min. Regular hydraulic cleaning removes the reversible fouling. The efficiency of backwash,
backflush or relaxation can be calculated as
TMPf TMP0
h= Eq. 24
TMPf TMPi

Where TMPi is the pressure at the beginning of the filtration cycle, TMPf at the end of the
filtration cycle and TMP0 after hydraulic cleaning.

Another method to find the best strategy for hydraulic cleaning is to calculate the net flux
(Jnet) defined in Eq. (9). An example of an optimization of relaxation time for a membrane
bioreactor shows that the optimal relaxation time is 3 min (Figure 10). Relaxation have been
used in a membrane bioreactor setup used to treat municipal wastewater. For example for
MBR systems with flat-sheet membrane, because backwash is not possible for these types
of membranes as it will destroy the membranes. Thus, relaxation have been used as a gentle
cleaning method to keep the flow high.

42
 Chapter 2

300 70
Optimum average flux
250 60

200
50
150

Net flux (LMH)

40
100
Flux (LMH)

30
50

0 20
0 1 2 3 4 5 0 5 10 15 20 25 30
Time (h) Relaxation time (min)

Figure 10 (a) Relaxation experiment at constant pressure during operation. (b) Net flux was
calculated including both operation time and relaxation time in calculation. Optimum flux
was obtained after 3 min. (Christensen et al., 2016) The process was operated at constant
transmembrane pressure.

At long term operations, the performance of the membrane will usually decline, and
hydraulic cleaning is not sufficient. In these cases, it is necessary to use chemical to clean
the membrane, one method is to add chemical to the backwash water known as chemically
enhanced backwash (Figure 9), but the membrane can also be chemical cleaned from the
feed side.

2.12 MEMBRANE CASCADES

Membrane system are designed to minimize the required membrane area. Membrane
fouling are more pronounced at higher dry matter concentration of the feed. Further, in feed
and-bleed system the concentration of feed near the membrane surface is almost the same
as in the retentate. Thus, it will sometimes beneficial to operate more membranes in series.
Modules in series reduce membrane area demand and lower energy consumption.

Stage 1 Stage 2 Stage 3

Figure 11 Feed-and-bleed in series to reduce required membrane area

An example is given for microfiltration of milk assuming the filtrate flux is given as

L L
J p = A+ B ln(CF ) = 40 2
−14 2 ln(CF )
mh mh

43
Experimental Methods for Membrane Applications

Where the flux decreases with increasing dry matter content in the feed and therefore with
the concentration factor (CF).

Table 2 Three stage feed-and-bleed system to treat 5000 L/h feed and concentrate it with a factor
of 5
Stage CF Qp (L/h) Jw (L/(m2h)) Area (m2)
1 1.5 1,667 34.3 48.0
2 2.5 1,333 27.2 49.1
3 5.0 1,000 17.5 57.2
Sum 5.0 4,000 154.9

Assuming a feed flow of 5000 L/h, rejection of SS is 1, and a required CF of 5, a single-stage

feed-and-bleed system will result in a permeate flow of 4000 L/h and filtrate flux of 17.5 L/
(m2h). Hence the required membrane area is 229.0 m2. This membrane area can be reduced
to 154.9 m2 for at three-stage feed-and bleed system (Table 2). This reduces the required
membrane area with more than 30%.

2.13 SUMMARY

Microfiltration and ultrafiltration are used to treat water, as well as for concentration of dry
matter or macromolecules. Membranes are selected based on their rejection, but it is also
important to consider the risk for membrane fouling, and the chemical and mechanical
stability of the membrane. Adsorption of molecules or pore blocking can sometimes be
avoided by selecting an alternative membrane.

Most operation is done at constant operating flux, since most plants are designed to produce
a certain amount of treated water per day. When designing the installation, it is important
to set up the operate at a flux below the sustainable flux. This helps to avoid a too steep
increase on trans-membrane pressure, and helps overall operating the microfiltration and
ultrafiltration plants in a sustainable way. Higher permeate flux can be obtained with a
high crossflow at the membrane surface, typical up to 2-3 m/s, but it also increases energy
consumption and increase cost. Membrane processes can be done as multi-stage operation
to reduce the required membrane area.

44
 Chapter 2

2.14 REFERENCES

Bacchin P. Aimar P., Field R.W. (2006) Critical and sustainable fluxes: Theory, experiments and
applications. Journal of membrane science 281, 42-69.
Christensen ML., Bugge TV. Hede BH, Nierychlo M. Larsen P. Jørgensen M. K. (2016) Effects of
relaxation time on fouling propensity in membrane bioreactors. Journal of Membrane Science.
504, 176-184
Chu R., J. Wei, M. Busch, Economic evaluation of UF+SWRO in seawater desalination. Chinese
Desalination Association Conference (2009), Qing Dao, China.
Daucik K., R.B. Dooley, Revised Supplementary Release on Properties of Liquid Water at 0.1 MPa, The
International Association for the Properties of Water and Steam, September 2008.
DuPont WaterApp, Pretreatment Advisor, Dupont (2022-09-09).
Gésan-Guiziou G., Boyaval E., Baufin G. (1999) Critical stability conditions in crossflow microfiltration
of skimmed milk: transition to irreversible deposition. Journal of membrane science 158
211-222.
Gilabert-Oriol, G. (2021). Ultrafiltration Membrane Cleaning Processes. In Ultrafiltration Membrane
Cleaning Processes. De Gruyter.
Gilabert Oriol, G., Hassan, M., Dewisme, J., Busch, M., & Garcia-Molina, V. (2013). High efficiency
operation of pressurized ultrafiltration for seawater desalination based on advanced cleaning
research. Industrial & Engineering Chemistry Research, 52(45), 15939-15945.
Gruskevica, K., & Mezule, L. (2021). Cleaning methods for ceramic ultrafiltration membranes affected
by organic fouling. Membranes, 11(2), 131.
Kim, Y. J., Jung, J. W., & Lee, S. (2015). Comparison of fouling rates for pressurized and submerged
ultrafiltration membranes. Desalination and Water Treatment, 54(13), 3610-3615.
Konieczny, K., Sakol, D., Płonka, J., Rajca, M., & Bodzek, M. (2009). Coagulation—ultrafiltration system
for river water treatment. Desalination, 240(1-3), 151-159.
Le Cleck P., Jefferson B., Chang I.S. Judd S.J. (2003) Critical flux determination by the flux-step method
in a submerged membrane bioreactor. Journal of membrane science 227, 81-93.
Liu, J., Huo, W., Zhang, X., Ren, B., Li, Y., Zhang, Z., Yang, J., (2018) Optimal design on the high-
temperature mechanical properties of porous alumina ceramics based on fractal dimension
analysis. Journal of Advanced Ceramic, 7, (2), 89-98.
Morris, Robert D., Elena N. Naumova, and Jeffrey K. Griffiths. ‘Did Milwaukee experience waterborne
cryptosporidiosis before the large documented outbreak in 1993?.’ Epidemiology (1998): 264-
270.
Mourato D., M. Singh, C. Painchaud, R. Arviv, Immersed membranes for desalination pre-treatment,
International Desalination Association (IDA) World Congress, Bahamas, Paradise Island,
Bahamas, 2003.
Nick, N., (2019) Pros And Cons Of Different Types Of Ultrafiltration Technology Configurations,
Weter Online.
Owen G., Bandi M. Howell J.A. Churchouse S.J. (1994) Economic assessment of membrane processes
for water and wastewater treatment. Journal of membrane science 102, 77-91.
Pressdee, J., Rezania, S., & Hill, C. (2005). Minneapolis Water Works’ ultrafiltration plant gets off to a
big start. Journal-American Water Works Association, 97(12), 56-63.

45
Experimental Methods for Membrane Applications

Raghunath B., Bin W,m Pattnaik P, Janssens J. (2012) Best practice for optimization and scale-up of
microfiltration TFF processes. Bioprocessing Journal. 11 (1) 30-39.
Sakola C, Konieczny K (2004) Application of coagulation and conventional filtration in raw water
pretreatment before microfiltration membranes. Desalination 162 61-67.
Sondhi, R., Bhave, R., & Jung, G. (2003). Applications and benefits of ceramic membranes. Membrane
Technology, 2003(11), 5-8.
Vyas H.K. Bennett R.J. Marshall A.D. (2002) Performance of crossflow microfiltration during constant
transmembrane pressure and constant flux operations. International dairy journal, 12, 473-479
Wang Z Field RW., Qu F., Han Y., Liang H Lia G., (2014) Use of threshold flux concept to aid selection
of sustainable operating flux: A multi-scale study from laboratory to full scale. Separation and
Purification Technology 123 (2014) 69–78.
Wu X., Xie Z., Wang H., Zhao C., Ng D., Zhang K (2018). Improved filtration performance and
antifouling properties of polyethersulfone ultrafiltration membranes by blending with
carboxylic acid functionalized polysulfone RSC ADV 2018, 8, 7774-7784.

46
 doi: 10.2166/9781789062977_0047

Chapter 3

Reverse Osmosis and

Nanofiltration
Guillem Gilabert-Oriol, DuPont Water Solutions, Spain

The learning objectives of this chapter are the following:

• give an understanding of the importance of reverse osmosis and nanofiltration

membranes and how they are used for water treatment

• give an introduction to the different equations that govern reverse osmosis

fundamentals and how to control, analyze and normalize a membrane installation

• give knowledge of how reverse osmosis and nanofiltration membranes are designed

• give knowledge on how membranes system is best operated to reduce membrane

fouling and ensure system availability.

3.1 THE RISE OF REVERSE OSMOSIS

Water scarcity is being recognized as one of the main threats that mankind is facing globally
(Fritzmann, et al., 2007). Reverse Osmosis (RO) membrane technology has developed as
a promising technology to address this problem, holding roughly 44% market share and
growing among all the desalinating technologies (Valavala, et al., 2011). This increased
market adoption has been driven as materials have been improved and costs have dropped
(Greenlee, et al., 2009. This is especially relevant in arid regions such in the Middle East
countries (ME), where population is located in arid and semi-arid regions, with a very limited
rainfall, and where due to high ambient temperatures, evaporation contributes to a higher
stress degree to the naturally available water sources. Moreover, water scarcity is aggravated
by the population increase this region is exposed, as well as the economic development
(Guo, et al., 2000). All these factors, together with the favorable energy to product quality
ratio that seawater reverse osmosis (SWRO) offers, has situated this technology as one key
driver to sustain population living standards in ME countries (Carroll, et al., 2000).

© 2024 The Authors. This is an Open Access book chapter distributed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0),
(https://creativecommons.org/licenses/by-nc-nd/4.0/). The chapter is from the book Experimental
Methods for Membrane Applications in Desalination and Water Treatment, Sergio G. Salinas-Rodriguez,
Loreen O. Villacorte (Eds).
Experimental Methods for Membrane Applications

The first technologies used to desalinate seawater were using thermal processes where
seawater is evaporated, and then the steam, which is free of salts, is recondensed to obtain
fresh water. These thermal driven technologies that rely on distillation are multistage-flash
distillation (MSF), multiple-effect distillation (MED), and vapor compression desalination
processes (VCD). The main drawback of these methods is the significant amount of energy
per cubic meter of water produced, compared to modern reverse osmosis based desalination.
As Figure 1 shows, reverse osmosis (RO) desalination, specially when coupled with energy
recovery devices (ERD) is 10 times more energy efficient than multistage flash desalination,
and 4 times more efficient than vapor compression distillation (Kumar et al., 2017, Kim
et al., 2019).

30

25

20

15

10

5
3
Energy consumption (kWh/m ) 0
MSF MED VCD RO RO+ERD

Figure 1 Energy consumption of different desalination technologies

One of the key aspects that has allowed reverse osmosis desalination to be so energy efficient
is the introduction of energy recovery devices (Kadaj and Bosleman, 2018). These systems
are like heat exchangers operation units, but instead of exchanging thermal energy, then they
exchange pressure. This allows to recovery almost all the energy that was previously lost in
the concentrate water stream of a seawater reverse osmosis system, and use it to pressurize
the same volumetric flow in the feed of the reverse osmosis system. If we take seawater
desalination as an example, this can reduce the energy expense in a seawater desalination
plant by 55%. Also, a smaller high pressure pump is needed to pressurize the feed of a revere
osmosis system, as 55% of the flow is already pressurized coming from the energy recovery
device. It is worth mentioning that previously, the concentrate stream of a reverse osmosis
system that could come pressurized up to 80 bar was discharge into the open atmosphere
and all this energy was lost.

3.2 SUSTAINAIBLITY OF REVERSE OSMOSIS

Reverse osmosis membranes membrane technology offers a solution to achieve the

sustainability development goals that the United Nations has set up for 2030. This has
been stated and recognized by the United Nations (SDG, 2023). Thanks to the desalination
technology, it is possible to fight against water scarcity, and obtain water of high quality.

48
 Chapter 3

If the energy is powered through renewable energies, such as through solar power or
photovoltaics, it is possible to achieve drinking water of high quality. This allows to fight
against climate change, as well as to provide unlimited amount of drinking water for the
population, that it is not linked to whether it rains or note in the nature.

It is also important to make sure that desalination concentrate discharge is done properly,
and that tis brine is properly managed through diffusors or mixing it with seawater, so its
discharge does not affect marine species or the environment (Fernández-Torquemada et al.,
2019).

The use of brine can also be used as a resource, to get value out of the brine, and being able
to extract valuable minerals such as sodium chloride, magnesium compounds, bromide,
and rubidium, among others. This is important, as this allows reducing the cost of water of
the desalination technologies, as well as preserving natural resources such as landscape and
mountains from invasive mining extraction operations (Casas et al., 2014).

Finally, some endeavors such as the Water Positive initiative, aims to take the sustainability
impact of desalination and water reuse one step further. This is inspired by the carbon credits
system, but for water credits. It aims to help those companies that aim to become water
neutral in terms of its water footprint, so that they can compensate their water negative use,
with those companies that are net producers of water, such as the desalination and water
reuse installations. This can help driving awareness of the importance to reduce the water
footprint, and helping preserve this valuable resource, as well as to help making sustainable
water treatment processes more affordable.

3.3 UNDERSTANDING THE OSMOSIS PROCESS

In order to understand why reverse osmosis is called with this name, it is important to first
understand what osmosis means. Osmosis is a natural process that only takes places when
there is a semi- permeable membrane.

3.3.1 Semi-permeable membranes

A semi-permeable membranes is defined by letting a solvent like water pass through it, but
not letting a solute like salt pass through the membrane. In order to better understand this
process, it can be useful to imagine a simplified scenario, where only water is considered
for a solvent, and only sodium chloride (NaCl) is considered for a solvent. When sodium
chloride is dissolved in water, both sodium and chloride are separated in terms of Na+ ions
and Cl- ions, following the equation detailed in Equation 1.

NaCl → Na+ + Cl − Eq. 1

Although both atomic radius of water molecule (H2O) and Na+ ions and Cl- ions are similar,
and therefore it could be expected that both water and sodium and chloride ions can pass
through the membrane in a similar rate, this is not the case. The reason for this phenomenon,
is the solvation phenomena. In order to maintain electrical neutrality, when ions are
dissolved in water, they become surrounded by water molecules. Since water molecules are

49
Experimental Methods for Membrane Applications

polar, they tend to orient their mostly negative charge, with the sodium positively charged
ions. The same happens for chloride negatively charged ions, as they get surrounded by
the negatively charged side of the water molecules. The ultimate consequence for both
chloride and sodium ions is that their effective sizes drastically increase, as a result of being
surrounded by water molecules. This is the main reason why small molecular weight species
that are charged are much better rejected from a solute like water. Therefore, the smaller
a species is, and the more charged it is, the better it will be rejected by a reverse osmosis
membrane.

The solvation effect is represented in Figure 2, where it can be seen the solvation effect of
water molecules, represented by red (oxygen) and grey molecules (hydrogen) to sodium
ions (blue) and chloride ions (green) as they pass through a space in the reverse osmosis
membrane.

– – – – – –
–
–
– – – –
–
– –
– +
– – – –
– –

Figure 2 Solvation effect of water (red and grey molecules) to sodium ions (blue) and chloride ions
(green) as they pass through a space in the reverse osmosis membrane

It should be noted that this “pore” drawn in this diagram, represent a space that is created
in the polyamide chain. As polymers rotate as a result of being above 0º Kelvin temperature,
small “pores” like this drawn in the diagram are created, where water can pass through it
freely through a pressure driven convective flow (Wang et al., 2023).

3.3.2 The reverse osmosis process

Once the concept of a semi-permeable membrane is defined, and it remains clear that it
lets a solvent like water to pass, while a solute like sodium chloride cannot pass, the natural
process of osmosis can be defined. As mentioned, the osmosis process only makes sense
when a semi-permeable membrane is present. Osmosis is defined as a natural process when
a specie that is at a higher concentration goes to dilute the other side of the membrane that
has lower concentration of this species. This is typically the case of solutes like water. Water
can move through a semi-permeable membrane, while salt cannot. Therefore, water with
lower salt concentration like fresh water will always go to dilute the water with higher
salt concentration like seawater, as this will have a lower water concentration. Since this
natural phenomenon is not the goal of producing fresh drinking water, as it is not desired
to get fresh water consumed. In order to achieve the opposite result, and be able to generate
fresh drinking water from seawater, the reverse osmosis process was invented. This process
consist into applying a pressure on the seawater that is enough in order to reverse this
process, and being able to generate fresh water instead of consuming it.

50
 Chapter 3

Osmosis process happens in nature. One example are cherries. When it rains, rain droplets
cover the cherry skin. The cherry skin acts like a semi permeable membrane, it lets the water
to pass through, but not salts or sugars to escape. Therefore, after it rains, water travels from
the outside of the skin, where the water concentration is higher, towards the inside of the
cherry skin, where the water concentration is lower as a result of the fiber and fructose that
cherries contain. The ultimate result is that as water from rain enters the inside of the cherry,
it can eventually crack the cherry skin as its volume increases, and the cherry skin might
not be able to expand properly before cracking to account for the increase in the cherries’
volume.

In order to visualize this, two different solutions with the same volume each one. These
solutions can be put in contact through a semi permeable membrane. The first solution is
seawater, which it is assumed to have 30 g salt with 970 g of water. This represents a 3.1%
salt concentration. The second solution is fresh water, which it is assumed to have 1 g of salt
with 999 g of water. This represents a 0.1% salt concentration This set up can be observed
in Figure 3. In this case, two solutions are separated. Fresh water will go to dilute seawater,
while salt will not be able to pass.
Seawater Fresh water

30 g Salt 1 g Salt
970 g Water 999 g Water

Time = 0
3.1% salt 0.1% salt

Figure 3 Initial experimental set up

As salt cannot pass, water from the fresh water side will go to dilute the seawater. Water will
keep passing until both sides salt concentration are the same. This will happen after 935 g
of water have passed from the fresh water side to the see water side. At this equilibrium
point, concentrations in both sides will be the same at 1.6%. It is important to notice that
salt mass in both sides is kept the same, but water mass has changed. This change in water
volume, which increases in the seawater side but decreases in the fresh water side, leads to
a difference in water height between the seawater and fresh water side. This difference in
height is what is called osmotic pressure. This osmosis process be seen in Figure 4.

Seawater Fresh water

30 g Salt 10 g Salt
1,905 g Water 64 g Water

935 g Water
Time = [0,∞]
1.6% salt 1.6% salt

Figure 4 Osmosis process

51
Experimental Methods for Membrane Applications

If one aims to reverse this naturally occurring osmosis process, one needs to apply a pressure
that at least is the same as this osmotic pressure. Once this pressure is applied, the water
flow is reverses. If exactly the same pressure as the osmositc pressure is applied, it will be
possible to reverse the 935 g of water flow from the seawater side to the fresh water side.
This process can be observed in Figure 5.

Pressure

Seawater Fresh water

30 g Salt 1 g Salt
1,905 g Water 64 g Water

935 g Water
Time = [0,∞]
1.6% salt 1.6% salt

Figure 5 Reverse osmosis process

After applying the same pressure as the osmotic pressure, the equilibrium will again be
reached, and the initial state will be created. Since typically the goal is to produce fresh water
from seawater, and not to prevent osmosis to happen by reaching an equilibrium as the
one depicted in Figure 6, a pressure greater than the osmotic pressure will be needed to be
applied to produce more fresh water by consuming seawater.

Pressure

Seawater Fresh water

30 g Salt 1 g Salt
970 g Water 999 g Water

Time = 0
3.1% salt 0.1% salt

Figure 6 Equilibrium step

3.4 EQUATIONS

Reverse osmosis membranes are defined by a simple set of equations. These equations are
used to control the flow of a solvent like water through the membrane, as well as the flow
of solutes like salt through a membrane, and also finally to calculate the osmotic pressure
needed to start producing fresh water from a higher concentration water. Also, the equations
that are used to characterize a reverse osmosis system are presented.

52
 Chapter 3

3.4.1 Fundamental equations

Osmotic pressure is defined by the Greek letter pi (Π). Osmotic pressure can be calculating
by multiplying the temperature (T) in Kelvin degrees, with the ideal gas constant (R) and the
solute concentration (C) and the osmotic pressure coefficient (Φ). This formula is shown in
Equation 2.

3.4.1.1 Osmotic pressure

The osmotic pressure coefficient represents how well a solute dissociates in water. For
NaCl, which fully dissociates in sodium and chloride ions following Equation 1 it will be
equal to 1. For other species that do not dissociate at all in water, its value will be equal
to 0. Concentration is typically expressed in molar mas (mol/L). Ideal gas law is typically
expressed as 0.08314 L·bar/K·mol. Temperature is expressed in kelvin degrees (K).

π = ϕ CRT Eq. 2

One easy way to remember the osmotic pressure equations is thinking how for diluted
solutions, the osmotic pressure resembles the Van’t Hoff equation for ideal gas laws. Van’t
Hoff equation is shown in Equation 3.

PV = nRT Eq. 3

Rearranging terms of Van’t Hoff Equation 3 the same equation than osmotic pressure
Equation 2 can be obtained, as shown in Equation 4. It is important however to highlight
that the osmotic pressure dissociation coefficient needs to be factor in this equation.
n
P= RT = CRT Eq. 4
V

The net driving pressure (NDP) represents how much of an energy driving for it exists across
the membrane. This is obtained by discounting the osmotic pressure (Π) to the pressure that
is applied to make a membrane permeate water (P). Pressure units are typically expressed in
bar or psi. This formula is shown in Equation 5.

NDP = P − π Eq. 5

To better understand how to calculate the osmotic pressure, the following example can
be studied. To calculate the osmotic pressure of a solution that has 2 g/L sodium chloride
dissolved in water, Equation 2 can be used.

The first step is to transform the mass concentration to molar concentration. This is achieved
in Equation 6.

2 g NaCl 1 mol NaCl 0.0342 mol NaCl

= Eq. 6
L 58.44 g NaCl L

53
Experimental Methods for Membrane Applications

Since sodium chloride fully dissociates in water following Equation 1, osmotic pressure
dissociation coefficient can be assumed 1. Therefore it is possible to calculate the
concentration of each individual ion dissolved in water as described in Equation 7.

0.0342 mol NaCl → 0.0342 mol Na+ + 0.0342 mol Cl − Eq. 7

Finally, each contribution of the osmotic pressure needs to be calculated for each individual
ion. This is achieved by using Equation 2 in each individual sodium and chloride ion. This
can be seen in both Equation 8 and Equation 9 respectively, where it can be seen that both
sodium and chloride ions have the same osmotic pressure individual contribution of 0.85
bar each.

0.0342 mol Na+ 0.08314 L bar

PNa+ = 298K = 0.85 bar Eq. 8
L K mol

0.0342 mol Cl − 0.08314 L bar

PCl− = 298K = 0.85 bar Eq. 9
L K mol

Finally, tot total osmotic pressure of sodium chloride can be calculated adding each
individual ion osmotic pressures. This is shown in Equation 10, where it can be seen that the
total osmotic pressure of a 2 g/L sodium chloride solution equals 1.7 bar. A rule of thumb
to quickly estimate the osmotic pressure is to divide the total dissolved solids by 100. This
gives an approximation of the osmotic pressure in psi. To have it in bar, this resulting value
needs to be multiplied by 14.5.

+
PT = PNa + PCl = 1.7 bar Eq. 10

3.4.1.2 Water flux

Water flux across a membrane (Fw) is defined as the multiplication of the water permeability
value, called A-value (A) with the net driving pressure, which is obtaining by subtracting
the osmotic pressure gradient (Π) to the pressure gradient applied to the membrane (P).
Flux of water is typically expressed in US gallons divided per square feet per day (gfd) or
in liters divided per hour per square meter per day (LMH). A-value expresses membrane
water permeability, and it is typically expressed in US gallons divided per square feet per
day per psi (gfd/psi) or in liters divided per hour per square meter per bar (L/m2·h·bar or
simply LMH/bar). Pressure is typically expressed in psi or bar. This formula is shown in
Equation 11.

Fw = A(P − π ) Eq. 11

54
 Chapter 3

It is important to notice how this equation resembles the Darcy’s law equation, which states
that a flow across porous membrane is proportional to the pressure that is applied. Darcy’s
Law can be seen in Equation 12, where k represents the permeability coefficient, μ the
dynamic viscosity, and L the membrane thickness. All these parameters can be incorporated
into the A-value membrane permeability coefficient.

k
Fw = P Eq. 12
µL

This is the same equation that governs the transport ort of water across an ultrafiltration
membrane. Therefore, it can be concluded that for a solvent like water, when it faces a
semi-permeable membrane where it can pass freely through it, it acts as a pressure-driven
convective flow filtration.

3.4.1.3 Salt transport

The flux of salt across a membrane (Fs) is described as the multiplication of the salt
coefficient value, also referred as B-value, with the concentration gradient of solutes across
the membrane. The flux of salt across a membrane is typically expressed in pounds divided
per square feet per day (lbfd) or in grams divided per hour per square meter per day (GMH).
B-value, or salt diffusion coefficient is typically expressed in US gallons divided per square
feet per day (gfd) or in liters divided per hour per square meter (L/m2·h or simply LMH).
Concentration is usually expressed in pounds divided by US gallon (lb/gal) or grams divided
per liter (g/L). This formula is described in Equation 13.

Fs = BC Eq. 13

It is important to notice how this equation resembles the Fick’s law equation. Fick’s law
describes de diffusion transport of mass across a membrane. As salt cannot pass through the
cavities that are created in a membrane, it needs to pass through diffusion. It can be observed
how diffusion is noted as with a D, and this corresponds to the B-value diffusion coefficient.
Concentration gradient stays the same. This formula is shown in Equation 12.

Fs = DC Eq. 14

3.4.1.4 The difference between convective and concentration driven flows

Typically, a pressure driven convective flow is several orders of magnitude higher than the
mass transfer coefficient that can be achieved through diffusion. This is why reverse osmosis
membranes are able to separate so well a solute from a solvent. This mainly happens since
for a semi-permeable membrane, a solvent like water can travel across the membranes just
following a pressure gradient. This means that water sees ‘pores’ across the membrane.
However, salt cannot pass through these ‘pores’, as because of solvation, dissociated species
in water are too big to pass through these ‘pores’, and the only way they have to pass through
a membranes is through diffusion.

55
Experimental Methods for Membrane Applications

The following examples illustrates the different order of magnitude difference between a
convective flow like water, and a diffusion flow like sodium chloride across the membrane.

To calculate the flux of water across a membrane, it can be assumed a water permeability
A-value of 4 LMH/bar, a 15 bar feed pressure, a 1 bar osmotic pressure.

Fw = A(P ) = 4(15 1) = 56 L / m2 h = 56,000 gm2 h Eq. 15

To calculate the flux of salt across this same membrane, a salt diffusion coefficient of 0.2
LMH and a concentration of salts of 2 g/L can be assumed.

Fs = BC = 0.2 2 = 0.4g / m2 h Eq. 16

As it can be observed from this example, reverse osmosis membranes are really selective
to water mass transport across the membrane when compared to salt mass transport.
These serval order of magnitude different in mass transport clearly illustrate the difference
between convective and diffusive flow.

3.4.2 System equations

A reverse osmosis system is mainly composed by a feed flow (Qf), and then this feed flow
gets divided between the filtrated flow that is treated with the membrane active layer, which
is called the permeate flow (Qp), and the concentrate flow (Qc), which has all the rejected
salts or spices that could not permeate the membrane. Concentrate flow is sometimes also
referred as brine or retentate. Flows are usually specified with in cubic meters per hour or day
(m3/h or m3/d), or in US gallons per day (gfd). Plant capacity represents the permeate flow a
desalination plant can produces, and it is usually expressed in millions liters per day (MLD).
1,000 m3/d equals 1 MLD. A simple reverse osmosis diagram can be found in Figure 7.

Qp
Qf
Qc

Figure 7 Reverse osmosis diagram

A reverse osmosis system is characterized by having close water mass balance. This means
that all the water that is entering the reverse osmosis membrane (Qf) needs to exit the
reverse osmosis system through either the permeate (Qf) or through the concentrate flow
(Qc). This formula is depicted in Equation 16.

Q f = Q p + Qc Eq. 17

56
 Chapter 3

A reverse osmosis system is also characterizing by having a neutral salts mass balance. This
means that all salts that are entering the system will be also exiting the reverse osmosis
membrane by either the permeate of the concentrate. Individual salts concentration for the
feed (Cf), permeate (Cp) and concentrate (Cc) are typically represented in g/L or in mg/L
(ppm). This is represented in Equation 18.

Q f C f = Q pC p + QcCc Eq. 18

Flux (F) represents a flow relative (Q) to the membrane active area (A) it is permeating. Flux
is typically measured in cubic meters per hour or day per square meter (m3/h or m3/d)
or in US gallons per day per square feet (US gall/(d·ft2) or gfd). This formula is shown in
Equation 19.

Q
F= Eq. 19
A

Reverse osmosis system recovery (R) represents the process water yield, and is calculated
dividing the permeate flow (Qp) by the feed flow (Qf). It is expressed as a percentage. This
formula is shown in Equation 20.

Qp
R= Eq. 20
Qf

Salt passage (SP) represented the percentage concentration of salt that is passing through
the membrane compared to the initial salt concentration being treated. It is calculated by
dividing the concentration of salt (Cp) in the permeate by the concentration of salt in the
feed (Cf). This parameter is useful from a physics point of view as it lets directly comparing
two different membrane performances. This formula is shown in Equation 21.

Cp
SP = Eq. 21
Cf

Salt rejection (SR) represents the percentage on how much salt a membrane is rejecting. It is
calculated by subtracting 1 minus salt passage (SP). This parameter is useful as it enables to
quickly realize how much solute or salts are being rejected by a membrane system. However,
in order to perform comparative evaluation of two different membranes performance, it is
usually necessary to do any comparative evaluation using the salt passage parameter. This
formula is shown in Equation 22.

SR = 1− SP Eq. 22

57
Experimental Methods for Membrane Applications

Another factor that is calculated is the plant availability (Av). This represents the amount
of time in percentage that the plant is in operation producing water (top) and therefore
not stopped versus the total time of the time period being considered (tT). This formula is
shown in Equation 20.

top
Av = Eq. 23
tT

3.4.3 Factors affecting membrane performance

Several factors can affect membrane performance. The three factors that are usually most
relevant are the effect that feed pressure increase, feed concentration increase, and feed
temperature increase have on membrane performance. In order to understand how these
factors changes, only three equations are required. These are the osmotic pressure equation,
shown in Equation 5, the water transport equation, shown in Equation 11 and the salt
transport equation, shown in Equation 13.

A summary table highlighting these interactions can be found in Table 1.

Table 1 Summary table of the effect of feed pressure, concentration and temperature on the salt
rejection and flux of a reverse osmosis membrane
Flux Salt Rejection

Pressure Fw = A · (P -π)
Fs = B · C

Concentration Fw = A · (P-π )

Fs = B · C

Temperature Fw = A · (P-π )

Fs = B ·C

3.4.3.1 Feed pressure

When feed pressure increases, water flux also increases as a result of an increase in the
pressure. Salt passage across the membrane stays constant, but since more water is passing
across the membrane, whet the water flow is divided by the same amount of salt, the final
salt concentration in the permeate decreases. Therefore, salt rejection increases.

3.4.3.2 Feed concentration

As feed concentration increases, osmotic pressure also increases. This leads to a direct
reduction in the water flux across the membrane, as there is less net driving pressure
available for the membrane to permeate. Additionally, as concentration increases, the flux
of salt directly increases. This leads to a decrease in water passing through the membrane,
which is divided by a higher salt passing through the membrane, thus increasing the salt
passage and decreasing the salt rejection.

58
 Chapter 3

3.4.3.3 Feed temperature

As feed water temperature increases, the water becomes less viscous. This means that
with the same amount of energy, more water can permeate through the membrane. This
eventually leads to improving the water permeability value (A-value) and therefore the
permeate water flux. With regards to the salt rejection, an increase in temperatures improves
much more the salt diffusion factor (B-value). Therefore, there is a higher increase in salt is
passing through the membrane than water passing through the membrane, and as a result,
salt rejection decreases.

3.4.3.4 Concentration polarization

Concentration polarization is the phenomenon in which as membrane removes water from
the feed solution, solutes are pushed towards the boundary layer of the membrane. This
leads to a decrease in performance in the reverse osmosis membrane system as, since the
membrane is filtration, the membrane sees the concentration in the boundary layer and
not in the feed solution. As this salt concentration is higher, this means that the membrane
experiences a decline in flux and salt rejection due to this phenomenon (Sablani et al., 2001).
Concentration polarization can be minimizing by controlling the membrane recovery rate,
as well as through membrane element design elements such as a feed spacer that is able
to provide a proper mixing and therefore minimizes the accumulation of solutes in the
membrane boundary layer.

3.5 REVERSE OSMOSIS MEMBRANES

Reverse osmosis and nanofiltration membranes are pressure-driven membrane filtration

processes, where feed pressure bigger than the osmotic pressure is used to filtrate water
through the membrane.

Commercial elements used in large industrial installations are standardized. They are usually
referred as spiral-wound polyamide based membrane configured in a cylindrical shape with
a typical diameter is 8-in (20 cm), with a typical length of 40-in (1 m). For smaller industrial
application where the water capacity required is lower, elements with 4-in diameter and
40-in elements are also used. Finally, elements used in home drinking applications are less
standardized, and their size in terms of diameter and length can vary depending on each
manufacturer. Examples of these elements are 1.8-in and 2.5-in diameter elements with
12-in or 14-in length. An example of a DuPont FilmTec™ BW30 PRO-400 membrane
can be found in Figure 8. In this example, feed flow will enter the membrane through the
anti-telescoping device on the left. The permeate flow will be collected in each membrane
leave, and finally collected through the inner permeate water tube. The permeate flow will
be exiting the membrane through the permeate water channel located in the center of the
membrane, leaving through the right of the membrane. The concentrate flow will also be
exiting the membrane through the right part through the anti-telescoping device on the
right part. It is useful to realize that the remaining feed water that exits the membrane is
what it is called concentrate.

59
Experimental Methods for Membrane Applications

Figure 8 DuPont FilmTec™ BW30 PRO-400 membrane

A reverse osmosis membrane is typically composed by an active polyamide base layer that is
around 0.2 μm thick. This polyamide membrane is also referred as the active layer, as is the
one responsible from separating the salt solutes from the water solvent. As this membrane
is so thin, in order to be able to precipitate it during the phase inversion process, a support
layer typically consisting of polysulphone is used. This allows the proper precipitation of
polyamide on the polysulphone. The polysulphone layer has a thickness of around 40 μm
thick. As still this thickness is rather seen, in order to enhance its mechanical structure, this
layer is put in a polyesther reinforcing layer, typically consisting of 120 μm thick layer. This
three multi-layer structure is usually referred as thin-film (TFC) composite layer (DuPont,
2023). A schematic of this arrangement can be found in Figure 9.

Polyamide 0.2 µm
Polysulfone 40 µm

Polyester 120 µm

Figure 9 Thin-film composite reverse osmosis membrane multi-layer composition

A scanning electron microscopy (SEM) image courtesy of DuPont FilmTec™ membranes

depicting the main three layers in a reverse osmosis membrane can be found in Figure 10.

60
 Chapter 3

Polyamide 0.2 µm

Polysulfone 40 µm

Polyester 120 µm

Figure 10 SEM image of a reverse osmosis membrane

Nanofiltration membranes are very similar to reverse osmosis membranes. Their main
difference is that the active layer usually consists of a polypiperazine polymer. Typically,
nanofiltration membranes are used when only certain solutes are needed to be separated,
but not all of them. This allows to significant energy savings. Examples of their use are
sulphate removing nanofiltration membranes. These membranes can let sodium chloride
pass through their active layer, but they remove sulphates and other divalent ions. This is
especially useful for oilfield applications, where seawater is used for injecting it into the oil
wells. In this application, no sodium chloride needs to be removed. However, to prevent
multiple problems, sulphate needs to be removed. By using nanofiltration membranes, the
operating pressure can be reduced from 70 bar to 15 bar, therefore saving a lot of energy.

Typically a revers osmosis membrane spiral-wound element consists of multiple polyamide

sheets that are rolled together. Each membrane sheet is separated from the one on its top
by a feed spacer. Inside each membrane there is a permeate spacer. The role of the spacers
is to provide mechanical integrity into the reverse osmosis element so that it can be
properly folded. The feed spacer also plays a crucial role into minimizing the concentration
polarization effect, as well as controlling biofouling and saving energy. Minimizing the dead
spaces inside a membrane is important to prevent biological growth inside of a membrane.
This schematic shown in Figure 11 shows the main parts of a spiral wound reverse osmosis
element.

Permeate spacer Qp
Membrane
Feed spacer
Qf Qc

Qp

Qf Qc

Qp

Figure 11 A spiral wound reverse osmosis elements with its parts

61
Experimental Methods for Membrane Applications

3.5.1 The significance of desalination

As of 2023, there are more than 21,000 desalination plants in the world. All these plants
provide a total install capacity of newly created fresh water equivalent to 100,000,000 m3/d
(Climate ADAPT, 2023).

3.6 PERFORMANCE MONITORING

Typically, systems are designed to provide a constant yearly water production capacity.
Therefore systems are designed at a constant flux rate. So, when systems suffer from fouling
or changes in operating condictiones such as temperature decrease or salinity increase,
typically their flux rate would decrease. To prevent this, more energy is used, so that the
membrane system can compensate for the decrease in membrane permeability and be able
to provide the same operating flux.

Therefore, it is of utmost importance to periodically monitor the performance of a reverse

osmosis systems. There are three key parameters that need to be monitored in a reverse
osmosis.

The first one is the energy consumption, monitored through the high pressure pump
operating pressure. This is an important parameter because the energy consumption directly
impacts the operating expenses (OPEX) of a reverse osmosis pump. Additionally, the high-
pressure pump needs to have enough capacity to increase the pressure it is delivering to
the membrane system. If the pump cannot deliver enough pressure, the whole desalination
installation can start suffering from a decrease in the water it is delivering. This can lead
to water shortages and even being outside the offset contract. This is why typically plants
are designed in a conservative way. To properly size the high pressure pump, typically the
lowest yearly temperature is used to size the pump, as the lowest water temperature will
provide the highest pressure needed to sustain the targeted design installation permeate
flow capacity.

The second parameter that is key to monitor a membrane system is the permeate water
quality. Typically, the water conductivity is monitored. Conductivity is typically expressed
in μS/cm. Conductivity is used to estimate the total dissolved solids (TDS) salinity of water.
A good rule of thumb for low salinities water is that 2 μS/cm are equal to 1 mg/L (ppm)
salinity. For seawater types, a good rule of thumb is that 1.4 μS/cm are equal to 1 mg/L
(ppm). Sometimes, beside general salinity measured by conductivity, specific spices that are
important are also specified. This can involve measuring specific targets such as alkalinity,
boron, and pH, among others. Water quality is of utmost importance, because ultimately,
water treatment plants are usually designed to provide a warrantied water quality that is
typically limited to be below a certain limit. Therefore, water quality typically acts as
the independent variable when designing a membrane system. Typically, a membrane
installation is designed taking into account the highest water temperature thought the yearly

62
 Chapter 3

temperature cycle. This is because at the highest temperature is when the water quality will
be the worst, and therefore show the highest permeate water total suspended solids value.
The third important parameter to monitor is the membrane pressure drop (dP). Pressure
drop is a factor that is important to measure as it directly affects the energy consumption.
However, this parameter is key as when pressure drop increases, this means that the
membrane feed-concentrate channel is getting blocked. This can happen if the membrane is
experiencing fouling, as when the membrane gets fouled, this means it is getting obstructed,
and therefore it is more difficult for water to travel across the membrane. They key problem
this issue presents, is as if the membrane gets too much blocked, the membrane can start
to get mechanically damaged, and it can eventually lead to its irreversible damage and the
membrane can stop working as intended. Therefore, when pressure drop starts to increase
to higher values, it is a good habit to perform a cleaning-in-place (CIP), in order to try to
recover the membrane performance to the initial pressure drop values. It should be noted
that the more a membrane has higher pressure drop, the more difficult it becomes to clean
the membrane and restore its performance to its initial values.

This is why it is so important to clean the membranes early, so that it is easier to recover
the membrane performance. Because membranes systems typically suffer from fouling,
especially in the areas of the planet where temperature is usually higher and there is
therefore a higher water demand due to water scarcity issues, typically plants are designed
with redundant trains. A train is a collection of pressure vessels. Each pressure vessel usually
contains up to 7 membranes in series, and a train usually has dozens of pressure vessels. This
means that for example, in a plant that has 11 reverse osmosis trains, 1 of this trains can be
used to be put in operation when one train needs to be going through a chemical cleaning.
This designs are called in the industry N-1 designs.

3.7 NORMALIZATION

Monitoring feed pressure, permeate conductivity and pressure drop is of utmost importance
in order to be able to anticipate to possible problems that might arise from changes in
operating conditions and in fouling. As it was explained previously, as a result of water
temperature or water salinity changes, the energy consumption and permeate water quality
can change. Therefore, it is of utmost importance to understand if these changes in water
quality or energy consumptions are due to unexpected problems such as fouling, or they are
normal and expected as a result of the physical principles that were previously explained.
This is when membrane normalization comes into approach.

The key normalized parameters that are needed to be analyzed in a reverse osmosis
membrane system are the normalized permeate flow, the normalized salt rejection, and the
normalized pressure drop.

Normalization can be done using the equations listed below, or using computer assisted
programs such as the FT-Norm PRO that DuPont offers.

63
Experimental Methods for Membrane Applications

3.7.1 Why normalization matters

The example detailed in Table 2 is a good example of why normalization matters. This
is an example of a reverse osmosis plant where after 5 days of operation, the feed water
temperature decreases from 25 ºC to 21 ºC. As it can be seen in the monitoring parameters,
the plant is delivering a constant permeate production of 50 m³/h with a constant feed
flow of 100 m³/h. This represents a 50% recovery. Feed salinity is also constant at 1,000
mg/L. However, it can be observed how feed pressure increases from 8 bar to 8.81 bar,
and permeate quality decreases from 30 mg/L to 27 mg/L. Without normalizing the data,
it would be challenging to understand if this increase in feed pressure and improvement
in water quality is a result of expected thermodynamics, or it is a result of the membrane
getting for example fouled. Normalization is the tool that allows to properly perform this
assessment.

In this particular example, it can be seen that normalized permeate flow stays constant,
while normalized salt rejection also stays constant. This means that the increase in feed
pressure and improvement in water quality is not related to fouling, but do to they normal
behavior of the membrane system. This happens as seen previously, when the temperature
decreases, water quality improves as less salt passes through the membrane. Additionally,
more energy is needed to pump the water through the membrane as its viscosity increases.
If the normalized permeate flow would show, for example, a decrease over time, it could be
suspected that fouling is responsible for this loss in membrane permeability. If normalized
salt rejection would be, for example, decreasing, this could also indicate the likelihood of
issues in the membrane system.

Table 2 Normalization example

Normalized
Permeate Feed Feed TDS Permeate permeate Normalized
Feed flow flow pressure (mg/L) TDS Temperature flow salt
Days (m³/h) (m³/h) (bar) (mg/L) (ºC) (m³/h) rejection
0 100 50 8.00 1,000 30 25 50 97.0%
1 100 50 8.00 1,000 30 25 50 97.0%
2 100 50 8.19 1,000 29 24 50 97.0%
3 100 50 8.39 1,000 28 23 50 97.0%
4 100 50 8.59 1,000 27 22 50 97.0%
5 100 50 8.81 1,000 27 21 50 97.0%

3.7.2 Equations
Normalization can be done with operating software’s such as DuPont’s FT-Norm PRO
software, or using the equations found in (DuPont, 2023).

3.7.2.1 Normalized permeate flow

To normalize the permeate flow (Qpn) the following formula found in Equation 24 can be
used. It must be noted that the subscript 0 references to the conditions being normalized to.
This can typically be the first data when the system is stabilized, or even an ideal projection

64
 Chapter 3

that the system is designed to operate at normal or ideal conditions. The other data without
subscript refers to the data at the time instance being normalized to. Units are in bar for the
pressure drop, and in m³/h or m³/day for the flows.

NDP0 TCF0
Q pn = Q p Eq. 24
NDP TCF

Net driving pressure (NDP) is calculate taking the feed pressure, and subtracting half the
pressure drop or differential pressure (dP) to the permeate pressure (Pp) and the average
feed concentrate osmotic pressure (Πfc). All pressures are measured in bar. This is shown in
Equation 25. It should be noted that the osmotic pressure in the permeate has been omitted
for simplification.

dP
NDP = Pf − − Pp − π fc Eq. 25
2

Pressure drop (dP) is calculated using Equation 26. To calculate it, the concentrate pressure
(Pc) needs to be subtracted to the feed pressure (Pf).

dP = Pf − Pc Eq. 26

The average feed-concentrate osmotic pressure (Πfc) can be calculated using Equation 27
(DuPont, 2021). It should be noted that Temperature (T) is in kelvin.

fc
= 0.00265 ×C f c ×T Eq. 27

The average feed-concentrate concentration (Cfc) can be calculated using Equation 28. It uses
the feed concentration (Cf) and the membrane system recovery (R). Its units are in mg/L.

1
ln
Cf c = C f 1 R Eq. 28
R

To calculate the temperature correction factor, Equation 29 can be used. It should be noted
that also Temperature (T) needs to be in kelvin. Also, the k parameter depends on water
temperature. If temperature ≥ 25 ºC, k = 2640. If temperature ≤ 25 ºC k = 3020.

1 1
k 298 T
TCF = e Eq. 29

65
Experimental Methods for Membrane Applications

3.7.2.2 Normalized salt rejection

In order to normalize the salt rejection, the following formula is used. This is displayed in
Equation 30 (DuPont, 2021).

Q p ×TCF0 × C fc0 ×C f
SPn = SP Eq. 30
Q p0 ×TCF × C fc × C f0

3.7.2.3 Normalized pressure drop

Normalized pressure drop (dPn) can be calculated using the formula displayed in Equation
31. To do this calculation, the standard pressure drop (dP) is used, together with the division
of the baseline average feed-concentrate flow (Qfc0) and the average feed-concentrate flow
(Qfc). Units are in bar for the pressure drop, and in m³/h or m³/day for the flows.

C fc0
dPn = dP Eq. 31
C fc

3.8 FOULING

Fouling is the phenomenon in which membranes get dirty. Fouling is one of the major
challenges that is nowadays affecting the reverse osmosis industry (Kucera, 2011). The key
problem that fouling offers is that it is difficult to foresee, and also it is difficult to deal with
it once fouling starts to appear. It is sometimes also difficult to identify the type of fouling
present.

Fouling is typically characterized because of the accumulation of unwanted spices that

get deposited inside the membrane, leading to a decrease in its performance. Typically
fouling leads to higher operating costs in membrane systems, as it can lead to higher
energy consumption, lower water quality, lower water production, higher chemical costs
because of the cleanings, and even an irreversible damage to the membrane. Fouling can
be characterized into four major types, plus a fifth one that relates to membrane integrity
failure.

3.8.1 Biofouling
The main type of fouling is biological fouling or simply biofouling. Biofouling happens
as certain bacteria, that have evolved to form a biofilm when they found a solid surface,
meet the membrane and the feed spacer. When this happens, these bacteria attach on the
membrane, and they start to form a biofilm.

Biofouling is characterized by an exponential increase in the feed-concentrate pressure drop

of the reverse osmosis membrane. It can lead to an increase in the energy consumption and
if not deal properly, it can mechanically damage the reverse osmosis membrane. Biofouling
is typically present in the first elements of a pressure vessel.

66
 Chapter 3

Once a biofilm is established on the membrane, it is difficult to remove. Useful strategies

involve cleaning the membranes with caustic cleanings (CIP). These cleanings are most
effective the higher the temperature and pH. If possible by the membrane characteristics,
effective cleanings will involve pH around 13 at a temperature around 35 ºC using NaOH.
It is important to have enough time for the chemicals to get soaked in the membranes,
and then flush the soaked solution effectively. This combination of steps can be repeated
multiple times until the soaking solution appears clean.

Other more preventive approaches involve starving the bacteria from growing. This can
be achieved using pre-treatment technologies like the DuPont B-Free™ technology, that
is able to eliminate nutrients before they reach the reverse osmosis system (Kucera, 2011).
This system has been proved extremely effective in preventing biofouling development
into reverse osmosis systems.

Other concepts that are practiced involve using non-oxidizing biocides, such as DBNPA or
2,2-dibromo-3-nitrilopropionamide (DBNPA). This biocide is described as non-oxidizing
the polyamide active layer of the reverse osmosis membrane. Some considerations that
need to be taken into account with this solution is its limited compatibility with drinking
water application, and also the fact that bacteria can get used to the biocide, and at certain
point, it can stop being effective and biofouling can develop again.

3.8.2 Organic fouling

The second type of fouling is the organic fouling. Organic fouling happens when organics
molecules get accumulated on the membrane surface. It typically leads to decline in
normalized permeate flow.

The main way to deal with organic fouling is through performing caustic based chemical
cleanings (CIP). These cleanings are most effective the higher the temperature and pH. The
same cleaning procedure as with biofouling can be followed.

3.8.3 Particulate fouling

Particulate fouling, or colloidal fouling, refers to the accumulation of particles in the
membrane. This can be due to an inadequate pretreatment. This type of fouling leads to a
rapid increase in the feed-concentrate pressure drop, as the feed spacer channel fills quickly
with particles.

In order to clean particular fouling effectively, it might be necessary to perform a caustic based
chemical cleanings (CIP). These cleanings are most effective the higher the temperature and
the pH is. If possible by the membrane characteristics, effective cleanings will involve pH
around 13 at a temperature around 35 ºC.

A good solution to deal with this type of fouling can involve upgrading the pretreatment
from a conventional one to a membrane based one, like ultrafiltration. Another solution
could be making sure there are no fiber breakages in the ultrafiltration part, and if there are,
repairing them with glue and pins, or replacing the damage modules with new ones.

67
Experimental Methods for Membrane Applications

3.8.4 Scaling
Scaling is a type of fouling that typically occurs when non too soluble salts starts to
precipitate on the membrane module. This can happen if water recovery is too high, or
if temperature or water composition changes. A recommended way to prevent scaling is
consulting a specialized anitscalant company. They have powerful software that simulate
the operating conditions at the targets recoveries and temperatures, and recommend the
best antiscalants and their concentration to used, based on the spices that have higher risk to
precipitate as they have the risk to surpass its solubility limit. Scaling typically happens in the
last elements of a pressure vessel, as there is where there is less water, and the water is more
concentrated. So dissolved spices have a higher risk to precipitate. Scaling typically leads to
an increase in pressure drop, as well as to a decrease in water quality or salt rejection. As salts
start to precipitate on the membrane surface, this affects the concentration polarization,
and increases the effective concentration of salts on the boundary layer. Scaling can also
therefore lead to a decrease in normalized permeate flow, since the osmotic pressure is
greatly increase, thus reducing the net driving pressure.

In order to clean a scaled membrane, it is important, if possible, to autopsy a membrane, to

understand the type of scaling present. There are multiple companies that are specialized in
offering this service. To clean the membrane, it is recommended to perform an acid chemical
cleaning (CIP). If the membrane characteristics allows it, a pH of 1 at room temperature
is effective. It is recommended to use HCl to prevent any further scaling cause by other
species such as sulfuric acid. Sometimes, membrane can have some organic or biofouling
too. Therefore, it is also recommended to always start with a caustic cleaning as previously
described, and then follow the acid cleaning step.

3.8.5 Integrity failure

Integrity failure occurs when the membrane is suffering from a non-fouling related damage.

This can involve chemical oxidation of the membrane. This can happen if for example,
sodium hypochlorite (NaOCl) from the ultrafiltration pre-treatment manages to reach the
reverse osmosis membrane, as the polyamide can get damage and eventually eliminated.
Symptoms involve an increase in the normalized salt passage, as since the membrane does
not have an active layer, it stops separating salt from water. Once oxidation is detected, it is
important to eliminate the source that is causing the chemical that is leading to oxidation to
leach. A strategy to properly address the leaching of NaOCl in ultrafiltration membranes is
described here (Gilabert-Oriol, 2021).

Other types of mechanical failures might involve o-ring failure. This can happen when the
o-ring that is used to separate a membrane gets pinched, there is a by-pass of water from
the feed side to the permeate side. This leads to an increase in the normalized salt passage.
In order to fix this, a probing test needs to be done in each membrane connection inside a
pressure vessel. A methodology to perform this test is described elsewhere (DuPont, 2021).

68
 Chapter 3

Compaction can happen when a membrane is operating at a too high pressure and
temperature. Compaction can be reversible or irreversible, or a combination of both. Typicall
effects of compaction involve a high increase in energy consumption, observed by a decline
in normalized permeate flow. Compaction can lead to an improvement in water quality, as
the membrane becomes more dense, it is more difficult for the salt to pass through it. When
compaction is identified, it is important to either use compaction resistant reverse osmosis
membranes, or to decrease, if possible, the target permeate flux, specially in periods of high
temperature, with the aim to reduce the operating pressure.

69
Experimental Methods for Membrane Applications

3.9 REFERENCES

Carroll T., S. King, S. R. Gray, B. A. Bolto, N. A. Booker, The fouling of microfiltration membranes by
NOM after coagulation treatment. Water Research 34.11 (2000) 2861-2868
Casas, S., Aladjem, C., Larrotcha, E., Gibert, O., Valderrama, C., & Cortina, J. L. (2014). Valorisation
of Ca and Mg by products from mining and seawater desalination brines for water treatment
applications. Journal of Chemical Technology & Biotechnology, 89(6), 872-883.
Climate ADAPT, consulted on August 19, 2023, https://climate-adapt.eea.europa.eu/en/metadata/
adaptation-options/desalinisation
DuPont, FilmTec™ FT-Norm software, 2021
DuPont, FilmTec™ Reverse Osmosis Membranes Technical Manual, 2023.
Fernández-Torquemada Y., A. Carratalá, J. L. Sánchez Lizaso. Impact of brine on the marine environment
and how it can be reduced. Desalination and Water Treatment 167, 27–37. (2019).
Fritzmann C., J. Löwenberg, T. Wintgens, T. Melin, State-of-the-art of reverse osmosis desalination,
Desalination 216.1 (2007) 1-76
Gilabert-Oriol G. Ultrafiltration Membrane Cleaning Processes: Optimization in Seawater Desalination
Plants. Walter de Gruyter GmbH & Co KG; 2021 Jun 8.
Greenlee L.F., D.F. Lawler, B.D. Freeman, B. Marrot, B., P. Moulin, P, Reverse osmosis desalination:
water sources, technology, and today’s challenges, Water research 43.9 (2009) 2317–2348
Guo W., H. H. Ngo,J. Li, A mini-review on membrane fouling, Bioresource technology 122 (2012),
27-34
Kadaj E., R. Bosleman, Energy recovery devices in membrane desalination processes. In Renewable
energy powered desalination handbook (pp. 415-444). Butterworth-Heinemann. (2018)
Kim J., K. Park, D. R. Yang, S. Hong, A comprehensive review of energy consumption of seawater
reverse osmosis desalination plants. Applied Energy, 254, (2019) 113652
Kucera J. Reverse Osmosis: Industrial applications and processes. Wiley; 2011.
Kumar M., T. Culp, Y. Shen, Water desalination: History, advances, and challenges. In Proc., Frontiers of
Engineering: Reports on Leading-Edge Engineering from the 2016 Symp,55-132 (2017)
Massons G., G. Gilabert-Oriol, S. Arenas-Urrera, J. Pordomingo, J.C. González-Bauzá, E. Gasia, M. Slagt.
Industrial scale pilot at Maspalomas I desalination plant demonstrates the efficiency of DuPont™
B-Free™ pretreatment–a new breakthrough solution against biofouling, Desalination and Water
Treatment, 1-5, 2022
Sablani S.S., M.F.A. Goosen, R. Al-Belushi, M. Wilf, (2001). Concentration polarization in ultrafiltration
and reverse osmosis: a critical review. Desalination, 141(3), pp.269-289.
Sustainable clean water through solar-powered desalination for water-scarce islands and coastal regions
(SDG: 2, 3, 6, 8, 11, 12, 14), Section #SDGAction42477, https://sdgs.un.org/partnerships/
sustainable-clean-water-through-solar-powered-desalination-water-scarce-islands-and,
consulted on August 19, 2023
Valavala R., J. Sohn, J. Han, N. Her, Y. Yoon, Pretreatment in Reverse Osmosis Seawater Desalination: A
Short Review, Environmental Engineering Research 16.4 (2011) 205-212
Wang L., J. He, M. Heiranian, , H. Fan,, L. Song, , Y. Li, M. Elimelech. (2023). Water transport in reverse
osmosis membranes is governed by pore flow, not a solution-diffusion mechanism. Science
Advances, 9(15), eadf8488.

70
 doi: 10.2166/9781789062977_0071

Chapter 4

Forward Osmosis
Maria Salud Camilleri-Rumbau, Aquaporin, Denmark/Eurecat, Spain
Xuan Tung Nguyen, Aquaporin, Singapore
Victoria Sanahuja-Embuena, Aquaporin, Denmark
Jan Frauholz, Aquaporin, Denmark/RWTH Aachen, Germany
Victor Augusto Yangali Quintanilla, Grundfos, Denmark
Alberto Tiraferri, Politecnico di Torino, Italy
Irena Petrinic, University of Maribor, Slovenia
Claus Hélix-Nielsen, Technical University of Denmark, Denmark

The learning objectives of this chapter are the following:

• To define principles of forward osmosis

• To define and apply forward osmosis parameters for assessing performance

• To present and discuss the basic equations governing forward osmosis performance
using typical experimental modes

• To understand the theoretical background of forward osmosis performance and

performance prediction using modeling tools.

4.1 INTRODUCTION: PRINCIPLES OF FORWARD OSMOSIS

Forward osmosis (FO) is an osmotically driven membrane technique which allows the
separation of water from a feed solution through a semi-permeable membrane using
osmotic pressure gradient as a driving force. Although during the last decade there have

© 2024 The Authors. This is an Open Access book chapter distributed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0),
(https://creativecommons.org/licenses/by-nc-nd/4.0/). The chapter is from the book Experimental
Methods for Membrane Applications in Desalination and Water Treatment, Sergio G. Salinas-Rodriguez,
Loreen O. Villacorte (Eds).
Experimental Methods for Membrane Applications

been important advances in FO in terms of material development and processes, the few
commercially available products and best practices for effluent processing makes the
standardization of the FO applicability challenging.

In terms of materials used for fabrication of FO membranes, cellulose acetate (CTA)

membranes and thin film composite (TFC) polyamide (PA) membranes are the most widely
known, both being commercially available (Xiao et al., 2017). New approaches using new
materials for FO purposes have also been developed. To name a few, these are double-
skinned membranes, membranes obtained by layer-by-layer techniques, mixed organic-
inorganic membranes and aquaporin-based membranes, which consist of TFC PA with
embedded aquaporins (Suwaileh et al., 2020).

Despite FO being a developing technology, in the recent years, there has been an increasing
interest for its use in industry. This opens countless opportunities for further developing
membrane configurations that can be used in an industrial setting. In terms of applications,
FO is a versatile membrane technique with a broad applicability spectrum within the water
treatment sector (desalination, municipal wastewater, industrial wastewater, potable
and non-potable water reuse, etc.), the management of process water (biorefineries,
pharmaceutical processes, etc.) and food and beverage processes (concentration of flavours
and aromas, juices, etc.).

Regarding membrane performance, FO relies on the osmotic pressure of the two solutions
separated by the semi-permeable membrane. The pass of water is allowed due to osmotic
gradient resulting in a concentrating process for the lower osmotic pressure solution
(known as feed) and a dilution process for the higher osmotic pressure solution (known
as draw solution). As for any other membrane-based process, water flux (Jw) and forward
solute rejection (R) can be obtained from experimental data (see section 4.3.1); while a
parameter known as reverse solute flux (Js), unique for FO, allows to track the loss of draw
solute into the feed solution.

During FO processing, concentration polarization can severely affect membrane performance

in both feed and draw sides of the membrane. Both external concentration polarization
(ECP) at the feed side of the membrane as well as internal concentration polarization (ICP)
at the draw side of the membrane, play a detrimental role that cannot be overlooked when
interpreting the FO process performance and when evaluating its applicability (see section
4.3.2. for more details). Concentration polarization can act in combination with fouling,
scaling and/or a combination of both fouling mechanisms, to decrease the net driving force
available for mass transport.

With this Chapter, the authors have put together the most relevant experimental practices
in the FO field in order to guide researchers and engineers towards getting hands-on
experiences in FO.

72
 Chapter 4

4.2 MATERIALS AND EXPERIMENTAL SET-UP

4.2.1 Membrane configurations

Membranes for FO processes and generally for liquid-liquid separations, can be found
commercially either as flat-sheets or hollow tubes. The modules of these can be classified,
similarly as done for other membrane processes, in four module types: plate-and-
frame, spiral-wound, hollow fiber and tubular. The first two types are made of flat-sheet
membranes, while the last two types are composed of hollow tube membranes.

Plate-and-frame modules are the simplest module configuration where membrane sheets
are mounted in frames closely together. The feed solution to be treated passes alongside the
sheets surface and it gets collected at the end as a concentrated retentate, while the permeate
is collected in its own channel. Spiral-wound are membrane sheets rolled in alternating
order together with turbulence promoting plastic grids, called spacers. The feed solution
is introduced at one end of the module and flows axially on the active layer and feed spacer
side of the membrane. The permeate is collected in the envelope and led to the permeate
channel which is centrally located. Hollow fiber modules consist oppositely of tubes, packed
closely together and placed inside a vessel. Here, the feed solution passes through the lumen
of the hollow fiber, permeates through the membrane towards the shell side and exits the
module. When their inner diameter rises to 5 mm or above, and the module packing density
significantly decreases, the module is known as a tubular module.

There are intrinsic advantages and disadvantages of using each module configuration and
their usage would depend on the intended process application. For example, plate-and-
frame and tubular configurations are usually used with extremely high-particulate streams
and/or high-viscosity solutions. The simplicity of the plate-and-frame configuration
allows for high cross-flow velocities, reducing fouling by increasing longitudinal shear
stress. Additionally, these membranes can be easily cleaned which increases the lifetime of
the membrane. However, this module type is expensive to manufacture, and the packing
density is low, which increases the membrane installation footprint considerably. Similarly,
tubular modules can be more tolerant to fouling and clogging due to the large inner diameter
of their hollow tube and the possibility of operating at high cross-flow velocities. However,
it suffers from the same disadvantages as its counterpart, the plate-and-frame module due
to a low packing density. Spiral-wound and hollow fiber configurations are thus the most
commonly used module configurations for membrane-based liquid-liquid separation and
FO processes, as they can cover a broad range of applications by balancing effectiveness and
price.

4.2.2 Experimental modes

In general, there are three main experimental modes, regardless of process configuration
being single-stage or multi-stage, that can be defined when working with FO membrane
processes:
- Single pass mode
- Batch-batch mode
- Semi-batch mode

73
Experimental Methods for Membrane Applications

The main differences between these operational modes are related to how the feed and
draw solutions are being processed within the membrane module. For example, in single
pass mode, the feed and draw solutions enter the module and have a unidirectional contact
area across the membrane, where the solutions are not recirculated. The feed recovery
or concentration is achieved in one pass. Oppositely, in batch mode, both feed and draw
solutions are recirculated to their respective holding tanks. This means that the feed
solution gets increasingly concentrated in the feed tank, while the draw solution gets
diluted with time. During semi-batch mode operations, one of the two solutions, either
feed or draw solutions, are run in a batch mode while the other solution runs in single pass
mode. Typically, the feed will run in a batch mode, allowing for continuous concentration
during processing, while the draw solution will run in single pass to minimize loss in FO
performance due to dilution of the draw solution.

The advantages of the operational modes above depend on the type of feed, on the process
that needs to be undertaken, and on the objective of the FO system. For instance, during
food valuable concentration processes, a semi-batch mode will be preferred, while a batch-
batch mode might be preferred during concentration of secondary effluent by using sea
water brine as a draw solution.

Test setup – Semi-batch mode (feed in batch mode vs. draw in single pass mode)
Figure 1 shows the schematic outline of a semi-batch FO setup. The draw solution will
become diluted during the feed concentration process. The draw outlet can be discarded to
the drain. The feed solution will be continuously recirculated to the FO membrane module
and thus concentrated.

T T
Draw inlet
Sample C C tank D1
point Sample
F point
F
P Transmembrane
Draw outlet Pressure can
pressure: P be adjusted by
tank D2 minimum 0.2 bar
opening/closing Clean water
T C F P P
needle valve on tank C
retentate line
FO module

Sample
retentate line
needle valve on
opening/closing C F
be adjusted by
Pressure can
For auto
point logging
T
Sample
Feed point
Feed inlet
Inlet
tank
TankF1F For recirculation
For auto
retentate line
needle valve on
opening/closing

Weighing logging
be adjusted by

Clean water
Pressure can

tank C balance

Figure 1 Setup for FO semi-batch operation

74
 Chapter 4

Recommendations for running an FO application test:

1. To select the type and strength of the draw solution, refer to section 4.2.3.
2. Start the feed pump to fill in the system. Afterwards, start the draw pump and adjust to
the operating conditions as indicated by the FO membrane/FO system manufacturers,
always ensuring that the system operating conditions are in agreement with the
membrane manufacturer operating limits. Ensure removal of air in the system. The
operating conditions can be modified during the application test. For example, the feed
inflow can be increased to enhance shear on the membrane surface and delay fouling,
while the draw inflow can be increased if flux (Jw) falls below 1 L/m2h.
3. TMP must always be kept positive where possible. For this, it is crucial to monitor that
the feed inlet and outlet pressure readings do not overcome the maximum TMP and
inlet pressure specified by the membrane module manufacturer. See more details on the
effects of negative TMP in section 4.3.3.
4. To be able to calculate compound mass balances, weight of samples taken from the feed
inlet, feed outlet and draw outlet should be considered throughout the concentration
process. This will allow monitoring of Jw and water recovery values accurately during the
FO process.
For further recommendations on process parameters and constraints, refer to Table 1.

Table 1 Process parameters to be considered during FO application test

Process parameter Process values - considerations
Recommended (and maximum) application flow Refer to FO membrane/system manufacturer’s
rates, L/h recommendations.
Minimum feed outlet flow, L/h Refer to FO membrane/system manufacturer’s
recommendations. The user needs to make sure that the
feed outlet has a minimum flow to avoid module damage.
TMP, bar Refer to FO membrane/system manufacturer’s
recommendations. Generally, it should be around or just
above 0 bar.
Feed and draw inlet pressure, bar Refer to manufacturer’s recommendations on pressure
tolerance for the specific FO product.
Recommended (and maximum) temperature, ˚C Refer to manufacturer’s recommendations. Generally, the
FO module should be able to run feed and draw solutions
at room temperature. Be aware that increases in operating
temperature could affect FO performance, e.g., increased
Jw and Js. The module temperature tolerance should not be
surpassed. Refer also to manufacturer’s recommendations
for CIP procedures.

Data collection during the application test

The data to be collected during the test is shown in Table 2. For detailed calculations, refer to
equations on ‘Typical parameters and phenomena’ in section 4.3.1.

75
Experimental Methods for Membrane Applications

Table 2 Process parameters to be measured during the semi-batch application test and where to
make the related readings
Purpose/calculated values Process parameters Where to measure
Water flux (Jw) Flow/Weight Feed bulk
TMP Pressure Feed inlet/outlet, draw inlet/
outlet
Feed inlet pressure Pressure Feed inlet
Maintain stable temperature (T) Temperature Feed outlet (if measuring
conductivity at feed outlet)
Osmotic pressure (indirect Conductivity Feed outlet
measurement)
Ensure minimum feed outlet flow Flow Feed outlet

4.2.3. Draw solutions: properties, regeneration, types and selection criteria

The performance of FO applications depends on the draw solution, which provides the
driving force for water permeation. An adequate choice of draw solute agent can maximize
the water flux (Jw) and the water recovery of the system. In addition, the reverse solute flux
(Js) can be reduced and the regeneration costs can be lowered, which usually represents
the largest operational costs in FO applications. Therefore, this subsection will give a
short summary of the most important properties of draw agents, their influence on the
membrane performance, available draw regeneration methods, and discuss advantages and
disadvantages of different classes/types of draw agents. Eventually, guidelines regarding the
selection of suitable draw solutes are given.

Main properties of a draw solution

Osmotic pressure
The driving force in forward osmosis processes is provided by the difference in osmotic
pressure across the active layer of the membrane, as defined as follows:

RT RT
Π=− ⋅ ln(aw ) = − ⋅ ln(γ w χ w ) Eq. 1
Vm Vm

Where R is the gas constant, T the temperature, Vm the partial molar volume of water, aw the
water activity, γw the activity coefficient, and cw the mole fraction.

In practice, assumptions are made to simplify osmotic pressures estimation, often also due
to unknown activity coefficients. In very dilute solutions the solvent activity coefficient can
be assumed to be close to 1, resulting in the validity of the van’t Hoff equation, as follows:

= cm RT Eq. 2

where cm is the osmotic concentration (osmolarity). The osmolarity is the molar

concentration of osmotic active solutes. For a salt solution, such as NaCl which dissociates
into two ions, the osmolarity equals twice the molarity assuming complete dissociation.

76
 Chapter 4

Even though the van’t Hoff’s equation is formally only valid for very dilute solutions,
it often provides acceptable accuracy for salt-based draw solutions, especially if direct
measurements of the osmolarity (e.g., freezing-point or vapor pressure osmometry) are
available. In comparison, prediction of osmotic pressures of organics, polymers, or other
draw solutes can lead to significant deviations.

Diffusivity and viscosity

In most technical applications, the draw solution is applied on the support layer side of
the membrane leading inevitably to driving force losses due to concentration polarization
effects (see section 4.3.2). Besides the properties of the membrane (structural parameter),
the extent of ICP is mainly related to the draw solution concentration and diffusivity. The
use of a higher draw concentration, hence nominal driving force, does not correlate with a
linear increase in the water flux. A higher draw concentration leads to larger relative driving
force losses due to stronger ICP related to the convective transport of draw solutes away
from the active layer. Since the transport of draw solutes towards the active layer is diffusive
only, the diffusivity of the draw agent significantly influences the extent of polarization
effects. Diffusivity depends mainly on the solution’s viscosity and the diffusion coefficient.
Overall, draw solutions of low viscosity and high diffusion coefficient are the best choice
when considering FO system productivity.

Regeneration of draw solutions

Continuous FO processes require a reconcentration of the diluted draw solution. Up to
now, the regeneration process remains the bottleneck in the draw solution selection.
Among the most studied draw regeneration methods are membrane-based processes,
such as reverse osmosis (RO), nanofiltration (NF), and membrane distillation, as well as
evaporative technologies and electrodialysis. While pressure-driven membrane processes
are limited in achievable osmotic pressures due to allowable pressure, the evaporation of
water is highly energy intensive. This has led to ideas for the implementation of FO within
applications that do not require a draw regeneration. Important to mention are (so-called
direct) FO processes using seawater or a concentrated fertilizer as draw solution, which may
be discarded or beneficially used once diluted, to avoid costly regeneration. Furthermore,
novel types of draw solutions have been designed to overcome existing challenges in draw
regeneration. These so-called responsive draw solutions exploit drastic changes in physical
and chemical properties of the draw agent provided by external stimuli, such as heat or pH,
enabling a practical and efficient draw regeneration. Please note that the energy required
for draw agent regeneration is always somewhat related to its target osmotic pressure
due to thermodynamics considerations, thus simplicity of regeneration should not be
confused with cost of regeneration and the two issues should be considered separately and
simultaneously to design a feasible and effective FO system.

Types of draw solutes

In theory, any water-soluble component exhibiting an osmotic pressure can be used as a
draw solute. Considering the above-described influence of draw properties on the process
performance, small solutes of high osmotic pressures (high solubilities) are preferred. A
variety of different draw solutes including salts, small organic molecules (e.g., sugars),
volatile organic compounds, nanoparticles, polymers, or hydrogels have been investigated
up to now.

77
Experimental Methods for Membrane Applications

Among the most studied and applied draw solutes are inorganic and organic salts offering
the advantages of high osmotic pressures, low viscosities, high diffusivity, electrical charge,
and low toxicity. As a result, salt-based draw solution can reach a high water flux, exhibit
comparable low driving force losses, enable a hazard-free operation and simple regeneration
by pressure-driven membrane processes such as reverse osmosis. Most salts are inexpensive,
available as food grade quality, and their replenishment costs are low.

Due to the larger variety of salts, salt-based draw solutions can be selected with regard to the
specific process requirements. For example, multivalent ions of higher molecular weights
can be selected for food and beverage application to minimize the reverse solute flux into
the product stream. Furthermore, studies indicate that the rejection of feed compounds can
be enhanced by selecting an appropriate draw solute.

Worth mentioning is sodium chloride (NaCl), one of the most studied draw solutes. It is
often used as a benchmark draw agent to evaluate membrane and process performance.
Many membrane manufacturers and published articles use the specific reverse solute flux
(Js/Jw) of NaCl as key characteristic to account for the membrane’s selectivity for water
transport.

Besides the comparable high reverse solute flux of salts, their regeneration remains
the bottleneck of salt-based draw solutions due to osmotic pressure limitation. Novel
approaches such as osmotically-assisted reverse osmosis (OARO) may enable higher draw
concentrations in the future (Peters and Hankins, 2020).

To exceed the osmotic pressure limitation of conventional draw solutes, a variety of

responsive draw solutes which can switch solubility properties by external stimuli were
studied. Among the most studied responsive draw solutions are thermo- and CO2-
responsive draw agents. Thermo-responsive draw solutions exploit temperature-dependent
miscibility gaps between water and polymers or ionic liquids. By exceeding the lower
critical solution temperature, the diluted draw solution separates into two phases whereof
one phase is rich in draw agent and the other is water-rich. CO2-responsive draw solutions
undergo acid base reaction and are often amine-based. Amines can react reversibly with
CO2 and form bicarbonate salts which can be used as draw agents. Upon heating or purging
with inert gas, the diluted amine bicarbonate draw solute decomposes into either a water-
insoluble liquid or gaseous amines such as trimethylamine or ammonia. Disadvantages
of responsive draw solutions are related to the costs, toxicity (amines), or low membrane
performance due to severe ICP (polymers).

Selection of draw solutes

The right choice of draw agent depends on the specific application and feed stream, the
target recovery, the process configuration, availability and costs of different energy forms
(electrical, waste heat, ...), space requirements, commercial assessment, as well as further
considerations. The following guidance can assist in selecting an adequate draw agent:

78
 Chapter 5

1. Osmotic pressure of the draw solution

- Too low osmotic pressure of the draw solution induces low membrane performance,
prolonging the process time or requiring larger membrane areas. Both may contribute
to target compound losses from the feed solution as well as draw solute contamination
of the feed solution.
- Too high osmotic pressures can accelerate membrane fouling and scaling. Regeneration
costs can increase due to dissipation of osmotic potential.
- Rule of thumb: ratio in osmotic pressure between the draw solution and concentrate
should not be below 1.1 - 1.2
2. Draw regeneration method:
- The draw regeneration process is often limiting the choice of draw type and strength
- Reverse osmosis:
• RO is the most energy efficient process to regenerate draw solutions
• Limited by hydraulic pressure (65 bar, high-pressure RO: 120 bar)
• Osmotically-assisted reverse osmosis is not established but can overcome osmotic
pressure limitations
- Evaporators
• Energy intense regeneration with no limitations regarding osmotic pressures
• Corrosive draw solutes (e.g., chlorides) can drastically increase the CAPEX due to
material requirements and should be avoided
- Electrodialysis
• High CAPEX
• Limited to ionic draw solutes
• Energy efficient at low osmotic pressure range
- Membrane distillation
• Energy intensive regeneration, but often low grade heat sources can be used
• Not-yet-established technology due to currently low performance and specific
current limitations related to module configurations, fouling/scaling, and long-
term stability
- Responsive draw recovery
• Still under development, offering the potential to concentrate draw solution to
high osmotic pressures
• Energy intensive, but often enabling the utilization of low-grade heat sources
3. Applications:
- Food and beverage
• Only food-approved draw solutes are applicable (sugars, salts)
• Multivalent ions and agent of higher molecular weight can reduce unwanted
reverse solute flux
• Target compound rejection can be increased by selecting draw solutes which are
already present in the feed stream
- Wastewater concentration
• Lower concentration of draw agents may be beneficial to reduce fouling propensity
• Target compound rejection can be increased by selecting draw solutes which are
already present in the feed stream

79
Experimental Methods for Membrane Applications

4.3 EXPERIMENTAL METHODS

4.3.1. Typical parameters and phenomena

The most important process parameters are the water flux Jw, the reverse solute flux Js, the
specific reverse solute flux Js/Jw, the recovery, and the rejection (both forward and reverse).

The water flux Jw is defined as the areal permeation rate of water as follows:
Q permeate Q feed Qconcentrate Qdraw out Qdraw in
Jw = = = Eq. 3
Amembrane Amembrane Amembrane

Where Q is the flow rate and the active membrane area (A) is in the denominator. As seen
above, Jw can be calculated based on the difference in feed in- and outlet flow rates as well as

based on the difference on the draw side. Its unit is L/m2h. In batch operation, Jw can be
determined by measuring the change in feed or draw weight under the assumption that only
water permeates the membrane.

The reverse solute flux Js is defined as follows:

msolute
Js = Eq. 4
Amembrane

Where the mass flux of draw solute is in the numerator and the active membrane area
(A) is in the denominator. The reverse solute flux is determined by measuring the draw
solute concentration in the concentrate stream. Depending on the feed composition, an
appropriate measurement of draw solute concentration, such as conductivity, ICP-OES, or
HPLC, can be used. The reverse solute flux is usually given in g/m2h.

The specific reverse solute flux Js/Jw is defined as the ratio between Js and Jw. It is a measure
of the selectivity for water permeation over draw solute transport given in g/L.

The transmembrane pressure (TMP) is defined as the average hydraulic pressure between
the feed side and the the draw side of the membrane, given as follows:
( p feed + p feed out ) ( pdraw in + pdraw out )
TMP = Eq. 5
2

The recovery (Rec) defines the ratio of the volume of recovered water to the volume of feed
solution. In single-pass operation the membrane recovery is defined by using the permeate
and feed flow rates as follows:
Q permeate Qconcentrate Q Qdraw in
Recmembrane = ×100% = 1 100% = draw out 100% Eq. 6
Q feed Q feed Q feed

80
 Chapter 4

In batch processes where the feed solution is constantly concentrated, the recovery is
defined as follows:

Vfeed (t)
Rec(t) = 1 ×100% Eq. 7
Vfeed (t0 )

Assuming only water to permeate the membrane and a constant density of the feed solution
rfeed(t) allows calculating the recovery by weights instead of volumes.

In food and beverage processes, concentration factors (CF) are often used instead of recovery,
where CF is:

1
CF = Eq. 8
R
1−
100%
The average membrane forward rejection R of a compound i (moving forward from the feed
to the draw side) is commonly defined using the concentration ratio between permeate and
feed. To take the concentration difference between incoming feed and outgoing concentrate
stream into consideration, the average concentration on the feed side of the membrane is
often used:

ci, permeate
Ri = 1 ×100%
Eq. 9
ci,concentrate + ci, feed
2

In contrast to most other membrane applications, the permeate concentration cannot be

directly measured due to its dilution by the draw. Therefore, its average must be calculated
based on a mass balance of component i on the draw side.

Taking the draw flow rate into consideration leads to:

ci, permeate × Q permeate = ci,draw out × Qdraw out ci,draw in × Qdraw in

Inserting equation 1 (Jw) and rearranging leads to:

ci,draw out × Qdraw out ci,draw in × Qdraw in
ci, permeate = Eq. 10
J w × Amembrane

While in (draw) batch operation the ingoing target solute concentration in the draw solution
needs to be considered, in single-pass operation Ci,draw in is in most cases negligible.

The achieved membrane forward rejection depends on the membrane type, operation
conditions (e.g., flow rates of draw and feed solutions), the water flux, as well as the recovery,
and it is different for different compounds.

81
Experimental Methods for Membrane Applications

It is important to note that the above membrane rejection calculations consider the observed
rejection and not the real compound rejection, as it is calculated considering:
1) The total mass that has passed through the membrane during the entire pass and the
average Jw (or entire time in batch mode) and not the mass that is passing across the
membrane in each location along the module (or at any given time in batch mode);
2) The feed or draw bulk concentration and therefore not considering the higher compound
concentrations reached at the active layer interface due to the polarization phenomena.
Larger molecules are often better rejected than smaller ones. In addition, uncharged
organic molecules show lower rejection than charged molecules due to missing
electrostatic repulsion. Even during batch concentration processes the rejection may
change significantly as shown in the case of urea (Figure 2).

Feed: Urea 200mgL-1

100

80
HF-O

60

HF-C
40

20

Rejection (% ) 0
0 20 40 60 80 100
Recovery (%)

Figure 2 Urea rejection variation with recovery rate for HF–C (chlorinated membranes) and HF–O
(non-modified membranes). Adapted from Sanahuja-Embuena et al. (2019).

4.3.2 FO process design constraints and considerations

To design a specific FO process and experimental setup, users are strongly advised to refer
to the manufacturer’s FO module datasheet to understand the operating limits of the given
modules. It is also recommended that the user reads any other documentation provided by
the FO manufacturer.

Concentration polarization (ECP/ICP)

As seen in Figure 3, concentration polarization occurs on both sides of the membrane due to
the permeation of water concentrating the feed solution while diluting the draw solution.
A distinction is made between external concentration polarization (ECP) on the active layer
side of the membrane and polarization effects in the membrane’s support layer referred to
as internal concentration polarization (ICP). Depending on the membrane orientation, i.e.
FO mode (where the feed solution is in contact with the active layer) or PRO mode (where
feed solution is in contact with the membrane support layer), these polarization effects can
either be dilutive or concentrative. In most applications, the draw solution is applied on the
support layer side leading to dilutive ICP and concentrative ECP.

82
 Chapter 4

Concentration polarization reduces the difference in osmotic pressure across the active
layer and leads inevitably to driving force losses for water permeation. Besides driving force
losses, concentrative ECP increases the risk of membrane fouling and scaling. Lower water
fluxes as well as turbulent flow conditions can contribute to reducing these risks.

The intensity of dilutive ICP depends on the porosity, tortuosity, and thickness of the
support layer (see structural parameter in section 4.4.1) as well as on the diffusivity of the
draw solutes and the water flux. Since draw solutes diffuse against the convective water flux,
draw solutions of low viscosity and high diffusion coefficients can mitigate dilutive ICP (see
draw solution in section 4.2.3). Additionally, highly porous, and thin support layers can
lower the extent of driving force losses.

AL-DS membrane orientation can significantly decrease dilutive CP of the draw solution.
Since concentrative CP of feed solutes in the support layer is increased, this membrane
configuration might only be beneficial in specific applications, where the feed presents low
fouling potential).

a) Active layer b) Support layer

Support layer Active layer

Feed solution C5 Feed solution C4

C4 C3

Jw Jw
∆πm ∆πb ∆πm ∆πe ∆πb
C3
∆πc C2
Draw solution C1
C2 Draw solution
Js
C1
Js

Figure 3 ECP and ICP at a) AL-FS mode, and b) AL-DS mode. Adapted from Wang and Liu (2021).

Pressure limit
Pressure limit is one of many important factors to consider as it affects the choice of flow
rate and cross flow velocity sent into each element. This typically already translates into the
recommended flow rate range on both feed and draw side. In addition, how much water is
transported into the draw solution side is primarily a function of draw solution flow rate
and concentration. Even at low draw solution inlet flow, high osmotic pressure difference
may result in a large water permeation rate and hence a higher flow rate on the draw side.

System projections are therefore useful to predict the behavior of pressure drop on both
feed and draw lines. However, calculating pressure drops in a system can be complicated as
this will depend on several factors such as module geometry, array configuration and liquid
properties among other factors. Few considerations need thus to be taken when projecting:
1) permissible pressures given by membrane manufacturer should not be exceeded, 2) TMP

83
Experimental Methods for Membrane Applications

usually increases during batch concentration (e.g., viscosity increase of feed, feed outflow
rate increases due to a lower Jw), or in continuous mode due to fouling, 3) system arrays
require special considerations such as accounting for local changes in pressure drops,
pressure build-up when more-than-one modules are connected in series, draw solution fed
in the system in series or in parallel, etc.

Flow rate limit

Flow rate limit, by extension, is determined by the maximum pressure limit of the module.
Flow rate should be selected within manufacturer’s recommendation in order not to exceed
pressure limit. In addition, users are advised to check for any minimum feed reject flow
requirement by manufacturers. In a batch process or semi-batch, feed and bleed process,
recovery of feed is time-dependent and not flow-dependent. It is therefore possible to
maintain as high cross-flow velocity as possible, while staying within pressure limit,
to minimize risks of fouling and scaling. This is especially so when feed streams contain
medium to high degree of foulants. In a single-pass continuous process, recovery of the
feed is flow-dependent. Designers of the FO process should determine, through projection,
whether concentrate flow rate at the last in-series element is below recommendation.

Moreover, draw flow rate in operation should be carefully selected and monitored because
it influences transmembrane pressure, permeation flux across FO membrane, and the
concentration of draw agent and of possible compounds permeated from the feed side.
Usually, for polyamide-based FO membrane, manufacturers may recommend a safe limit of
negative TMP, beyond which there poses a risk of delamination of polyamide active layer.
Having a high draw flow rate increases the overall permeation. However, an excessively high
draw flow rate might raise pressure on the draw solution side and result in a high chance that
the negative TMP limit is exceeded.

Flow direction
Flow direction, whether counter-current or co-current, is also a tool available for FO process
designers. In co-current operation, feed and draw solutions enter the module through the
same end of the module, leading to a constantly reducing driving force along the module
length. Counter-current operation enables to maintain a more constant osmotic pressure
difference along the lengths of the module (see Figure 4). Additionally, counter-current
operation maximizes the average water flux across the FO module or system and the
permeate recovery, while minimizing local differences in water flux. This means that the
difference between water flux across FO membrane across inlet and outlet of FO system is
less for counter-current, as compared to co-current flow direction.

It should however be noted that the selection of the flow mode (i.e., co-current or counter.
current) depends on module type. For spiral wound or some plate-and-frame module type,
flow path is designed to be in cross flow, where feed and draw solutions are perpendicular
to each other. For hollow fiber and tubular membrane type, flow path can be selected to be
counter-current or co-current.

Conventionally, a counter-current flow path is the is the preferred option in most

applications and experimental setups as it allows maximization of the driving force.

84
 Chapter 4

In practice, FO process designers should pay attention to ease of filling up the shell side
chamber in counter current mode, assuming that the module is mounted vertically (i.e., feed
side flow is upwards and draw side flow is downwards). Modules of larger size which are
mounted vertically may however require ingoing streams to enter on the bottom side of
the module to remove any trapped air from the module. In such a case, if the draw inlet
flow rate is too small, partial filling of shell chamber may occur resulting in underutilized
membrane area. In this specific case, operating in co-current operation may be advantageous
even though process performance is reduced.

a)

Drawin Drawout Drawout Drawin

Feed Concentrate Feed Concentrate

Co-current Counter-current

Draw Draw
∆π ∆π
Feed
Feed

π π
Module length Module length

b) Feed: DI H2O
40 1.5

1.2
30
Jw-HF-C Jw-HF-C
0.9
JwHF-O JwHF-O
20
0.6
Js/Jw HF-C
10
0.3
Js/Jw HF-O

Jw (Lm-2h-1) 0 0.0 Js/Jw (gL-1)

Counter-current mode Co-current mode

Figure 4 a) Driving force for counter-current and co-current flow direction; b) Jw and Js/Jw for
HF–C (chlorinated membranes) and HF–O (non-modified membranes) in co-current and
counter-current when DI water was used as FS. Operating conditions were: Feed flow
rate was 100 L.h-1, draw flow rate was 25 L.h-1, draw concentration was 1 M NaCl and
TMP was 0.2 bar. (n = 2). Adapted from: Sanahuja-Embuena et al. (2019).

Limiting flux
Lastly, water permeation limit or design flux limit is a major factor affecting the FO design.
On the one hand, this is related to feasible feed inlet flow rate for FO module and system. A
lower FO feed inlet flow rate limit by manufacturers’ recommendation or by system design

85
Experimental Methods for Membrane Applications

indicates lower maximum design flux limit. However, this is in fact generally related to
fouling potential and reversibility of fouling.

While there is no consensus on what design flux limit for FO membrane should be, there
are research reports indicating limiting flux to be between 10-20 LMH and designed flux
for reversible fouling to be 5-15 LMH. Here, limiting flux is defined to be the starting flux
value at which there is decline of flux over time at constant osmotic driving force difference.
Design flux is defined to be the starting flux at which there is minimal flux decline and there
is flux restoration upon cleaning if there is any flux decline over time. These flux values vary
depending on membrane type and feed quality or foulants present in feed.

Water permeation limits may be controlled by draw inlet flow rate and draw inlet
concentration. As mentioned above, a higher draw inlet flow rate means less dilution effect
on the draw side, allowing osmotic driving force to be sustained from inlet to outlet of
module. This comes with the drawback of having a higher pressure drop on the draw side.

A higher draw inlet concentration means higher osmotic driving force across entire FO
modules or system, at the same draw inlet flow rate. The disadvantage is the high likelihood
of exceeding design or limit FO flux at certain sections of FO membrane within a module
or system. This may lead to sustained high ECP in those regions, increased likelihood of
fouling and scaling and premature module failure.

4.3.3 Best practices

Transmembrane Pressure (TMP)

Most forward osmosis membrane suppliers recommend running FO processes under low
positive transmembrane pressure. The positive TMP can hinder the transport of draw
solutes towards the feed solution due to the pressure gradient, which helps in preventing
the immediate contamination of feed solution by draw solutes in the event of membrane
breakage or defects on the selective layer. In the case of small defects on the polyamide layer,
a positive TMP will also be beneficial. However, a positive TMP may also aid the transport
of feed solutes into the draw solution and thus, the quality of the selective layer would need
to be checked. A tight and highly cross-linked polyamide layer should not be significantly
or drastically affected by slight positive and negative TMPs, and if this happens, it may be a
sign of membrane deterioration.

Nevertheless, a negative TMP should be strictly avoided, even for brief periods of time, due
to the polyamide layer configuration (where the layer is on the lumen side of the membrane).
When a negative TMP is applied, the pressure gradient direction can cause the delamination
of the polyamide layer, and consequently, the breakage of the membrane.

During the FO module operation, pressure losses from inlet to outlet for both feed and draw
side are expected, regardless of the flow mode selected (i.e., counter-current or co-current),
which could provoke negative TMP at the feed outlet or draw inlet locations. It is therefore
of paramount importance to maintain a positive TMP at the feed outlet ensuring following
the manufacturing guidelines on pressure limits.

86
 Chapter 4

In summary, since FO is a virtually pressure-less membrane process, membranes are not

designed for high hydraulic pressures on either side of the membrane. Therefore, commonly
recommended TMPs are around 0.2 bar. Allowable pressures given by the membrane
manufacturers should not be exceeded to ensure safe operation. Here, pressure relief valves
in the experimental setup can protect the membrane from maloperation.

Avoiding ‘over-recovery’
High recoveries of feed solution can lead to the precipitation and deposition of feed particles
on the membrane (fouling and scaling). While membrane fouling is characterized by the
deposition of (mainly organic) suspended solids, scaling refers to the precipitation and
crystallization due to exceeding salt solubilities. In process configuration consisting of serial
connected FO modules, fouling will occur in the first stages while scaling usually occurs in
the consequent stages.

Although FO is generally considered a low fouling propensity membrane technology which

can handle more difficult-to-treat feed stream, fouling and scaling will ultimately reduce
the membrane performance. Indications are a reduced water flux, increased pressure drops
on the feed side of the membrane, as well as reduced rejections. Besides an appropriate pre-
treatment of the feed solution to remove suspended solids, frequent cleaning-in-place (CIP)
can mitigate the deposition of solids on the membrane surface and performance detriment.
Scaling should be prevented by estimating the scaling risk of a certain feed composition by
using the Scaling Index and avoiding working at water recoveries that could provoke severe
compound precipitation.

4.4 DATA ANALYSIS: BASIC FO PROCESS DESIGN

4.4.1 FO Fundamental Equations

In a typical FO process, the equation for water flux flowing from feed side to draw side is
given by:

J w = A⎡⎣( PF − PD ) = π Dm − π Fm ⎤⎦
( ) Eq. 11

Where Jw is water permeation flux, A is the water permeability, PF and PD is hydraulic

pressure of feed side and draw side respectively, and π is osmotic pressure of draw side and
feed side at membrane surface, respectively. In FO operation, hydraulic pressure difference
tends to be zero or close to zero.
The salt flux equation is given by:

( )
JS = B C Dm C Fm = B Cm Eq. 12

where Js is sat flux from draw to feed, B is the salt permeability, and C is solute concentration
in draw and feed solution at membrane surface, respectively.

87
Experimental Methods for Membrane Applications

The salt transport across the FO membrane is also described by the convection-diffusion
model with a diffusive term proportionally related to solute concentration gradient and
a convective term related to water permeate flux across the membrane in the opposite
direction.
dC(x)
Js = D J wC(x) Eq. 13
dx
Where D is the solute diffusion coefficient. The solution of the transport equations above
differ depending on the orientation of the membrane.

In active layer facing feed side (AL-FS) mode or FO mode, water permeates from feed side into
the support layer on the draw side, leading to dilutive internal and external concentration
polarization (i.e., ICP and ECP, respectively). On the feed side, the convective water flux
carries solutes from bulk feed solution to membrane surface, at which they are rejected and
accumulate, causing concentrative ECP. The solution of the convective-diffusive equation
above, for AL-FS mode, become:

Jw 1 S
C Db exp C Fb exp J w +
kD k F DF Eq. 14
Cm,ALDS =
B 1 S Jw
1+ exp J w + exp
Jw k F DF kD
For active layer facing draw side (AL-DS) mode or PRO mode, water permeates from
feed side with solutes that are rejected and accumulate across the support layer, resulting
in concentrative ICP and ECP on the feed side. On the draw side, there is dilutive ECP as
pure water permeates into the draw side. The solution of the convective-diffusive equation
above, for AL-FS mode, becomes:

Jw 1 S
C Db exp C Fb exp J w +
kD k F DF
Cm,ALDS = Eq. 15
B 1 S Jw
1+ exp J w + exp
Jw k F DF kD

Where k is mass transfer coefficient, and the term ‘exp(-Jw/kD)’ indicates external
concentration polarization in general whereas the term exp[Jw(1/kF + S/Df)] denotes internal
concentration polarization with S being structural parameter of membrane, consisting of
porosity and tortuosity term used in modifying solute diffusion coefficient from the bulk
solution to the inside support layer.

The mass transfer coefficient k value is dependent on the type of membrane form factor and
module. In general, mass transfer coefficient is:
Sh × D
k= Eq. 16
dh

88
 Chapter 4

where Sh is Sherwood number and dh is hydraulic diameter, both being geometry-

dependent.

Combining above stated equations, one is able to calculate the expected water permeation,
Jw, and reverse solute flux, Js, of a FO membrane, given its bulk feed and draw solution
characteristics and some basic hydrodynamic information to obtain mass transfer
coefficients.

4.4.2 FO Module Mass Balance

To simulate transport inside a membrane module, mass balance equations should be
considered. In addition, the effect of volume change due to dissolved solute should also be
taken into account. This means the differential term of density and concentration of solute
cannot be neglected.

Typically, mass balance equations for pressure, velocity and concentration along module
length can be established. For instance, the velocity and concentration differential equation
on the feed side can be seen below for a rectangular flat plate channel type.

w d F× 1
Jw× + Js J ×
dv F
dc F s H Eq. 17
= F
dx F ,bulk d
×cF
F
dc

dv F 1
F cF × + ×J
dc dx H s Eq. 18
=
dx vF

Where rW is density of pure water, vF is the differential term to account for volume change
with solute concentration and H is the height of flat plate flow channel. In other geometries,
such as for hollow fiber or tubular types, these terms are referred to inner diameter of hollow
fiber (this assumes an inside out FO module with active layer being on the lumen side).

Similarly, the velocity and concentration differential equation on the draw side for a
rectangular flat plate channel type can be seen below.

w d D× 1
Jw × Js + J × Eq. 19
dv D
dc D s H
=
dx D d D D
×c
dc D

dv D 1
D cD× + ×J
dc dx H s Eq. 20
=
dx vD

89
Experimental Methods for Membrane Applications

For a hollow fiber bundle, the H hydraulic radius term becomes the following for lumen and
shell respectively:
Hlumen = 4/di Eq. 21

Hshell = (4 × n × di)/(n × do2 - Di2) Eq. 22

where di is fiber inner diameter, do is fiber outer diameter, Di is shell housing inner diameter
and n is the total number of fibers.

Similarly, the velocity and concentration differential equation on the draw side for a
rectangular flat plate channel type can be seen below.

w d D 1
Jw × Js + ×J ×
Eq. 23
dv D
dc D s H
=
dx D d D D
×c
dc D

dv D 1
D c D× + ×J
dc dx H s
= Eq. 24
dx vD
It should be noted that the sign of the velocity differential equation is reversed in the event
of counter current flow.

The pressure drop equation across the module strongly depends on the type of module
used. As an example, for the hollow fiber form factor, the analogy of flow through a packed
bed with the Ergun equation could be used to model pressure drop across the tube bundle
on the shell side.

150 (1 ) ×μ
2 D D 2
dP D ×v 1.75× (1 )× D
×vD Eq. 25
= × ×
3 2
+ 2
dx ×d o
× do

where θ is empirical pressure drop correction factor and α is flow direction (1 for counter
current, and -1 for co-current). ε is packing density of hollow fiber bundle, μ is fluid dynamic
viscosity, r is fluid density, and u is fluid velocity. D denotes draw solution side, which
typically flows on the shell side of a hollow fiber module.

Meanwhile, the Hagen-Poiseuille model for pressure drop across cylindrical tube is used for
the lumen side pressure drop:

dP F 32 ×μ F × v F Eq. 26
=
dx di2

where P is pressure, F denotes the feed solution side, which typically flows on the lumen
side of a hollow fiber bundle, μ is fluid dynamic viscosity, u is fluid velocity and di is the
fiber inner diameter.

90
 Chapter 4

4.4.3 FO Design Considerations

For the batch or feed-and-bleed type of system, feed solution is re-circulated and water
extraction happens over time. Sensors may be installed to automate the feed-and-bleed or
cycle shutdown operation. In such a system, because the feed side is being concentrated,
water flux will start high and decrease over the course of a cycle, assuming that draw inlet
flow rate and concentration are constant. Care should be taken to ensure that initial flux is
below design flux limit for the given process, and final flux is non-zero so that cycle time is
still productive and reverse salt flux into feed batch is minimized.

For a single pass process where the FO feed outlet is expected to reach a desired concentration
factor, the number of modules and their array should be designed to achieve recovery
outcome, while balancing all design constraints above.

For instance, FO modules may be arranged in parallel to sub-divide flow to be within

recommended flow rate. FO modules, and hence membrane area, may be added in series
to achieve recovery in single pass, while design flux limit is obeyed. For the same flow rate
extraction requirement, added area means lower operating flux, ensuring that it is within
design limit. The maximum number of modules in series is dictated by pressure drop across
the system, while the maximum number of lines of modules in parallel is dictated by the
minimum FO outlet flow rate for each line.

One way to circumvent minimum FO outlet flow rate being below limit is to implement
multi-stage design. That means, the flow rate of multiple lines of FO modules of the so-
called first stage are combined and redistributed over a smaller number of lines in the second
stage. This allows more flow per module when recovery is at the highest point and by design,
above module limit by manufacturers’ recommendation.

Lastly, it should be noted that process limits should be considered for both flushing or
cleaning process as well as FO process. In the former, cross flow velocity on the feed side is
highest, and on the later cross flow velocity on the draw side is highest. It should be ensured
that design considerations are met for both operation types for successful commissioning
of a FO system.

Other design considerations

Beside technical considerations, there are other parameters that FO process designers should
pay attention to as it influences the operating cost of such a system. Most directly, increasing
the number of FO modules used will increase the cost of membrane replacement and initial
capital investment on the system. This will also increase the hold-up volume and volume of
flushing water or chemicals required for the cleaning process, even though this tends to take
a small fraction of overall operating cost.

If operating flux is still within design flux limit, increasing draw solute concentration results
in less membrane required and reduces membrane initial investment. However, this would
result in increased reverse salt flux from draw to feed side, increased salt passage into draw
regeneration permeate stream and increased energy cost of draw regeneration step.

91
Experimental Methods for Membrane Applications

4.5 APPLICATION EXAMPLES

Textile industry application of FO for lowering water footprint

Textile production is estimated to be responsible for about 20% of global clean water
pollution from dyeing and finishing products (Morlet et al., 2017 ). Given the increasing
need for the textile industry to lower the environmental impact it is necessary not only
to design appropriate wastewater treatment technologies but also to enable reuse and
recycling of water. Here FO can be used for water reclamation using concentrated dyeing
salt solutions as draw solutions where the diluted dyeing salt solutions can be used in the
dyeing baths directly (see Figure 5).

Following the study of Sheldon et al., (2019) it was evaluated the potential of dye solutions
as a novel draw solution by screening, assessing and identifying suitable reactive dyes,
e.g., Reactive Black 5 and Basic Blue 41 GRL dyeing solutions were investigated as draw
solutions in FO with a dye-to-salt 1:10 mass ratio, see Figure 1. Synthetic seawater (SSW)
and two types of textile wastewater (TWW1 and TWW2) were evaluated as feed solutions
for water reclamation. Reactive Black 5 and Basic Blue 41 GRL were diluted 10 and 5 times
respectively.

With Reactive Black 5 as draw solution and SSW as feed solution a water recovery of 75%
was achieved. Using TWW1 and TWW2 as draw solutions, water recovery was around
30%. Using Basic Blue 41 GRL with SSW, TWW1, and TWW2 as feed solutions, water
recoveries of 50%, 20% and 20%, respectively, were achieved. The average reverse solute
fluxes were between 0.06 and 0.34 g/m2h. Results indicated the potential of FO in the
textile industry leading to substantial water savings.

Concentrated
salt solution
Concentrated batching
dyeing rolls
wastewater fabric

FO unit RO unit
Textile
dyeing
Water process

Diluted Dyeing
salt solution wastewater

Figure 5 Implementing forward osmosis (FO) into the textile wastewater treatment process can
provide high value to an industry segment which is a large consumer of fresh water and
one of the biggest polluters. The scheme shows the FO process integrated in a textile
wastewater treatment plant using inorganic salt as a draw solution. For using the salt
solution as a draw solution there is an integrated reverse osmosis unit for the reuse of the
diluted salt.

92
 Chapter 4

Concentrating distillery wastewater for subsequent antioxidant retrieval

Alcohol distillation from sugarcane molasses constitutes an important industry in several
countries. Molasses-based distillation is a water intensive method with a freshwater
consumption in the range of 9-21 L per alcohol and concomitant wastewater production
of 7-15 L per L alcohol (Gol, 2014). The resulting wastewater has a high organic load, low
pH, and high total dissolved solids. About 2% (w/v) of the wastewater is melanoidins,
a product of Maillard reaction obtained from reducing sugars and amino acids during
distillation. From a classical wastewater treatment point of view this makes this particular
stream problematic as melanoids are not readily biodegradable. However, melaniodins have
antioxidant properties which could be a valuable sub-product. The high organic load and
the high total dissolved solids makes separation based on classical filtration challenging but
due to the inherently low fouling potential FO has attracted attention as a method for up-
concentration of this potential antioxidant source.

Singh et al. (2018) studied the concentration of distillery wastewater by FO with magnesium
chloride hexahydrate (MgCl2.6H2O) as draw solution. They used a 10% v/v melanoidins
model feed solution to optimize the operational parameters. Subsequently they achieved
85-90% melanoidins rejections with as-received distillery wastewater and 3M MgCl2.6H2O
as draw solution. The water flux was 2.8 L m-2h-1 with water recovery over 24 h was around
70% which is significantly higher than reported for RO (35-45%). However, further
investigations on membrane fouling and draw solution recovery are required to establish
the superiority of FO over RO for the concentration of this type of wastewater.

Concentrating electroplating wastewater

Chromium plating and chromate processes are widespread technologies for electroplating
of pristine or nickel-coated plastics as chromium and chromate endow surfaces with special
properties such as hardness and corrosion resistance (Korzenowski et al, 2018; Sorme et
al., 2002). In this process, large quantities of wastewaters, residues, and sludge is generated
which can be categorized as problematic waste requiring extensive waste treatment (Sorme
et al. 2002).

In the study of Bratovcic et al. (2022) FO was investigated for concentration of hexavalent
chromium (Cr(VI)) in electroplating wastewater from processing plastics to enable the
reuse of recovered Cr(VI) in the plating baths, see Figure 6. The feed solution was chromium
galvanic wastewater, while the draw solution was an underground brine (close to the factory
location) with osmotic pressures of 28 and 226.8 bar, respectively.

Baseline and FO filtrations were performed using Aquaporin Inside(R) membrane hollow
fibre FO (AIM™ HFFO) modules with a sequence of baseline, filtration (1.5h) and cleaning
(30 min with DI water) steps. During the initial filtration (F1), the water flux decreased on
average from an initial value of 28.7 LMH at 46.7 % water recovery to 18.5 LMH. For the
second filtration (F2) the water flux decreased from 20.1 LMH at 28.4 % water recovery to
16.8 LMH. The corresponding feed solution (wastewater) volume reduction factors were
1.9 and 1.4 with a concomitant Cr(VI) concentration factor of 1.6 and 1.3 for F1 and F2,
respectively. After 1.5 h of filtration, the Cr(VI) rejection was 99.7 % and 95.8 % for F1 and
F2, respectively. As the AIM™ HFFO membrane is negatively charged electrostatic repulsion

93
Experimental Methods for Membrane Applications

between the membrane surface and the negative ions (HCrO4− and Cr2O72-) will contribute
to the rejection of Cr(VI). The appearance of Cr(VI) in the draw solution indicated a loss of
membrane integrity which was ascribed to chemical degradation of the membrane due to
oxidation from Cr(VI). Local guidelines for standard chromium discharge from industrial
wastewater into the environment is 0.5–1 mg L-1. Since the diluted brine draw solution
contained 0.07 gL-1 and 0.65 gL-1 of Cr(VI) for F1 and F2 respectively, it cannot be directly
discharged into the salt groundwater resource.

In conclusion, brine-driven FO could concentrate chromium galvanic wastewater taking

advantage of the high chemical potential gradient provided by the high salinity brine, but
the membrane material must be adapted to withstand harsh environments.

102

Cr(VI) Plating Factory 100

Cr(VI)
Cr(VI) 98
Cr(VI) 96
Cr(VI)
Plating bath 94

92
1h 1.5h 1h 1.5h
90
24 F01 F02
Rejection of Cr(VI)
Cr
Chromium
51.996 30 2.2
25 2.0
Iw (LMH)

20 1.8

CF
15 1.6
10 1.4
5 1.2

Chrome Underground 0 1
wastewater brine 0 10 20 30 40 50
R (%)
Hollow fibre FO
Water flux and concentration
membrane module
versus recovery

Water passage through

the membrane

Figure 6 FO tests using Aquaporin Inside membrane hollow fibre FO (AIM™ HFFO) modules
for concentration of hexavalent chromium (Cr(VI)) in electroplating wastewater from
processing plastics to enable the reuse of recovered Cr(VI) in the plating baths. Chromium
galvanic wastewater was used as feed solution while the draw solution was underground
brine close to the factory location. The results show that FO can be used in this type of
application, but the membrane material must be adapted to withstand harsh environments
(Bratovcic et al, 2022)

94
 Chapter 4

4.6 OUTLOOK

FO is a relatively new technology which presents numerous advantages, especially when

a direct FO system can be implemented (i.e., draw solution is available and regeneration is
not needed) or when the resulting feed concentrate can bring an added value to the final
product. However, FO presents the drawbacks of a developing technology. These are
mainly the scarce availability of FO membrane manufacturers, the development of materials
which ensure a high water flux, high compound rejection, withstand harsh environments,
high selectivity to water and a reduced concentration polarization. The unique system
design characteristics required by the FO technology (i.e., draw solution regeneration
and membrane configurations) also involve an additional level of system complexity.
The availability of non-expensive draw solutions with the desired characteristics and the
suitability of these in those applications that require high safety levels, such as in food and
pharma industries, are also challenging. However, overall, FO technology can still bring
unique advantages in niche applications, although more research in membrane materials
and processing are needed to fully understand its capabilities in industry.

95
Experimental Methods for Membrane Applications

4.7 REFERENCES

Baker, R. W. (2004). Membrane Technology and Applications. John Wiley & Sons, Ltd.
DOI:10.1002/0470020393
Bratovcic, A., Buksek, H., Helix-Nielsen, C., Petrinic, I., Concentrating hexavalent chromium
electroplating wastewater for recovery and reuse by forward osmosis using underground brine
as draw solution Chemical Engineering Journal, 431, #133918, 2022.
Gol, 2014. Report by Principal Scientific Advisor to Government of India. Electronic source. http://
psa.gov.in/publications-reports/opportunities-green-chemistryinitiatives-molasses-based-
distilleries-2014
Im Sung-Ju, Lee H., Jang A. (2021). Effects of co-existence of organic matter and microplastics on the
rejection of PFCs by forward osmosis membrane. Environmental Research, 194, 110597
Jørgensen, M. K., Keiding K., Christensen, M.L. (2014). On the reversibility of cake buildup and
compression in a membrane bioreactor. Journal of Membrane Science, 455, 152-161
Korzenowski C., Rodrigues M.A.S., Bresciani L., Bernardes A.M., Ferreira J.Z., Purification of spent
chromium bath by membrane electrolysis, J. Hazard. Mater. 152 (3) (2008) 960–967.
Morlet, A., Opsomer, R., Herrmann, S., Balmond, L., Gillet, C., and Fuchs, L. (2017). A new textiles
economy: Redesigning fashion’s future. Ellen MacArthur Foundation.
Mulder, M. (1996). Basic Principles of Membrane Technology. Springer. DOI: 10.1007/978-94-009-
1766-8
Peters, Christian D. and Hankins, Nicholas P. (2020). The synergy between osmotically assisted reverse
osmosis (OARO) and the use of thermo-responsive draw solutions for energy efficient, zero-
liquid discharge desalination, Desalination, Volume 493, 114630
Sanahuja-Embuena, V.; Khensir, G.; Yusuf, M.; Andersen, M.F.; Nguyen, X.T.; Trzaskus, K.; Pinelo, M.;
Helix-Nielsen, C. Role of Operating Conditions in a Pilot Scale Investigation of Hollow Fiber
Forward Osmosis Membrane Modules. Membranes 2019, 9, 66. https://doi.org/10.3390/
membranes9060066
Sheldon et al. Water Sci Technol (2019) 80 (6): 1053–1062
Singh et al. Water Research 130 (2018) 271-280
Sorme L. and Lagerkvist R., Sources of heavy metals in urban wastewater in Stockholm, Sci. Total
Environ. 298 (1-3) (2002) 131–145.
Suwaileh W., Pathak, N., Shon H., Hilal N., Forward osmosis membranes and processes: A
comprehensive review of research trends and future outlook, Desalination, Volume 485, 2020,
114455, ISSN 0011-9164, https://doi.org/10.1016/j.desal.2020.114455.
Wang J, Liu X (2021) Forward osmosis technology for water treatment: Recent advances and future
perspectives. Journal of Cleaner Production 280: 124354 DOI https://doi.org/10.1016/j.
jclepro.2020.124354
Xiao, T., Nghiem, L. D., Song, J., Bao, R., Li, X. & He, T. (2017). Phenol rejection by cellulose triacetate
and thin film composite forward osmosis membranes. Separation and Purification Technology,
186 45-54.

96
 doi: 10.2166/9781789062977_0097

Chapter 5

Membrane Distillation
Mohammad Mahdi A. Shirazi, Aalborg University, Denmark

Cejna Anna Quist-Jensen, Aalborg University, Denmark

Aamer Ali, Aalborg University, Denmark

The learning objectives of this chapter are the following:

• Introduction to membrane distillation

• An overview of the membrane materials and setup used in MD

• Main techniques to characterize the membranes for MD

• Present and discuss the main applications of MD

• Provide an overview of the outlook of the process.

5.1 INTRODUCTION

Membrane distillation (MD) is a non-isothermal membrane process (Lawson and Lloyd,

1997a). In MD, a hydrophobic membrane with porous structure separates the feed and
permeate channels. The feed channel contains a heated solution with an elevated temperature,
while the permeate channel contains a cooling solution with lower temperature (Curcio and
Drioli, 2005). This temperature difference can provide a vapor pressure difference between
the feed and permeate channels. As the membrane is hydrophobic, the liquid feed does not
penetrate the pores (El-Bourawi et al., 2006). However, due to the vapor pressure difference
between the hot side and the cold side (i.e., the driving force), the vapor molecules can
transfer across the membrane pores (Wang & Chung, 2015). Figure 1 illustrates a general
scheme of the MD process.

© 2024 The Authors. This is an Open Access book chapter distributed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0),
(https://creativecommons.org/licenses/by-nc-nd/4.0/). The chapter is from the book Experimental
Methods for Membrane Applications in Desalination and Water Treatment, Sergio G. Salinas-Rodriguez,
Loreen O. Villacorte (Eds).
Experimental Methods for Membrane Applications

Tf

Tfm

Tpm

Tp

Water vapors
Ions
Other solids

Feed solution

Figure 1 A general scheme of the MD process (Tf: bulk temperature in the feed channel, Tfm:
temperature on the membrane surface in the feed channel, TPm: temperature on the
membrane surface in the permeate channel, and Tp: bulk temperature in the permeate
channel).

MD has numerous attractive advantages for water treatment applications. As only the vapor
molecules can pass through the membrane pores a complete solute rejection (i.e., ~100%,
theoretically) can be achieved by the MD process. MD uses membranes with pores in the
range of 0.1-0.5 μm, which is much larger than the pore size range in the pressure-driven
membranes, such as reverse osmosis (RO) and nanofiltration (NF) membranes. Moreover,
the operating pressure is much lower in MD compared with RO/NF. These can make
MD more cost-effective, less demanding on the membrane properties, and less sensitive
to fouling/scaling, as well (Shirazi et al., 2016; Tibi et al., 2020). Furthermore, the low
operating pressure allows MD to utilize less expensive materials with better anti-corrosive
properties, such as plastics, for module design and fabrication (Hussain et al., 2022). In MD,
the process liquid in the feed stream is not necessarily heated up to reach the boiling point
and the operating temperature ranges between 40-80 oC. This can reduce the required
energy for MD in comparison with the conventional distillation processes. Moreover,
working with low temperatures enables MD to utilize low-grade and renewable energy
sources, such as solar energy (Ahmed et al., 2020; Drioli et al., 2015b). All these can make
MD an attractive alternative for the desalination of seawater and brackish water, however,
in lower production rates in comparison with RO. Thus, MD may not still be competitive
with RO for freshwater production via seawater desalination for large scales but perform
effectively for small-scale desalination. Moreover, MD could find a way as an efficient
technology for treating challenging wastewater streams, such as RO brine, textile dyeing
wastewater, radioactive contaminated wastewater, etc., where other separation processes
cannot perform efficiently (Shirazi & Dumée, 2022). Table 1 highlights the advantages and
challenges associated with MD for various applications.

98
 Chapter 5

Table 1 Advantages and challenges associated with MD

Advantages Challenges
Low operating temperature (40-80 ˚C) Lack of specific membranes
High non-volatile solute rejection (~100%) High risk of pore wetting
Very low operating pressure (<0.2 bar) Heat loss
Less sensitivity to the feed concentration Temperature polarization
Modular and scalable Complexity in terms of process optimization
Integrate-able with other separation techniques Lack of experience in large scale

Competitive production cost with other processes

Possible to utilize renewable and low-grade energy
source
Less carbon footprint
Able to treat challenging wastewater streams

The very first MD experiments were conducted using commercially available microfiltration
(MF) membranes, which were made of hydrophobic polymers, such as polytetrafluorethylene
(PTFE), polypropylene (PP), and polyvinyl difluoride (PVDF) (Curcio & Drioli, 2005;
Shirazi et al., 2013a; Shirazi et al., 2012). However, they suffer from some drawbacks, such
as pore wetting, delamination of the active layer, etc., as they were not specifically fabricated
for MD (Wang & Chung, 2015). An extensive review of commercial membranes for MD
can be found in the literature (Khayet, 2011). The next generation of MD membranes
have been fabricated using commercial and synthesized polymers, such as PVDF. The MD
membranes can be fabricated using various techniques. The most investigated one is the
phase inversion technique (Eykens et al., 2017). However, the membranes by this technique
may suffer from some drawbacks, such as limited porosity, and dead-ended pores with
twisted structure (i.e., high tortuosity) (Qasim et al., 2021b). Thus, new research directions
have been focused on new fabrication techniques, such as electrospinning and 3D printing
(Tijing et al., 2014; Tijing et al., 2020). Moreover, extensive research was also carried out on
MD modelling and optimization (Ali et al., 2015a; Ali et al., 2016a; Ali et al., 2018; Hitsov
et al., 2015; Jantaporn et al., 2017; Olatunji & Camacho, 2018).

The new generation of membrane materials and fabrication techniques, and process
development and optimization have enabled MD to extend their applications in new
directions for more efficient and greener desalination towards zero-liquid discharge, treating
challenging wastewater streams (e.g., wastewater from biological and pharmaceutical
processes, metal finishing, and electronic industry), removal of specific gas streams (e.g.,
H2S) from process water, concentrating fruit juices in the food industry; and recovery of
minerals and value-added chemicals from wastewater streams (Tibi et al., 2020; Hussain et
al., 2022; Sanaeepur et al., 2022; Julian et al., 2022; Gontarek-Castro et al., 2022; Afsari et
al., 2021). Despite of all developments in this separation technique, MD still needs further
development for specific membranes with durable properties, high pore wetting resistance,
and long-term performance, new module design for higher energy efficiency and lower
polarization effect as well as experiencing new and challenging feed samples for treatment.

99
Experimental Methods for Membrane Applications

5.2 MATERIALS, EXPERIMENTAL SET-UP

5.2.1 MD membranes

5.2.1.1 Membrane properties

As MD is a non-isothermal membrane separation and only the water vapor molecules must
pass through the pores, the applied membrane should possess some essential requirements
(Alklaibi and Lior, 2005).

The most important characteristic of an MD membrane is surface hydrophobicity. The

membrane structure can possess a single layer or even multi-layers. However, the layer
which is in direct contact with the feed stream must be hydrophobic to repel the liquid and
only pass the vapor molecules (Shirazi et al., 2014).

To ensure a sustainable performance without pore wetting, the MD membrane should have
high liquid entry pressure (LEP). LEP defines as the minimum required pressure that allows
the feed liquid to enter the pores (He et al., 2011). To ensure a proper LEP value, further to
high surface hydrophobicity, the membrane pore size should be as small as possible. The
typical pore size for MD membranes has been reported in the range of 0.1-0.5 μm (Khayet,
2011). Moreover, the pore size distribution and the maximum pore size should be as narrow
as possible and as small as possible, respectively, to provide a high LEP value (McGaughey
et al., 2020a).

Porosity is another important parameter for MD membranes. It defines as the void volume
fraction which is openly available for transferring the vapor molecules (Gryta, 2007).
The membrane porosity is proportional to its permeability. Thus, the more porous the
membrane, the higher the permeate flux, regardless of the MD configuration. It is worth
noting that the fabrication technique can considerably affect the membrane porosity
(Ravi et al., 2020). For example, nanofiber membranes possess higher porosity (≥85%) in
comparison with phase-inverted membranes (40-80%) (Tijing et al., 2014).

The pore structure of MD membrane is often assumed to be a straight cylinder for modelling
purposes, while it is not true in real life. The deviation of the pore structure from the standard
cylindrical shape is defined by the tortuosity factor. Unlike porosity, tortuosity is inversely
proportional to the permeability of the MD membrane. Therefore, the lower the tortuosity
factor, the higher the permeate flux (Tai et al., 2019; Kim, 2021).

In MD, both mass and heat transfer through the membrane happen simultaneously
(Qtaishat et al., 2008). Although a high mass transfer (higher permeate flux) is favorable
for MD, high heat transfer through the membrane is considered as heat loss (Phattaranawik
et al., 2003a). With reference to the membrane thickness, a thin membrane can have
lower mass transfer resistance (higher permeate flux) and higher heat loss, while a thicker
membrane can have lower heat loss and higher mass transfer resistance (lower permeate
flux). Thus, the membrane thickness should possess an optimum value to compromise the
heat and mass transfer in MD. To achieve this, membranes with multi-layer structures have
been introduced for MD applications, where a thin hydrophobic membrane provides lower

100
 Chapter 5

mass transfer resistance and higher permeate flux, while the thicker and less hydrophobic
or hydrophilic support layer can reduce the heat loss (Bonyadi & Chung, 2007; Cheng et al.,
2018; Shirazi et al., 2020a; Zuo et al., 2017).

Table 2 Guideline for desired properties of MD membranes

Property Description Recommendation
Hydrophobicity The higher the hydrophobicity (i.e., the lower the As high as possible
surface energy), the higher the liquid retention.
Liquid entry pressure (LEP) The higher the LEP, the more pore wetting resistance. >250 kPa
Pore size Larger pore size can provide higher permeate flux; 0.1 – 0.45 µm
however, it reduces the LEP.
Porosity The higher the porosity; the higher the permeate flux, 60-80%
the lower the heat loss, and the lower the mechanical
strength.
Thickness The lower the thickness; the higher the permeate flux 30-450 µm
and heat loss.
Thermal conductivity The lower the thermal conductivity, the higher the As low as possible
permeate flux and energy efficiency.
Tortuosity The higher the tortuosity, the lower the permeate flux. As low as possible
Chemical resistance The higher the chemical resistance, the better the As high as possible
membrane integrity.
Mechanical strength The higher the mechanical strength, the better the As high as possible
membrane integrity.

As mentioned earlier, heat loss through the thermal conduction of the membrane is an
important challenge for MD (Phattaranawik et al., 2003b; Susanto, 2011a). Therefore, the
thermal conductivity of the membrane material should be as low as possible. Moreover,
the higher the porosity, the lower the heat loss through the membrane, as the heat transfer
coefficient of the trapped air in the membrane pores is much lower than that of the polymer
thermal conductivity (Eykens et al., 2016). Therefore, low thermally conductive polymers
can be used for fabricating a highly porous membrane to enhance the MD performance
(Shirazi et al., 2020a).

Further to all the above-mentioned characteristics, MD membranes should also possess anti-
fouling properties and should be chemically and thermally durable for stable performance
in the long-term operation (Chen et al., 2020). Table 2 provides a guideline for the desired
properties of MD membranes.

5.2.1.2 Membrane materials

The very first MD experiments were mostly focused on the process (e.g., proving the
concept, optimization of operating parameters, enhancing energy efficiency, etc.) (Lawson
and Lloyd, 1997b). These experiments were mostly carried out using commercial MF
membranes which were made of hydrophobic polymers (Franken et al., 1987a; Fane et al.,
1987; Schofield et al., 1987; Kimura et al., 1987; Schofield et al., 1990). However, these

101
Experimental Methods for Membrane Applications

membranes have not been specifically designed and fabricated for MD, which is a non-
isothermal separation based on the vapor-liquid interface equilibrium. This was found
as a serious challenge for optimizing and developing the application of MD. Therefore,
a considerable part of research in the next step has been focused on developing novel
membranes with enhanced performance in terms of anti-wetting properties, permeate flux,
solute rejection, and energy efficiency (Wang & Chung, 2015).

Table 3 Commercial polymers for the fabrication of MD membranes (Wang and Chung, 2015;
Qasim et al., 2021a)
Thermal
Chemical Surface energy conductivity Thermal Chemical
Polymer structure (×10-3 Nm-1) (Wm-1K-1) stability stability
PTFE F F 9-20 ~0.26 Very good Very good

C C

F F n

PVDF 30.3 ~0.18 Moderate Good

H F

C C

H F
n

PP 30 ~0.14 Moderate Good

CH3

PE H H 28-33 ~0.4 Poor Good

C C

H H n

MD membranes can be fabricated from both polymeric and inorganic materials. However,
the considered material must be hydrophobic (i.e., with low surface energy) intrinsically,
or by proper modification (Tibi et al., 2020). The common commercial polymers for the
fabrication of MD membranes include PTFE, PP, PVDF, and polyethylene (PE) (Qasim et
al., 2021b; Wang & Chung, 2015). Table 3 presents the characteristics of some common
hydrophobic polymers. As could be observed, PTFE has the lowest surface energy (9-
20*10-3 N.m-1), which can provide a relatively high surface hydrophobicity. Moreover, it is
chemically and mechanically stable. However, due to the insolubility of PTFE polymer in the
majority of chemical solvents, PTFE membranes should be fabricated using complicated and
expensive techniques, such as melt-extrusion (Feng et al., 2018). PVDF has higher surface
energy than that of PTFE, which means it is less hydrophobic. However, it has a lower
thermal conductivity in comparison with PTFE (Table 3). It is worth noting that PVDF and
their derivatives, such as poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP),

102
 Chapter 5

are the most investigated polymers for developing MD membranes (Qasim et al., 2021a).
This is due to the numerous advantages of these polymers, such as proper solubility in a
wide range of solvents and good processability, for membrane fabrication (Kang and Cao,
2014; Ji et al., 2015; Zou and Lee, 2022).

Recently, with developing new techniques for membrane fabrication, such as electrospinning
(which can provide nanofibers), new commercial polymers have also been investigated for
fabricating the MD membranes, such as polystyrene (PS), high-impact polystyrene (HIPS),
styrene-acrylonitrile (SAN), poly (methyl methacrylate) (PMM), acrylonitrile-butadiene-
styrene (ABS), styrene-butadiene-styrene (SBS), and polydimethylsiloxane (PDMS) (An et
al., 2017; Duong et al., 2018; Niknejad et al., 2021; Sadeghzadeh et al., 2020a; Shirazi et al.,
2020b).

Table 4 Comparison of polymeric and inorganic membranes for MD

Item Polymer membrane Inorganic membrane
Membrane
Material - Widely available - Available
Fabrication - Well-developed techniques - Developed techniques
- Inexpensive - Expensive
Mechanical strength - Flexible - Brittle
- Less durable in long term - Durable in long term
Fouling and cleaning - Sensible to fouling - Sensible to fouling
- Cleaning is challenging - Flexible in cleaning
Corrosion resistance - High - Moderate
Performance
perating temperature* - Low to moderate - Low to high
Permeate flux - Low (AGMD) to moderate (VMD) - Moderate (AGMD) to high (VMD)
Solute rejection - High - High

* Temperature range: Low (30˚C) to High (90˚C)

Further to polymers, inorganic materials can also be used for the fabrication of MD
membranes (Ramlow et al., 2019). The inorganic membranes can be made of a wide range of
materials, such as alumina, zirconia, titania, silicon nitride, and metal oxides of iron (Camacho
et al., 2013). However, these materials are hydrophilic in nature (e.g., due to the presence
of the hydroxyl group). Thus, inorganic membranes should be modified properly for their
surface to impart the required hydrophobicity for MD applications (Ferreira et al., 2021).
Inorganic membranes, in particular ceramic membranes, are more chemically, mechanically,
and thermally stable in comparison with polymeric membranes. Moreover, they can be
cleaned several times, even using extreme cleaning techniques, such as chemicals, steam, and
high-pressure backwash. However, inorganic membranes are expensive, and brittle, and show
higher fouling and scaling tendency in comparison with polymeric membranes (Omar et al.,
2022). Table 4 compares the polymeric and inorganic membranes for MD.

103
Experimental Methods for Membrane Applications

5.2.2 Experimental set-up

5.2.2.1 MD configurations
MD has four main configurations, as shown in Figure 2. All these configurations are identical
for the feed channel, where the hot solution stream is in direct contact with the hydrophobic
surface of the membrane (Adewole et al., 2022).

Condenser
Feed in Permeate out Feed in Sweeping
gas out

Permeate
Permeate

Permeate
Feed

Feed
(a) (b)

Feed out Permeate in Feed out Sweeping gas in

Feed in Feed in Cooling out

Permeate

Air gap
Feed

Feed
(c) (d)

Vacuum
pomp

Feed out Permeate Feed out Cooling in

Permeate

Figure 2 A general scheme of the main MD configurations: (a) direct contact membrane distillation
(DCMD), (b) sweeping gas membrane distillation (SGMD), (c) vacuum membrane
distillation (VMD), and (d) air-gap membrane distillation (AGMD).

When further to the feed side, the applied membrane is in direct contact with the liquid in
the permeate side, which is pure and cold, it is known as the direct contact MD (DCMD)
(Figure 2-a). This is the most investigated and simplest MD configuration (Khayet,
2011). DCMD has been used for different applications, such as seawater desalination and
wastewater treatment. However, the most important drawback of DCMD is the high heat
loss through the membrane thermal conduction (Ashoor et al., 2016; Niknejad et al., 2021).

104
 Chapter 5

Table 5 A comparative overview of MD configurations

Configuration Description Highlights
DCMD • Both feed and permeate channels • Simple modules design.
contains liquid streams which • Internal condensation.
are in direct contact with the • Possibility for internal heat recovery.
membrane surfaces. • Pure/fresh water is required for cooling stream.
• The most investigated MD configuration.
• Wide range of applications.
• High heat loss and low thermal efficiency.

SGMD • An inert gas stream with low • Lower permeate flux compared to VMD.
temperature is used for sweeping • External condenser is required.
the vapor molecules in the • The least investigated MD configurations.
permeate channel. • Promising for concentration application, where the
permeate can be vented.
• The operating condition in the permeate. channel is
more effective on flux compared to DCMD.
• Good when feed stream contains volatile
compounds.
• Low heat loss.
• Challenging in terms of heat recovery.

VMD • Vacuum is used to enhance the • High permeate flux.

permeation of vapor molecules in • Higher risk of pore wetting.
the permeate channel. • Low heat loss.
• Challenging in terms of heat recovery.
• External condenser is required.
• Good for processing solutions with low. vapor
pressure at a desire temperature.
• Feasible for large scale.

AGMD • An air gap is maintained between • High energy efficiency.

the membrane and a condensing • Internal condensation.
surface in the permeate channel. • Possibility for internal heat recovery.
• Seawater and non-hazardous (pre-treated)
wastewater can be used as cooling stream.
• Lower permeate flux compared to other
configurations.
• Wide range of applications.
• Feasible for large scale.

The water vapor molecules in the permeate channel can be collected either by imposing
a vacuum pressure or by imposing a sweeping gas flow. Under these circumstances, the
configurations are known as the vacuum MD (VMD) and the sweeping gas MD (SGMD),
respectively (Figures 2-b and 2-c, respectively) (Abu-Zeid et al., 2015; Said et al., 2020).
Although SGMD and VMD can provide relatively higher permeate flux in comparison with
DCMD along with moderately better energy efficiency, an external condenser is required
for both configurations, which can increase the operation cost and energy consumption
(Khayet et al., 2003; Huayan et al., 2011). To overcome this challenge, an air gap can be
imposed between the membrane and a condensing surface in the permeate channel (Figure
2-d). This MD configuration is known as the air gap MD (AGMD) (Khayet and Cojocaru,

105
Experimental Methods for Membrane Applications

2012). This module design for MD can considerably enhance energy efficiency without
needing an external condenser (Kalla et al., 2019). However, the stagnant air gap can impose
extra mass transfer resistance for transferring the water vapor molecules from the interface
of the membrane support in the permeate channel towards the condensing surface (Shahu
and Thombre, 2019a). Table 5 presents the comparison among the main configurations of
the MD process.

5.2.3 Process

5.2.3.1 MD system
Figure 3 illustrates a general scheme of the MD system. The system consists of various parts
which are introduced briefly.

Thermocouple
T T

Cooler

pH T Heater
Peristaltic meter EC meter
pump pH
meter EC meter

Permeate Peristaltic
pump

Balance
Computer Computer
Feed tank

Figure 3 A general scheme of the MD experimental system.

The MD system consists of at least two containers, one to store the feed solution (feed tank)
and another one to collect the product (permeate tank). The permeate tank should be placed
on a balance to record the variation in the permeate mass and calculate the permeate flux.

The mass transfer and permeate production take place inside the membrane module, which
can have different configurations, including plate and frame, tubular, hollow fiber, and spiral
wound (Francis et al., 2022). A pump is required to recirculate the feed stream in the close
loop of the feed tank-module-feed tank. Depending on the MD configuration, the cooling
flow in the permeate channel can be provided by another pump in DCMD and AGMD or
by a compressor or blower in SGMD. In the case of VMD, the permeate channel is under
vacuum pressure using a vacuum pump.

A heating system is required for heating up the feed stream to a desired temperature.
Likewise, a cooling system is required to keep the temperature of the cooling stream low
and constant. In the case of the SGMD and VMD configurations, an external condenser is

106
 Chapter 5

also required to condense and collect the permeate. To monitor and control temperature, at
least four thermal sensors should be placed as close as possible to the inlet and outlet points
of the MD module. Thermal sensors are shown by ‘T’ in Figure 3.

The quality of the product in the permeate tank can be monitored using an EC meter
and a pH meter. Likewise, the variation of the feed quality can be monitored in the feed
tank. Depending on the requirements of the research, other equipment, such as an in-line
microscope or an injection, can also be considered.

5.2.3.2 Operating parameters

MD is a non-isothermal separation process, and the feed temperature is a dominant operating
parameter in all MD configurations (Curcio and Drioli, 2005). The operating temperature
in the feed channel can vary from 40 to 80 ˚C, depending on the available energy source,
the thermal resistance of the membrane, and the system design (Ahmed et al., 2020).
Moreover, the temperature of the cooling stream in the permeate channel is also important
for DCMD, SGMD, and AGMD configurations. For VMD, however, the vacuum pressure
in the permeate channel is an important operating parameter further to feed temperature
(Mohammadi et al., 2015; Peydayesh et al., 2015).

Fluid flow rate (i.e., the recirculation rate) in both the feed and permeate channels is another
operating parameter. Different values were reported in the literature for the flow rate,
depending on the MD module and capacity of the system. One should be considered the
inlet pressure at high flow rate. When the fluid flow is provided by a peristaltic pump, the
pressure of the fluid flow in the feed and permeate channels is quite low, i.e., very close to the
atmospheric pressure. However, if other types of pumps, such as centrifugal or diaphragm
pumps, are used, a proper pressure reducer device/tool (e.g., the pressure regulator) should
be used before the MD module.

The feed solution in MD can contain various compounds, such as chemicals, alcohols in
dilute concentration, different salts (e.g., NaCl, MgCl, CaCO3, Na2CO3, Na2SO4, etc.),
sugars (lactose, sucrose, glucose, etc.), etc. (Shirazi & Kargari, 2019; Ali et al., 2021; Ali et
al., 2015, 2018; Park et al., 2020; Peydayesh et al., 2015; Quist-Jensen et al., 2016a; Quist-
Jensen et al., 2017; Quist-Jensen et al., 2016; Shirazi et al., 2014). Thus, the feed type and its
concentration are other operating parameters, which can affect the MD performance, then
they can also be considered.

Depending on the MD configuration, the vacuum pressure, sweeping gas flowrate, and the
distance of the air gap in VMD, SGMD, and AGMD configurations, respectively, can also
be investigated as other operating parameters. Moreover, membrane properties (i.e., type,
material, pore size, etc.) have also been investigated in various research (Abu-Zeid et al.,
2015; Chamani et al., 2021; Curcio & Drioli, 2005; Eykens et al., 2017; Hitsov et al., 2015;
Shahu & Thombre, 2019a; Shalaby et al., 2022a, 2022b).

5.2.4 MD modules
A typical membrane module is a unit consisting of a membrane mounted in a housing and
containing feed inlet, retentate outlet and permeate outlet channels (Yang et al., 2013). As
new membrane applications emerge and new module designs are developed, the definition

107
Experimental Methods for Membrane Applications

of modules evolves as well. For example, in submerged membrane modules, as the name
suggests, the module is directly immersed in the feed solution without an outer casing and
has only ports for permeate removal. Other examples are air-gap and permeate-gap MD
modules (AGMD and PGMD, respectively), which require additional channels at the inlet
and outlet of the cooling flow.

The main purpose of this module is to properly secure the membrane so that it can be used
in its designated application. However, a well-designed module must also meet several
other requirements. A proper module design should ensure a high packing density of the
membrane. Packing density is taken as the size of the surface area of a functional membrane
in a given volume. Generally, high packing density is desirable to avoid inefficient use of
module housing. However, it should be noted that for hollow fiber membranes, increasing
the packing density beyond a critical value can result in stagnant or ‘dead’ regions of poor heat
and mass transfer within the module. important. For plate and frame flat sheet membrane
modules, typical packing densities range from 100 to 400 m2/m3, whereas hollow fiber
membrane modules can have higher packing densities of up to 3000 m2/m3 (Peng et al.,
2012).

MD systems involve mass transport steps through the feed, membrane, and permeate, with
each region having a specific transfer coefficient. To mitigate mass transfer resistance at the
boundary layer, appropriate module designs should demonstrate good hydrodynamics,
minimize temperature and concentration polarizations, and minimize energy consumption.
Modules should maximize heat recovery, be easy and economical to fabricate, and minimize
leakage issues. They should also facilitate scale-up and integration into existing processes.
The module’s performance should be predictable using mathematical models under various
operating conditions and feed characteristics. The module must maintain integrity during
long-term operation, minimize foulant deposition, and be resistant to heat and chemical
degradation.

New variants of flat sheet and hollow fiber membrane modules have been introduced for
MD applications. Flat sheet membranes are typically assembled in plate and frame or spiral
wound configurations, while hollow fiber modules can be classified into shell and fiber/tube
and submerged configurations. These new variants aim to improve process performance by
improving heat and mass transport, heat recovery, and membrane area.

The simplest module design for MD experiments is the plate and frame module. In this
design, the membrane is placed between two frames and plates. The membranes for this
type of module are in flat sheet form. This module can possess different sizes and is useful
for lab-scale experiments. However, the membrane area is small, and this module does not
have that much chance for industrial applications. Various flow arrangements can also be
considered for this module, including the co-current, counter-current, and crossflow. The
efficiency of these flow arrangements in terms of the permeate flux in the plate and frame
modules is in the order of: crossflow>counter-current>co-current (Shirazi et al., 2014).

108
 Chapter 5

(A)
Cold stream

Hot stream

(B)
Cold stream

Hot stream

(C)

Hot stream
Cold stream

Figure 4 The flow arrangement of hot and cold streams in the plate and frame module; (a) co-current,
(b) counter-current, and (c) cross-current flows (Shirazi et al., 2014).

Vacuum multi-effect MD (VMEMD) systems operate under reduced pressure and can
achieve higher water recovery rates compared to AGMD. These systems use multiple stages
of MD in series, working at lower operating temperatures and pressures. They are compact,
high-efficiency systems with solar thermal collectors and solar-photovoltaic sources as heat
sources. VMEMD designs enable internal heating and condensation, saving heat energy
(Zhao et al., 2013).

The air gap width is a crucial parameter in AGMD module design, determining distillate
production rate. It aims to prevent condensing media from contacting the membrane surface,
reducing heat loss, and increasing vapor transport distance. The lower limit is determined
by thermal efficiency and water bridging. Various membrane configurations, including flat
sheet, tubular, hollow fiber, and spiral wound, have been applied in AGMD studies.

The fundamental module design of the AGMD has undergone numerous modifications,
including the introduction of spacers in the feed channel and the use of cooling plates on the
coolant channel. These modifications have improved the efficiency of heat removal from
the coolant and increased system flux, with the flat and channelled plates being effective
in increasing system flux. Multi-effect AGMD (ME-AGMD) is another novel approach to
improving module performance and industrialization of MD. Pangarkar and Deshmukh
developed a new ME-AGMD module for water treatment applications, which performed
better in terms of permeate flow and energy utilization (Pangarkar and Deshmukh, 2015).
The parallel stage MS-AGMD system generated 2.6 and 3 times more permeate volume
than a single-stage system, but its precise energy usage was only 1.5 times that of the
single-stage system. Operational modifications of the conventional AGMD module have
reduced pressure inside the gap below atmospheric pressure, increasing distillate flux up to
3 times. The generation of vacuum in the gap requires an additional vacuum pump and extra

109
Experimental Methods for Membrane Applications

electric energy input. The traditional AGMD module has been modified to an air-cooled
AGMD, which has the potential to significantly reduce energy consumption and costs for
desalination by minimizing or eliminating plant components associated with coolant flow
systems.

The spiral-wound module is a variation of the plate-and-frame configuration, where

a membrane envelope with a spacer is wound around a porous tube. It was first used
in pressure-driven processes like RO and UF and has since been used in research. The
Fraunhofer Institute for Solar Energy Systems has produced single and multi-channel spiral
wound membrane modules for MD. Fabrication involves rolling a membrane, condenser
foil, and spacer materials around a motorized main spindle, and sealing the channels with
resin. Multichannel spiral wound modules are built using multiple pairs of evaporator/
condenser flow channels. The main advantage of spiral wound modules is better counter-
current circulation of hot and cold streams, allowing for more efficient heat recovery
(Winter et al., 2012; Winter et al., 2011).

Traditional flat sheet membranes have been used for AGMD modules, but hollow fiber
has gained interest. Two strategies have been proposed for designing hollow fiber AGMD
modules: using the inside surface of the module to condense vapor passing through the
membrane or condensing the condensing surface inside the shell (Abu-Zeid & ElMasry,
2020; Alpatova et al., 2019; Shahu & Thombre, 2019b). Different variants of the second
approach have been proposed, including porous and dense hollow fibers packed inside
a module, water gap MD (WGMD), and multiple copper tubes enclosed in the shell of a
hollow fiber AGMD module. These approaches have shown promising results in improving
heat transfer efficiency and reducing cooling channel replacements.

The second conventional module design for MD is the hollow fiber module, which can have
hundreds of hollow fibers in a shell tube. Thus, the hollow fiber module can provide a much
larger surface area for MD experiments in comparison with the plate and frame modules. The
feed stream and the permeate stream can flow through either the fibers or the shell. In terms
of the high salinity brine, for example, it would be better to introduce the feed stream to
the shell instead of the fibers, as the membrane cleaning would be easier and more efficient
(Quist-Jensen et al., 2017). Same as the plate and frame modules, all MD configurations
can be performed using the hollow fiber modules. However, in the case of AGMD, special
design considerations may be necessary.

A hollow fiber module is a system consisting of a hollow fiber membrane bundle, cartridge,
tube sheets, and side caps. The bundle consists of hollow fiber membranes packed together,
with a liquid potting substance forming the tube sheet. The tube sheet acts as a fluid-tight
barrier, separating streams flowing through the lumen and shell sides of the module. The
packing density of the hollow fiber module is crucial for its productivity, as it directly affects
the module’s productivity(Mat et al., 2014). The packing density can be arranged in various
configurations, such as parallel, crisscross, or other precise geometric arrangements.

In DCMD, hollow fiber modules are typically in shell and tube heat exchanger configuration,
with feed flow on one side and permeate on the other. The feed compartment is based on
the feed solution properties. Axial flow can be divided into co-current and counter-current

110
 Chapter 5

flows, with the counter-current flow arrangement being the most widely used configuration
for MD applications. Cross flow is often used in axial flow designs to reduce stagnant regions
and concentration polarization effects.

Traditional parallel hollow fiber modules are susceptible to high temperature and
concentration polarizations, especially at low fluid flow rates. The non-uniformity of fiber
packing is a challenge due to the production of parallel fiber bundles. This results in sluggish
or dead zones, reduced separation efficiency, and channelling or bypassing in poorly packed
zones. To improve uniformity, fibers can be weaved into different structural geometries,
such as helical, wavy, or twisted shapes (Ali et al., 2015b; Shahu & Thombre, 2021).
This results in more uniform shell flow and less concentration polarization due to fluid
mixing. Studies have shown that using these geometries can increase flux enhancement
in membrane applications. Yang et al., (2012) compared five types of hollow fiber module
designs, revealing that the space-knitted fiber design showed the best performance, with
over 90% increase in permeate flux. This configuration improved fluid dynamics and even
flow distribution, increased vapor permeability, and reduced thermal polarization with
lower energy loss.

Submerged hollow fiber MD modules are increasingly popular in membrane bioreactors

(MBRs) due to their simplicity and ease of fabrication (Francis et al., 2015; Meng et al.,
2015; Gryta, 2020). These modules eliminate the need for feed stream circulation, reducing
electric energy consumption. They have been proposed for VMD and DCMD configurations,
allowing for recirculation of either feed or permeate stream while submerged in the other
stream. The MDBR system, combining MD with a thermophilic MBR, produces high-
quality permeate with a flux two orders higher than competitors. Submerged MD hollow
fiber modules have also been proposed for desalination of Red Sea water, using PTFE-based
hollow fibers immersed in clean water and hot feed introduced inside.

However, some inherent issues have been identified, such as high temperature and
concentration polarizations, fouling, and scaling at the membrane surface. Strategies such
as mixing feed solution with a magnetic stirrer, transmembrane vibrations, and low-power
ultrasound have been proposed to improve the efficiency of these modules.

Module design in membrane-based membranes (MD) has been influenced by overall

contact length, which refers to the membrane length over which hot feed streams stay in
contact without intermittent heating. High contact lengths can be achieved by increasing
membrane length, connecting multiple modules in series, or increasing membrane length
in each envelope of flat sheet membrane. Recovery of latent heat of condensation from
permeate can reduce the thermal energy consumption of MD, but in DCMD, a sufficient
length is required to keep the feed and permeate in contact. Recent studies have shown
that increasing flow channel length can reduce gain-to-output ratios, specific thermal
energy consumption, and channel depth (Abu-Zeid et al., 2021; Ali et al., 2016b). (Tsai
et al., 2023) proposed multipass hollow fiber membrane modules, which studied the
effect of operational modes, number of passes, length, and operating temperature on the
performance of the multipass modules. The results may open new windows for future MD
modules for better and more energy-efficient performance, particularly for desalination and
brine management purposes.

111
Experimental Methods for Membrane Applications

5.3 METHODS

5.3.1 Process measurements and calculations

5.3.1.1 Permeate flux

The permeate flux in MD can be measured as follow:

m
J= Eq. 1
A× t
where Δm, A, and t are the collected permeate mass (kg), membrane surface (m2), and the
interval time (sec), respectively (Drioli et al., 2013).

5.3.1.2 Solute rejection

The solute rejection in the permeate stream can be measured as follow:
Cp
Rejection(%) = (1 ) ×100 Eq. 2
Cf

where Cp and Cf are the solute concentrations in the permeate and feed sides (mg/L),
respectively. Cp can be calculated based on the following equation with reference to the
dilution effect:

C1m1 C0 m0
Cp = Eq. 3
m1 m0

where m0 and m1 are the initial and final masses of the cold stream. C0 and C1 are also the
initial and final salt concentrations of the cold stream, respectively (Lu et al., 2016).

5.3.1.3 Logarithmic temperature difference

The logarithmic temperature difference along the MD module is defined as follow (Quist-
Jensen et al., 2018):

(Tf i Tp0 ) (Tf 0 Tpi )

dTln = Eq. 4
(Tf i Tp0 )
ln
(Tf 0 Tpi )

where Tfi and Tfo are the inlet and outlet temperatures of the feed stream, and Tpi and Tpo
are the inlet and outlet temperatures of the permeate stream, in the membrane module,
respectively.

5.3.2 Membrane characterization

5.3.2.1 Physical and morphology properties

(a) Scanning electron microscopy
Membrane surface can affect the MD performance in terms of flux, fouling, etc. Thus, surface
morphology should be investigated, specifically for fabricated membranes (Talukder et al.,
2022; Wang et al., 2023). Scanning electron microscopy (SEM) is the most investigated

112
 Chapter 5

method for morphology observation of various materials, including membranes (Khulbe

and Matsuura, 2017). SEM has numerous advantages for determining the membrane
morphology, such as simplicity and ease of operation. Moreover, as SEM is a non-destructive
technique, the samples are not damaged during the analysis and can then be used many
times for further imaging (Naresh-Kumar et al., 2012). Figure 5 shows the example of SEM
images of a commercial PTFE and a fabricated membrane with nanofiber structure for MD.

Nanofiber membrane PTFE membrane

Figure 5 SEM images of a fabricated membrane with nanofiber structure (left) and a commercial
PTFE membrane (right) for MD (Shirazi et al., 2020a).

Before conducting an SEM test, sample preparation is required. To prepare the sample, a
small piece of membrane is cut, and placed on a stub. As the sample is in small size, tweezers
are usually used along with double-sided glue tape for fixing the sample on top of the
holder. To prevent charge accumulation, the sample should be sputter coated with a thin
layer of a highly conductive metal, such as gold or platinum (Sharma & Bhardwaj, 2019;
Vladár & Hodoroaba, 2020). Moreover, if cross-sectional images are required, the sample
should carefully be freeze-dried in liquid nitrogen, and then should properly be cut using a
razor blade (Zhu et al., 2012; Conners and Banerjee, 2020; Vladár and Hodoroaba, 2020). It
is worth noting that the sample should not be touched to avoid adding contamination and
footprint.

If more informative images are required, for example for morphology observation and
detection of nanoparticles on the membrane surface, field emission SEM (FESEM) with the
platinum coating for the sample are recommended, as it can provide images with higher
resolution (Lewczuk and Szyry ska, 2021; Kirk, 2017). The SEM or FESEM images can
be used for morphological observation, determination of pore size and its distribution,
thickness measurement (from the cross-sectional images), investigating the homogeneity,
and presence of particles or fouling layer on the membrane surface.

Also, SEM utilizes imaging software to measure the dimensions of, e.g., the size of particles,
on the surface at various magnification ranges. Moreover, various external software, such as
ImageJ which is an open-source software for image processing, can be used for measuring
the pore size, pore size distribution, and porosity (Guillen et al., 2010; Shirazi et al., 2013;
Ziel et al., 2008).

(b) Transmission electron microscopy

If clear view of the internal structure of the membrane sample is required, the transmission
electron microscopy (TEM) should be applied for imaging the membrane samples. TEM is
an alternative to SEM for investigating the structural morphology and crystalline (Sharma

113
Experimental Methods for Membrane Applications

et al., 2018; Shyam Kumar et al., 2017). For example, it can be used when nanoparticles are
incorporated into the membrane structure. Thus, the internal morphology and distribution
of nanoparticles can be investigated using TEM (Qin et al., 2015; Dadari et al., 2022; He et
al., 2023). Despite of SEM, which is more practical for surface observation, TEM can provide
accurate information about the structure and the body of the membrane sample (Mousa et
al., 2022; Talukder et al., 2022; Wang et al., 2023; Wiktor et al., 2017).

(c) Pore size and its distribution

Pore size is an important parameter for MD membranes, as only vapor molecules should
pass through the pores (Eykens et al., 2016). For example, although membranes with large
pore size are considered to provide higher permeate flux, they also suffer from high pore
wetting risk (McGaughey et al., 2020b). Membrane pore size can be measured using various
techniques (Nakao, 1994; Tanis-Kanbur et al., 2021; Tung et al., 2014).

As mentioned earlier, the pore size and its distribution can be measured using SEM images of
the membrane surface (Ahmed et al., 2015; Sadeghzadeh et al., 2020b; Shirazi et al., 2013a).
However, it should be noted that the obtained results are based on surface observation.
The membrane pore size can also be determined using the filtration test. To do this, fine
particles with a known particle size distribution can be filtered using the membrane sample.
The permeate sample should be tested for the particle content and their sizes. By comparing
particle size in the permeate with the original value of the particle size in the feed sample,
the pore size range can then be determined. The results of the particle filtration test can then
be compared with the obtained results based on the image processing of SEM images for
pore size measurement (Gopal et al., 2006; Sadeghzadeh et al., 2020).

More accurate data for pore size and pore size distribution of MD membranes can be provided
using the capillary flow porometry technique (Jena and Gupta, 2005a). In this technique, a
small piece of membrane sample should be placed in a holder and get wet using a proper
wetting liquid of known surface tension, such as Topor or Galwick (Jena and Gupta, 2010;
Kolb et al., 2018). Afterward, the different flows of inert gas should be used to displace the
liquid inside the pores on the membrane structure. Using this technique, pore size and the
pore size distribution can be obtained (AlMarzooqi et al., 2016; Jena and Gupta, 2005b; Li
et al., 2006).

(d) Wetting properties

(i) Water contact angle
In MD, the membrane pores must not get wet with the feed solution, and only vapor
molecules should be passed through the pores. Therefore, the wetting property of the
membrane surface is crucial for MD applications (Chamani et al., 2021). This can be
determined using the surface contact angle. According to the standard definition, the
membrane surface is hydrophobic if the water contact angle on the sample surface is greater
than 90o, while the contact angle lower than this represents the hydrophilicity of the
membrane surface (Ismail et al., 2022; Rezaei et al., 2018). Figure 6 illustrates this concept.
Figure 7 also shows a typical system for contact angle measurements.

114
 Chapter 5

θ > 90˚

θ < 90˚

Hydrophile Hydrophobe

Figure 6 A general scheme of the surface hydrophobicity.

Figure 7 A typical contact angle measurement system (www.kruss-scientific.com).

To measure the contact angle, a 5-μL liquid droplet (usually deionized water) is placed on
the membrane surface and a high-resolution speed camera takes the image of the shape of
the droplet. The sessile drop technique can then be used for calculating the water contact
angle and surface energy of the membrane sample (Franken et al., 1987b; Lu et al., 2019). To
have high accuracy in the reported results, it is recommended to conduct the contact angle
test at least for five different points on the membrane surface, and then report the average
value.

115
Experimental Methods for Membrane Applications

(ii) Liquid entry pressure

Liquid entry pressure (LEP) determines the minimum required pressure to penetrate the
feed solution to the pores. When the pressure of the feed stream is greater than LEP, the
liquid enters the pores and pore wetting happens (Eykens et al., 2016). Thus, the membranes
for MD should possess as high LEP value as possible (Sadeghzadeh et al., 2020b). LEP is
proportional to some parameters, including the surface hydrophobicity, surface tension of
the feed solution, pore structure, and the pore size. LEP can therefore be calculated as follow:
−2Bγ l cosθ
LEP = Eq. 5
rmax

where θ is the surface contact angle (can be measured using the water contact angle test), ϒl
is the surface tension of the feed solution, B stands for the pore geometry factor, and rrmax is
the maximum pore size (Silva et al., 2021; Rácz et al., 2014).

0.4 bar
6 5

3 4

1
Membrane
2

Figure 8 A general scheme of the typical setup for measuring the LEP value of MD membranes. The
system consists of (1) an inert gas cylinder, (2) a pressurized container, (3) a membrane
cell filled with water, (4) a flowmeter, (5) a digital manometer, and (6) a pressure
regulator (Essalhi and Khayet, 2013).

LEP can also be measured experimentally for the membrane samples. Figure 8 illustrates the
general scheme of an experimental LEP measurement set-up. To do this, a dry membrane
sample should be placed in a plate and frame module (for flat sheet membrane samples),
and distilled water should be exposed on the hydrophobic surface of the membrane. The
pressure of the module should then be increased stepwise (10 kPa per minute would be
recommended followed by a few seconds time laps) using an inert gas (e.g., nitrogen). As
soon as the first water droplet is observed on the other side of the membrane sample, the
corresponding pressure represents the LEP value (Khayet and Matsuura, 2001; Rácz et al.,
2014). It is also recommended to evaluate the LEP using the feed solution further to the
distilled water, as the membrane is in direct contact with the feed solution rather than the
distilled water in MD experiments (Silva et al., 2021).

116
 Chapter 5

(e) Porosity
The membrane porosity can be measured using the gravimetric method. In this technique,
a small piece of membrane should be cut and then the dry weight should be recorded. The
sample should then be immersed in a proper wetting liquid (e.g., isopropanol alcohol) to get
completely wet and re-weighted again (Khayet and Matsuura, 2001). The porosity (ε) of the
membrane sample can then be calculated as follow:

w1 w2
( )
= 1 m Eq. 6
w1 w2 w2
( )+( )
m l

where w1 and w2 are the weights of the dry and wet samples, respectively. Moreover, rm and
rl are the density of the membrane sample and the density of the wetting liquid, respectively
(Alkhudhiri et al., 2012). It is worth noting that the weight measurement of the wetted
membrane sample should be carried out carefully.

As it was mentioned earlier, the membrane porosity can also be measured using the image
processing of SEM images. However, it should be noted that this will be the surface porosity
(Sadeghzadeh et al., 2020).

(f) Thickness
Membrane thickness can directly be measured using a precise micro calliper. It is
recommended to measure the thickness at least for 10 points and then report the average
value, to be sure to minimize the compression effect (Zhang et al., 2017).

Further to this, the membrane thickness can also be measured using the optical microscope
along with a proper scale bar (Vicente et al., 2013). More accurate thickness data, however,
can be provided by SEM through the cross-sectional imaging (Attia et al., 2018a; Attia et al.,
2018b).

(g) Surface roughness and topography

Both the surface roughness and topography can affect the performance of MD membranes.
These parameters can be determined using the atomic force microscopy (Khayet et al.,
2004). To evaluate the membrane surface using atomic force microscopy (AFM), a sample
with defined dimensions should be cut, placed, and then fixed on the top of the holder using
the double-sided glue tape (Wyart et al., 2008).

AFM uses a nanometric prob to move along the membrane surface and collect the
topographical data using a laser diode and a detector. Imaging can be performed via three
different modes, i.e., contact mode, semi-contact mode, and non-contact mode. The
generated data should be analysed using a collector system, and then topography images
(with Angstrom resolution) can be made. It is worth quoting that the non-contact mode
can provide 3D topographic images with higher resolution (Johnson and Hilal, 2015;
Hilal et al., 2004). When performing the AFM analysis for a membrane sample, the type
of probe (e.g., silicon nitride), scanning environment (e.g., in air at ambient conditions),

117
Experimental Methods for Membrane Applications

the specifications of the cantilever and its tip (i.e., length, width, resonance frequency,
slope, etc.), the scanning speed (e.g., 5 μm/s at 1 Hz), the applied force (e.g., 0.15 nN), the
scan size, and the resolution (250 points per line) are important parameters (Shirazi et al.,
2013a). Figure 9 presents the 3D AFM images of three commercial MF membranes for MD
applications based on the non-contact mode imaging.

PP

Figure 9 3D AFM image of commercial MF membrane for MD applications (Shirazi et al., 2013b).

Table 6 Typical roughness parameters for evaluating the membrane surface topography (Shirazi
et al., 2013a)
Roughness parameter Expression
Average roughness (Ra)
1 n
Ra = ∑Z
n i−1 i

Root-mean-square roughness (Rq) n

1
Rq = Z i2
n i 1

Peak-to-valley height (Rz) Rz = Z max − Z min

Surface skewness (Rsk) n

1
Rsk = Z i3
nRq3 i 1

Surface kurtosis (Rku) n

1
Rku = Z i4
nRq4 i 1

Zi The height at point i

n Number of points in the image
Zmax and Zmin The highest and the lowest Z values

118
 Chapter 5

Table 6 introduces some AFM parameters which are useful for evaluating the membrane
topography. For example, the average roughness (Ra), which provides an overall view of the
surface roughness, is the most reported topography parameter for MD membranes. In other
words, the higher the Ra value, the rougher the membrane surface (Johnson & Hilal, 2015;
Shirazi et al., 2013b). The skewness factor (Rsk) represents the height distribution symmetry.
While the positive Rsk parameter shows the domination of peaks on the surface, the negative
Rsk values are associated with a porous surface. The kurtosis factor (Rku) corresponds to the
sharpness of the height distributions (Johnson and Hilal, 2015). Rku values greater than 3
represent a surface with sharper height distribution, while values lower than 3 indicate a flat
surface. Further detailed descriptions and applications of these parameters for characterizing
MD membranes can be found in the literature (Johnson & Hilal, 2015; Shirazi et al., 2013b;
Shirazi et al., 2013a, 2013b).

It is worth noting that the AFM parameters and their results are scale and mode dependent.
Therefore, AFM images that have been provided with the same scale and the same mode can
be compared together. Moreover, compared to SEM analysis, AFM is a non-vacuum analysis
technique, and the membrane sample is not coated. Therefore, the AFM results can be closer
to the real features of the membrane in real life (Shirazi et al., 2013b).

Mechanical properties
Although MD membranes do not require very strict mechanical properties compared to
membranes in pressure-driven processes, such as RO, a minimum mechanical strength is
still required for handling and modulation of MD membranes (Essalhi and Khayet, 2014;
Essalhi and Khayet, 2013). Tensile test can be employed to evaluate the mechanical strength
of MD membranes. Figure 10 illustrates the general scheme of a typical tensile measurement
system.

Force
measurement
Grips for Fixed head
holding
specimen
firmly

Test specimen

Fixed head Thickness 1/8”

Constant rate
of motion

Figure 10 A general scheme of the tensile test (Ismail et al., 2019).

119
Experimental Methods for Membrane Applications

To perform the tensile test, a piece of membrane sample should be cut according to ASTM
D882-10 and taped for both ends to secure the grip. The membrane sample should then
be placed and fixed in a grip between two jaws. To run the tensile test, some operating
parameters are important, such as the load cell (e.g., 10 N), and the cross-head speed of the
load cell (e.g., 5 or 10 mm/min). Based on the data recorded in the tensile machine, the stress-
strain curves can then be illustrated and investigated (Tijing et al., 2013; Wang et al., 2017).

5.3.2.2 Chemical properties

(a) Elemental composition

Energy-dispersive X-ray spectroscopy (EDS) is a well-known, analytical technique for
investigating the elemental composition of MD membranes. It usually performs along with
the SEM test. EDS can provide the elemental composition of the membrane surface after
modification (e.g., successful adsorption of nanomaterials on the membrane surface) or the
composition of the fouling layer on the membrane surface (Kamaz et al., 2019; Shirazi et
al., 2020a). It can also be used for investigating the uniform dispersion of nanoparticles on
surface or in the membrane matrix. To perform the EDS, when the surface of the membrane
sample is bombarded by SEM electron beams, the EDS detector can then detect the emitted
X-ray and analyse the elemental composition of the membrane surface (Yang et al., 2018).
X-ray photoelectron spectroscopy (XPS) is another practical technique for elemental
analysis of MD membranes (Khayet and Matsuura, 2003). XPS can provide useful and
detailed information such as the chemical formula and chemical composition of the
membrane surface. Thus, XPS can practically be employed to evaluate the efficiency of a
proposed surface modification technique (Suk et al., 2006; Wei et al., 2012; Zhao et al.,
2017).

(b) Functional groups

Fourier transform infrared (FTIR) spectroscopy is a practical, analytical technique to
investigate the functional groups on the membrane surface, either after surface modification
or deposition of foulants during the MD test (Korolkov et al., 2018). FTIR can detect
different bonding types as well as organic and inorganic functional groups in molecular level.
FTIR works based on the absorbance of the wavelength of the IR light of various functional
groups in a wide range (400 to 4000 cm-1). Thus, the detector can analyse the absorbed
wavelength and identify the target functional group, and its density, as well. When the FTIR
graph is plotted, the sharper peaks represent the higher amount of the specific functional
group. Likewise, the small peaks represent trace amount of the that group (Belfer et al.,
2000; Jhaveri & Murthy, 2016; Rahman et al., 2018). Moreover, same as SEM, FTIR is a
non-destructive technique, and thus samples can be used for further measurements.

(c ) Chemical structure
When a new membrane is fabricated or modified for investigating the MD performance, it
should also be evaluated for the chemical structure. This can be performed using the Raman
spectroscopy technique (Intrchom et al. 2018; Pouya et al., 2021). Raman technique can
also provide informative curves for determining molecular bonds, crystallinity, and the
orientation of polymeric chains. Moreover, Raman is a non-destructive analysis technique
(Bhadra et al., 2016; Dumée et al., 2011; Huang et al., 2020).

120
 Chapter 5

(d) Crystalline structure

Each material can be specified using special X-ray diffraction. This can then be used for
investigating the size of atoms, the crystal size, the length of chemical bonds, and the
layer spacing (Bunaciu et al., 2015). These have made X-ray diffraction (XRD) a powerful
method to determine the structural crystallinity of materials, such as inorganic substances
(e.g., metals, ceramics, etc.) (Kahle et al., 2002; Norby, 2006). Thus, this analytical technique
is practical for the characterization of inorganic and mixed matrix MD membranes, as
polymers usually have a low crystalline structure. Moreover, XRD can be applied for
investigating the inorganic scaling on the membrane surface and the produced crystals in
the MD crystallization process (Garofalo et al., 2016; Zuo and Chung, 2016; Gryta, 2011;
Mokhtar et al., 2014). Using XRD, the change in the chemical bonds and the crystallinity
of materials can be detected. When XRD data is plotted, the narrow and sharp peaks
indicate ordered material with large particle size, and vice a versa (Petkov, 2008; Flack and
Bernardinelli, 2008).

5.3.2.3 Thermal properties

(a) Thermal conductivity

Thermal properties of MD membranes are important for long term performance. This can
be highlighted in terms of heat loss through the membrane thermal conduction (Eykens et
al., 2016). The thermal conductivity of MD membranes can be calculated as follow:

km = (1− ε )ks + ε k g Eq. 7

where ε is the membrane porosity, and ks and kg are the thermal conductivity of polymer
and the trapped gas in the pores (e.g., air), respectively. As could be observed, the thermal
conduction is proportional to the membrane porosity. Thus, the higher the porosity, the
lower the membrane’s thermal conductivity. Therefore, membranes with higher porosity,
such nanofiber membranes, can perform better in MD experiments due to lower heat
loss (Alkhudhiri et al., 2012; Shirazi et al., 2020a). Moreover, the precise examination
of the membrane porosity along with accurate conductivity data for the used polymer
in membrane fabrication can provide better results for the thermal conduction of MD
membranes. However, in terms of mixed matrix MD membranes, other correlations should
be investigated (Eykens et al., 2016).

(b) Thermal stability

As the MD membrane works under a range of feed temperatures (40-80 ˚C), the thermal
stability matters in long-term performance (Saffarini et al., 2013; Susanto, 2011b; Gryta,
2005). The thermal stability of MD membranes can be investigated using the differential
scanning calorimetry (DSC) method (Ding et al., 2021; Prince et al., 2012). DSC provides
very useful structural information of the membrane sample at various temperatures (Khayet
et al., 2018; Essalhi et al., 2021; García-Fernández et al., 2015).

121
Experimental Methods for Membrane Applications

5.4 APPLICATIONS AND EXAMPLES

MD is traditionally used for desalination purposes as an alternative to RO or to overcome

the limited recoveries of RO and other thermal desalination techniques. Though the most
investigated application of MD has been desalination, the low fouling propensity of the
process, the ability to handle complex feed solutions, and the fact that separation occurs
through temperature-induced phase equilibria at specific locations have led to many other
interesting applications of MD being explored (Drioli et al., 2015a). The temperature
gradient-based nature of MD also opens new possibilities for use in vapor/gas separation
applications where the equilibrium composition contains more volatile components at
each temperature. As a result, the scope of the process has expanded beyond traditional
desalting applications. MD could also be operated in an osmotic configuration where the
mass transport through the membrane pores is driven by the osmotic pressure difference
across the membrane. This operational mode is interesting for temperature-sensitive
products such as pharmaceutical compounds, juices, dairy products, natural aromatic
compounds, and various chemical solutions. MD has also demonstrated the potential to
treat the solutions where the final effluent quality must meet strict criteria such as nuclear
waste, radioactive water, or water treatment for the semiconductor industry.

In the oil and gas sector, shale gas has been recognized as a game changer due to its abundant
availability in different parts of the world. However, the negative environmental impact of
shale gas exploration remains a major obstacle to large-scale adaptation. Water produced
along with the oil - so-called produced water - is a major contributor to the dangerous
environmental impacts of shale gas exploration. The produced water contains a very high
proportion of salts, various hydrocarbons, and production chemicals. Handling such
complex liquids using state-of-the-art processes is a real challenge. Additionally, the high
pressure and temperature of the water generated during the manufacturing process further
complicate the immediate handling. MD has been shown to be a potential candidate for
treating this water after certain physical processes that remove hydrocarbons from the
stream.

Traditional MBR also has the biggest pollution problem. MD as a standalone process or
integrated with other processes (such as FO) yielded very interesting results. Similarly,
the removal of heavy metals that act as trace contaminants is a challenge for existing RO
plants. For example, since boron exists as boric acid, under normal pH conditions it can
diffuse through RO membranes, so conventional RO fails to meet the required removal
criteria (0.5-2.5 ppm). Current alternative technologies are either expensive or not robust
to changing operating conditions. The application of MD successfully removed boron far
below the specified limit.

Another area of potential interest for MD is the recovery/removal of phosphorus from

agricultural, domestic, and industrial wastewater. The presence of phosphorus in soil
is essential for crop growth. Phosphorus influx, on the other hand, causes a condition
called eutrophication. Eutrophication is characterized by the excessive growth of algae
in the water, reducing oxygen levels and adversely affecting marine life. As phosphorus
reserves are limited, MD can be used as a stand-alone process or in combination with other

122
 Chapter 5

membrane-based processes to not only control the phosphorus content in wastewater,

but also to increase phosphorus enrichment. Phosphorus crystals can also be recovered
from large streams (Quist-Jensen et al., 2018b; Simoni et al., 2021). The same is true for
protein crystallization and crystallization of pharmaceutical compounds by membrane
crystallization.

5.5 OUTLOOK

MD has made great progress during the last two decades or so. It is expected that the process
will attract further research and commercial interest for sustainable desalination as well
as resource recovery from different liquid streams such as desalination and geothermal
brines and wastewaters. However, it should be highlighted that further road to progression
should not consider MD as the replacement of large-scale reverse osmosis units but rather
a complementary process to augment the deficiencies of RO. For instance, MD could be
used to concentrate the retentate of RO brine with the ambition to approach zero liquid
discharge in seawater and brackish water desalination. Due to its capability to run with
solar energy, MD can be a suitable candidate as a standalone desalination process in off-
grid water-scarce regions. Due to its capability to concentrate, the process could also be a
valuable tool to recover valuable dissolved components from different liquid streams. This
is evident from the current interest in the process of the concentration of lithium brines.
MD has also demonstrated the potential to produce electricity when operated in pressure
retarded mode. This could turn the process into the simultaneous producer of freshwater,
electricity, and raw materials (Rahman et al., 2023).

Realization of the full potential of the process, however, is associated with overcoming
some key challenges. On the membrane front, the development of membranes with long-
lasting hydrophobic and anti-scaling/fouling characteristics is desired. Further research and
development in material development and synthesis routes are needed to achieve this goal.
As MD is more feasible for the treatment of high-concentrated solutions, the membrane
scaling issues are expected to be more significant in MD than the conventional pressure-
driven membrane processes such as NF and RO. Therefore, the development of improved
techniques to overcome scaling issues is of paramount importance. In this context, the
development of anti-scaling membranes as well as appropriate pre-treatment strategies is
expected to offer an important contribution. MD is also gaining traction in food processing,
and treatment of oily water and organic-rich wastewater where the scaling issues will be
significantly higher than the conventionally investigated desalination applications. The
membrane and process development, therefore, should also consider the appropriate
strategies to tackle fouling issues when tackling such complex feed solutions. The presence
of natural organic matter like humic acid, carbohydrates, proteins, lipids, and other low
molecular weight species causes organic fouling in MD. Organic matter can adhere to the
membrane surface through hydrophobic interactions, chemical affinity, and electrostatic
forces. This adherence can cause reduced vapor permeability and can interfere with the
hydrophobic character of the membrane. However, currently, very little attention has
been devoted to developing strategies to tackle this type of fouling. Due to the low flux
and different separation mechanism than the conventional pressure-driven membrane
processes, particulate fouling has not received significant attention. However, the solid

123
Experimental Methods for Membrane Applications

particles inherently present in the feed or crystals precipitating from the feed can create
particulate fouling in MD as reported in the literature (Chimanlal et al., 2022). Therefore,
it becomes relevant to develop mitigation strategies for such fouling in MD. The possible
strategies include appropriate pre-treatments (e.g., filtration, chemical precipitation),
module designs, fouling-resistant membrane configurations and materials and optimized
process conditions. Biofouling, or biological fouling, is caused by the accumulation of
bacteria and living microorganisms on the surface of a membrane. It leads to the formation
of a biofilm, which can significantly reduce membrane performance and productivity.
Biofouling is less common in MD compared to other membrane processes, but it still
occurs, especially in MD bioreactors. Biofouling in MD inhibits the process through pore
wetting and pore blockage mechanisms, allowing particles to penetrate the permeate side
and causing distillate contamination. Factors such as feed flow rate, membrane properties,
microorganism properties, pH, and feed water source influence the attachment and growth
of microorganisms on the membrane surface. To control biofouling in MD, techniques
(appropriate pre-treatments, quorum quenching, membrane, and process design) developed
for other membrane processes could become of interest.

MD is also becoming increasingly relevant for the treatment of acidic wastewater (e.g., the
one originating from the battery recycling process). This will require the development of
membranes that are tolerable to exposure to the low-pH solutions.

Future efforts on the process front should focus on minimizing the electric as well as thermal
energy consumption of the process. This can be achieved by developing more energy-
efficient membranes, process configurations, and module designs. More rational integration
of different energy sources (solar, geothermal, and industrial) will also provide an important
contribution. In particular, the studies should focus more on the integration of the process
with geothermal heat, which is a more stable and broadly available source of energy.

124
 Chapter 5

5.6 REFERENCES

Abu-Zeid, M. A.E.R., X. Lu, and S. Zhang. 2021. Influence of Module Length on Water Desalination
Using Air Gap Membrane Distillation Process: An Experimental Comparative Study. Arabian
Journal for Science and Engineering 46, no. 8 (August 1): 7989–8008.
Abu-Zeid, M.A.E.-R., Y. Zhang, H. Dong, L. Zhang, H.-L. Chen, and L. Hou. 2015. A Comprehensive
Review of Vacuum Membrane Distillation Technique. Desalination 356 (January): 1–14.
Abu-Zeid, Mostafa Abd El Rady, and G. ElMasry. 2020. Experimental Evaluation of Two Consecutive
Air-Gap Membrane Distillation Modules with Heat Recovery. Water Science and Technology:
Water Supply 20, no. 5 (August 1): 1678–1691.
Adewole, J.K., H.M. al Maawali, T. Jafary, A. Firouzi, and H. Oladipo. 2022. A Review of Seawater
Desalination with Membrane Distillation: Material Development and Energy Requirements.
Water Supply 22, no. 12 (December 1): 8500–8526.
Afsari, M., H.K. Shon, and L.D. Tijing. 2021. Janus Membranes for Membrane Distillation: Recent
Advances and Challenges. Advances in Colloid and Interface Science 289 (March): 102362.
Ahmed, F.E., B.S. Lalia, and R. Hashaikeh. 2015. A Review on Electrospinning for Membrane
Fabrication: Challenges and Applications. Desalination 356 (January): 15–30.
Ahmed, F.E., B.S. Lalia, R. Hashaikeh, and N. Hilal. 2020. Alternative Heating Techniques in Membrane
Distillation: A Review. Desalination 496 (December 15): 114713.
Al-hadidi, A., Kemperman, A., Wessling, M. & van der Meer, W. (2008). The influence of membrane
properties on the silt density index. In: EDS (ed.) Membranes in drinking and industrial water.
Toulouse: EDS-INSA.
Ali, A., P. Aimar, and E. Drioli. 2015a. Effect of Module Design and Flow Patterns on Performance of
Membrane Distillation Process. CHEMICAL ENGINEERING JOURNAL 277: 368–377. http://
dx.doi.org/10.1016/j.cej.2015.04.108.
Ali, A., C.A. Quist-Jensen, F. Macedonio, and E. Drioli. 2016. Optimization of Module Length
for Continuous Direct Contact Membrane Distillation Process. Chemical Engineering and
Processing: Process Intensification 110 (December): 188–200.
Ali, Aamer, C. Quist-Jensen, F. Macedonio, and E. Drioli. 2015. Application of Membrane Crystallization
for Minerals’ Recovery from Produced Water. Membranes 5, no. 4 (November 24): 772–792.
Ali, Aamer, C.A. Quist-Jensen, E. Drioli, and F. Macedonio. 2018. Evaluation of Integrated
Microfiltration and Membrane Distillation/Crystallization Processes for Produced Water
Treatment. Desalination 434 (May): 161–168.
Ali, Aamer, C.A. Quist-Jensen, M.K. Jørgensen, A. Siekierka, M.L. Christensen, M. Bryjak, C. Hélix-
Nielsen, and E. Drioli. 2021. A Review of Membrane Crystallization, Forward Osmosis and
Membrane Capacitive Deionization for Liquid Mining. Resources, Conservation and Recycling
168 (May 1): 105273.
Ali, Aamer, C.A. Quist-Jensen, F. Macedonio, and E. Drioli. 2016. Optimization of Module Length and
Membrane Thickness for Membrane Distillation. In 2nd International Workshop on Membrane
Distillation and Innovating Membrane Operations in Desalination and Water Reuse, Ravello,
Italy. Ravello.
Ali, Aamer, J.-H. Tsai, K.-L. Tung, E. Drioli, and F. Macedonio. 2018. Designing and Optimization of
Continuous Direct Contact Membrane Distillation Process. Desalination 426 (January): 97–107.
Alkhudhiri, A., N. Darwish, and N. Hilal. 2012. Membrane Distillation: A Comprehensive Review.
Desalination 287 (February): 2–18.
Alklaibi, A.M., and N. Lior. 2005. Membrane-Distillation Desalination: Status and Potential.
Desalination 171, no. 2 (January 10): 111–131.

125
Experimental Methods for Membrane Applications

AlMarzooqi, F.A., M.R. Bilad, B. Mansoor, and H.A. Arafat. 2016. A Comparative Study of Image
Analysis and Porometry Techniques for Characterization of Porous Membranes. Journal of
Materials Science 51, no. 4 (February 27): 2017–2032.
Alpatova, A., A.S. Alsaadi, M. Alharthi, J.G. Lee, and N. Ghaffour. 2019. Co-Axial Hollow Fiber Module
for Air Gap Membrane Distillation. Journal of Membrane Science 578 (May 15): 172–182.
An, A.K., J. Guo, E.-J. Lee, S. Jeong, Y. Zhao, Z. Wang, and T. Leiknes. 2017. PDMS/PVDF Hybrid
Electrospun Membrane with Superhydrophobic Property and Drop Impact Dynamics for
Dyeing Wastewater Treatment Using Membrane Distillation. Journal of Membrane Science 525
(March): 57–67.
Anderson, D. M., Boerlage, S. F. E. & Dixon, M. B. (2017). Harmful Algal Blooms (HABs) and
Desalination: A Guide to Impacts, Monitoring and Management. In: UNESCO, I. O. C. o. (ed.)
IOC Manuals and Guides No. 78. Paris: UNESCO.
Ashoor, B.B., S. Mansour, A. Giwa, V. Dufour, and S.W. Hasan. 2016. Principles and Applications of
Direct Contact Membrane Distillation (DCMD): A Comprehensive Review. Desalination 398
(November): 222–246.
ASTM D4189 - 14 (2014). Standard Test Method for Silt Density Index (SDI) of Water. West
Conshohocken, PA: ASTM International.
Attia, H., D.J. Johnson, C.J. Wright, and N. Hilal. 2018a. Comparison between Dual-Layer
(Superhydrophobic–Hydrophobic) and Single Superhydrophobic Layer Electrospun Membranes
for Heavy Metal Recovery by Air-Gap Membrane Distillation. Desalination 439 (August): 31–45.
———. 2018b. Robust Superhydrophobic Electrospun Membrane Fabricated by Combination
of Electrospinning and Electrospraying Techniques for Air Gap Membrane Distillation.
Desalination 446 (November): 70–82.
Belfer, S., R. Fainchtain, Y. Purinson, and O. Kedem. 2000. Surface Characterization by FTIR-ATR
Spectroscopy of Polyethersulfone Membranes-Unmodified, Modified and Protein Fouled.
Journal of Membrane Science 172, no. 1–2 (July): 113–124.
Belfort, G., Davis, R. H. & Zydney, A. L. (1994). The behavior of suspensions and macromolecular
solutions in crossflow microfiltration. Journal of Membrane Science, 96, 1-58
Bhadra, M., S. Roy, and S. Mitra. 2016. Desalination across a Graphene Oxide Membrane via Direct
Contact Membrane Distillation. Desalination 378 (January): 37–43.
Bonyadi, S., and T.S. Chung. 2007. Flux Enhancement in Membrane Distillation by Fabrication of Dual
Layer Hydrophilic–Hydrophobic Hollow Fiber Membranes. Journal of Membrane Science 306,
no. 1–2 (December): 134–146.
Bunaciu, A.A., E. gabriela Udri tioiu, and H.Y. Aboul-Enein. 2015. X-Ray Diffraction: Instrumentation
and Applications. Critical Reviews in Analytical Chemistry 45, no. 4 (October 2): 289–299.
Camacho, L., L. Dumée, J. Zhang, J. Li, M. Duke, J. Gomez, and S. Gray. 2013. Advances in Membrane
Distillation for Water Desalination and Purification Applications. Water 5, no. 1 (January 25):
94–196.
Chamani, H., J. Woloszyn, T. Matsuura, D. Rana, and C.Q. Lan. 2021. Pore Wetting in Membrane
Distillation: A Comprehensive Review. Progress in Materials Science 122 (October 1): 100843.
Chen, L., P. Xu, and H. Wang. 2020. Interplay of the Factors Affecting Water Flux and Salt Rejection
in Membrane Distillation: A State-of-the-Art Critical Review. Water 12, no. 10 (October 13):
2841.
Cheng, D., J. Zhang, N. Li, D. Ng, S.R. Gray, and Z. Xie. 2018. Antiwettability and Performance Stability
of a Composite Hydrophobic/Hydrophilic Dual-Layer Membrane in Wastewater Treatment
by Membrane Distillation. Industrial & Engineering Chemistry Research 57, no. 28 (July 18):
9313–9322.

126
 Chapter 5

Chimanlal, I., L.N. Nthunya, C. Quist-Jensen, and H. Richards. 2022. Membrane Distillation
Crystallization for Water and Mineral Recovery: The Occurrence of Fouling and Its Control
during Wastewater Treatment. Frontiers in Chemical Engineering 4 (November 29).
Conners, T.E., and S. Banerjee. 2020. Surface Analysis of Paper. CRC Press.
Curcio, E., and E. Drioli. 2005. Membrane Distillation and Related Operations—A Review. Separation
& Purification Reviews 34, no. 1 (January): 35–86.
Dadari, S., M. Rahimi, and S. Zinadini. 2022. Novel Antibacterial and Antifouling PES Nanofiltration
Membrane Incorporated with Green Synthesized Nickel-Bentonite Nanoparticles for Heavy
Metal Ions Removal. Chemical Engineering Journal 431 (March): 134116.
Ding, Z., Z. Liu, and C. Xiao. 2021. Excellent Performance of Novel Superhydrophobic Composite
Hollow Membrane in the Vacuum Membrane Distillation. Separation and Purification
Technology 268 (August): 118603.
Drioli, E., A. Ali, S. Simone, F. Macedonio, S.A. AL-Jlil, F.S. al Shabonah, H.S. Al-Romaih, O. Al-Harbi,
A. Figoli, and A. Criscuoli. 2013. Novel PVDF Hollow Fiber Membranes for Vacuum and Direct
Contact Membrane Distillation Applications. Separation and Purification Technology 115
(August): 27–38.
Drioli, Enrico, A. Ali, and F. Macedonio. 2015. Membrane Distillation : Recent Developments and
Perspectives. Desalination 356: 56–84. http://dx.doi.org/10.1016/j.desal.2014.10.028.
Dumée, L., V. Germain, K. Sears, J. Schütz, N. Finn, M. Duke, S. Cerneaux, D. Cornu, and S. Gray. 2011.
Enhanced Durability and Hydrophobicity of Carbon Nanotube Bucky Paper Membranes in
Membrane Distillation. Journal of Membrane Science 376, no. 1–2 (July): 241–246.
Duong, H.C., D. Chuai, Y.C. Woo, H.K. Shon, L.D. Nghiem, and V. Sencadas. 2018. A Novel Electrospun,
Hydrophobic, and Elastomeric Styrene-Butadiene-Styrene Membrane for Membrane Distillation
Applications. Journal of Membrane Science 549 (March): 420–427.
El-Bourawi, M.S., Z. Ding, R. Ma, and M. Khayet. 2006. A Framework for Better Understanding
Membrane Distillation Separation Process. Journal of Membrane Science 285, no. 1–2
(November): 4–29.
Essalhi, M., and M. Khayet. 2013. Self-Sustained Webs of Polyvinylidene Fluoride Electrospun
Nanofibers at Different Electrospinning Times: 1. Desalination by Direct Contact Membrane
Distillation. Journal of Membrane Science 433 (April): 167–179.
———. 2014. Self-Sustained Webs of Polyvinylidene Fluoride Electrospun Nano-Fibers: Effects of
Polymer Concentration and Desalination by Direct Contact Membrane Distillation. Journal of
Membrane Science 454 (March): 133–143.
Essalhi, Mohamed, M. Khayet, N. Ismail, O. Sundman, and N. Tavajohi. 2021. Improvement of
Nanostructured Electrospun Membranes for Desalination by Membrane Distillation Technology.
Desalination 510 (August): 115086.
Eykens, L., K. de Sitter, C. Dotremont, L. Pinoy, and B. van der Bruggen. 2017. Membrane Synthesis for
Membrane Distillation: A Review. Separation and Purification Technology 182 (July): 36–51.
Eykens, Lies, K. de Sitter, C. Dotremont, L. Pinoy, and B. van der Bruggen. 2016. How To Optimize the
Membrane Properties for Membrane Distillation: A Review. Industrial & Engineering Chemistry
Research 55, no. 35 (September 7): 9333–9343.
Fane, A.G., R.W. Schofield, and C.J.D. Fell. 1987. The Efficient Use of Energy in Membrane Distillation.
Desalination 64 (January): 231–243.
Feng, S., Z. Zhong, Y. Wang, W. Xing, and E. Drioli. 2018. Progress and Perspectives in PTFE Membrane:
Preparation, Modification, and Applications. Journal of Membrane Science 549 (March): 332–
349.

127
Experimental Methods for Membrane Applications

Ferreira, R.K.M., H. Ramlow, C. Marangoni, and R.A.F. Machado. 2021. A Review on the Manufacturing
Techniques of Porous Hydrophobic Ceramic Membranes Applied to Direct Contact Membrane
Distillation. Advances in Applied Ceramics 120, no. 5–8 (November 17): 336–357.
Flack, H.D., and G. Bernardinelli. 2008. The Use of X-Ray Crystallography to Determine Absolute
Configuration. Chirality 20, no. 5 (May 15): 681–690.
Francis, L., F.E. Ahmed, and N. Hilal. 2022. Advances in Membrane Distillation Module Configurations.
Membranes 12, no. 1 (January 12): 81.
Francis, L., N. Ghaffour, A.S. Al-Saadi, and G.L. Amy. 2015. Submerged Membrane Distillation for
Seawater Desalination. Desalination and Water Treatment 55, no. 10 (September 4): 2741–
2746.
Franken, A.C.M., J.A.M. Nolten, M.H.V. Mulder, D. Bargeman, and C.A. Smolders. 1987. Wetting
Criteria for the Applicability of Membrane Distillation. Journal of Membrane Science 33, no. 3
(October): 315–328.
García-Fernández, L., M. Khayet, and M.C. García-Payo. 2015. Membranes Used in Membrane
Distillation: Preparation and Characterization. In Pervaporation, Vapour Permeation and
Membrane Distillation, 317–359. Elsevier.
Garofalo, A., M.C. Carnevale, L. Donato, E. Drioli, O. Alharbi, S.A. Aljlil, A. Criscuoli, and C. Algieri.
2016. Scale-up of MFI Zeolite Membranes for Desalination by Vacuum Membrane Distillation.
Desalination 397 (November): 205–212.
Gontarek-Castro, E., R. Castro-Muñoz, and M. Lieder. 2022. New Insights of Nanomaterials Usage
toward Superhydrophobic Membranes for Water Desalination via Membrane Distillation: A
Review. Critical Reviews in Environmental Science and Technology 52, no. 12 (June 18): 2104–
2149.
Gopal, R., S. Kaur, Z. Ma, C. Chan, S. Ramakrishna, and T. Matsuura. 2006. Electrospun Nanofibrous
Filtration Membrane. Journal of Membrane Science 281, no. 1–2 (September 15): 581–586.
Gryta, M. 2005. Long-Term Performance of Membrane Distillation Process. Journal of Membrane
Science 265, no. 1–2 (November 15): 153–159.
———. 2007. Influence of Polypropylene Membrane Surface Porosity on the Performance of Membrane
Distillation Process. Journal of Membrane Science 287, no. 1 (January 5): 67–78.
Gryta, M. 2011. The Influence of Magnetic Water Treatment on CaCO3 Scale Formation in Membrane
Distillation Process. Separation and Purification Technology 80, no. 2 (July): 293–299.
———. 2020. The Application of Submerged Modules for Membrane Distillation. Membranes 10, no.
2 (February 1).
Guillen, G.R., T.P. Farrell, R.B. Kaner, and E.M. v. Hoek. 2010. Pore-Structure, Hydrophilicity, and
Particle Filtration Characteristics of Polyaniline–Polysulfone Ultrafiltration Membranes. Journal
of Materials Chemistry 20, no. 22: 4621.
He, K., H.J. Hwang, M.W. Woo, and I.S. Moon. 2011. Production of Drinking Water from Saline
Water by Direct Contact Membrane Distillation (DCMD). Journal of Industrial and Engineering
Chemistry 17, no. 1 (January): 41–48.
He, Y., J. Wang, X. Fu, H. Lin, W. Zhang, X. Wang, and F. Liu. 2023. Electrosprayed Thin Film
Nanocomposite Polyamide Nanofiltration with Homogeneous Distribution of Nanoparticles
for Enhanced Separation Performance. Desalination 546 (January): 116206.
Hilal, N., H. Al-Zoubi, N.A. Darwish, A.W. Mohamma, and M. Abu Arabi. 2004. A Comprehensive
Review of Nanofiltration Membranes:Treatment, Pretreatment, Modelling, and Atomic Force
Microscopy. Desalination 170, no. 3 (November): 281–308.

128
 Chapter 5

Hitsov, I., T. Maere, K. de Sitter, C. Dotremont, and I. Nopens. 2015. Modelling Approaches in
Membrane Distillation: A Critical Review. Separation and Purification Technology 142 (March
4): 48–64.
Huang, J., Y. Hu, Y. Bai, Y. He, and J. Zhu. 2020. Novel Solar Membrane Distillation Enabled by a
PDMS/CNT/PVDF Membrane with Localized Heating. Desalination 489 (September): 114529.
Huayan, C., W. Chunrui, J. Yue, W. Xuan, and L. Xiaolong. 2011. Comparison of Three Membrane
Distillation Confi Gurations and Seawater Desalination by Vacuum Membrane Distillation.
Desalination and Water Treatment 28, no. 1–3 (April 3): 321–327.
Hussain, A., A. Janson, J.M. Matar, and S. Adham. 2022. Membrane Distillation: Recent Technological
Developments and Advancements in Membrane Materials. Emergent Materials 5, no. 2 (April
5): 347–367.
Intrchom, W., S. Roy, M. Humoud, and S. Mitra. 2018. Immobilization of Graphene Oxide on the
Permeate Side of a Membrane Distillation Membrane to Enhance Flux. Membranes 8, no. 3
(August 15): 63.
Ismail, A.F., K.C. Khulbe, and T. Matsuura. 2019. RO Membrane Characterization. In Reverse Osmosis,
57–90. Elsevier.
Ismail, M.F., M.A. Islam, B. Khorshidi, A. Tehrani-Bagha, and M. Sadrzadeh. 2022. Surface
Characterization of Thin-Film Composite Membranes Using Contact Angle Technique: Review
of Quantification Strategies and Applications. Advances in Colloid and Interface Science 299
(January): 102524.
Jantaporn, W., A. Ali, and P. Aimar. 2017. Specific Energy Requirement of Direct Contact Membrane
Distillation. Chemical Engineering Research and Design 128 (December): 15–26.
Jena, A., and K. Gupta. 2010. Advances in Pore Structure Evaluation by Porometry. Chemical
Engineering & Technology 33, no. 8 (June 25): 1241–1250.
Jena, Akshaya, and K. Gupta. 2005. Pore Volume of Nanofiber Nonwovens. International Nonwovens.
Journal os-14, no. 2 (June 10): 1558925005os–14.
Jhaveri, J.H., and Z.V.P. Murthy. 2016. A Comprehensive Review on Anti-Fouling Nanocomposite
Membranes for Pressure Driven Membrane Separation Processes. Desalination 379 (February):
137–154.
Ji, J., F. Liu, N.A. Hashim, M.R.M. Abed, and K. Li. 2015. Poly(Vinylidene Fluoride) (PVDF) Membranes
for Fluid Separation. Reactive and Functional Polymers 86 (January): 134–153.
Johnson, D., and N. Hilal. 2015. Characterisation and Quantification of Membrane Surface Properties
Using Atomic Force Microscopy: A Comprehensive Review. Desalination 356 (January): 149–
164.
Julian, H., N. Nurgirisia, G. Qiu, Y.-P. Ting, and I.G. Wenten. 2022. Membrane Distillation for
Wastewater Treatment: Current Trends, Challenges and Prospects of Dense Membrane
Distillation. Journal of Water Process Engineering 46 (April): 102615.
Kahle, M., M. Kleber, and R. Jahn. 2002. Review of XRD-Based Quantitative Analyses of Clay Minerals
in Soils: The Suitability of Mineral Intensity Factors. Geoderma 109, no. 3–4 (October): 191–
205.
Kalla, S., S. Upadhyaya, and K. Singh. 2019. Principles and Advancements of Air Gap Membrane
Distillation. Reviews in Chemical Engineering 35, no. 7 (October 25): 817–859.
Kamaz, M., A. Sengupta, A. Gutierrez, Y.-H. Chiao, and R. Wickramasinghe. 2019. Surface Modification
of PVDF Membranes for Treating Produced Waters by Direct Contact Membrane Distillation.
International Journal of Environmental Research and Public Health 16, no. 5 (February 26): 685.
Kang, G., and Y. Cao. 2014. Application and Modification of Poly(Vinylidene Fluoride) (PVDF)
Membranes – A Review. Journal of Membrane Science 463 (August): 145–165.

129
Experimental Methods for Membrane Applications

Khayet, M., and C. Cojocaru. 2012. Air Gap Membrane Distillation: Desalination, Modeling and
Optimization. Desalination 287 (February): 138–145.
Khayet, M., M.C. García-Payo, L. García-Fernández, and J. Contreras-Martínez. 2018. Dual-Layered
Electrospun Nanofibrous Membranes for Membrane Distillation. Desalination 426 (January):
174–184.
Khayet, M., M.P. Godino, and J.I. Mengual. 2003. Possibility of Nuclear Desalination through Various
Membrane Distillation Configurations: A Comparative Study. International Journal of Nuclear
Desalination 1, no. 1: 30.
Khayet, M., K. Khulbe, and T. Matsuura. 2004. Characterization of Membranes for Membrane
Distillation by Atomic Force Microscopy and Estimation of Their Water Vapor Transfer
Coefficients in Vacuum Membrane Distillation Process. Journal of Membrane Science 238, no.
1–2 (July 15): 199–211.
Khayet, M., and T. Matsuura. 2003. Application of Surface Modifying Macromolecules for the Preparation
of Membranes for Membrane Distillation. Desalination 158, no. 1–3 (August): 51–56.
Khayet, Mohamed. 2011. Membranes and Theoretical Modeling of Membrane Distillation: A Review.
Advances in Colloid and Interface Science 164, no. 1–2 (May): 56–88.
Khayet, Mohamed, and T. Matsuura. 2001. Preparation and Characterization of Polyvinylidene Fluoride
Membranes for Membrane Distillation. Industrial & Engineering Chemistry Research 40, no. 24
(November 1): 5710–5718.
Khulbe, K.C., and T. Matsuura. 2017. Synthetic Membrane Characterisation – a Review: Part I.
Membrane Technology 2017, no. 7 (July): 7–12.
Kim, W.-J., O. Campanella, and D.R. Heldman. 2021. Predicting the Performance of Direct Contact
Membrane Distillation (DCMD): Mathematical Determination of Appropriate Tortuosity Based
on Porosity. Journal of Food Engineering 294 (April): 110400.
Kimura, S., S.-I. Nakao, and S.-I. Shimatani. 1987. Transport Phenomena in Membrane Distillation.
Journal of Membrane Science 33, no. 3 (October): 285–298.
Kirk, T.L. 2017. A Review of Scanning Electron Microscopy in Near Field Emission Mode. In , 39–109.
Kolb, H.E., R. Schmitt, A. Dittler, and G. Kasper. 2018. On the Accuracy of Capillary Flow Porometry
for Fibrous Filter Media. Separation and Purification Technology 199 (June): 198–205.
Korolkov, I. v., Y.G. Gorin, A.B. Yeszhanov, A.L. Kozlovskiy, and M. v. Zdorovets. 2018. Preparation
of PET Track-Etched Membranes for Membrane Distillation by Photo-Induced Graft
Polymerization. Materials Chemistry and Physics 205 (February): 55–63.
Lawson, K.W., and D.R. Lloyd. 1997. Membrane Distillation. Journal of Membrane Science 124, no. 1
(February): 1–25.
Lewczuk, B., and N. Szyrynska. 2021. Field-Emission Scanning Electron Microscope as a Tool for Large-
Area and Large-Volume Ultrastructural Studies. Animals 11, no. 12 (November 27): 3390.
Li, D., M.W. Frey, and Y.L. Joo. 2006. Characterization of Nanofibrous Membranes with Capillary Flow
Porometry. Journal of Membrane Science 286, no. 1–2 (December): 104–114.
Lu, K.J., Y. Chen, and T.-S. Chung. 2019. Design of Omniphobic Interfaces for Membrane Distillation –
A Review. Water Research 162 (October): 64–77.
Lu, K.-J., J. Zuo, and T.-S. Chung. 2016. Tri-Bore PVDF Hollow Fibers with a Super-Hydrophobic
Coating for Membrane Distillation. Journal of Membrane Science 514 (September): 165–175.
Mat, N.C., Y. Lou, and G.G. Lipscomb. 2014. Hollow Fiber Membrane Modules. Current Opinion in
Chemical Engineering 4: 18–24.
McGaughey, A.L., P. Karandikar, M. Gupta, and A.E. Childress. 2020. Hydrophobicity versus Pore Size:
Polymer Coatings to Improve Membrane Wetting Resistance for Membrane Distillation. ACS
Applied Polymer Materials 2, no. 3 (March 13): 1256–1267.

130
 Chapter 5

Meng, S., Y.C. Hsu, Y. Ye, and V. Chen. 2015. Submerged Membrane Distillation for Inland Desalination
Applications. Desalination 361 (April 1): 72–80.
Mohammadi, T., P. Kazemi, and M. Peydayesh. 2015. Optimization of Vacuum Membrane Distillation
Parameters for Water Desalination Using Box–Behnken Design. Desalination and Water
Treatment 56, no. 9 (November 27): 2306–2315.
Mokhtar, N.M., W.J. Lau, A.F. Ismail, and B.C. Ng. 2014. Physicochemical Study of Polyvinylidene
Fluoride–Cloisite15A® Composite Membranes for Membrane Distillation Application. RSC
Adv. 4, no. 108: 63367–63379.
Mousa, H.M., M. Hamdy, M.A. Yassin, M.M. El-Sayed Seleman, and G.T. Abdel-Jaber. 2022.
Characterization of Nanofiber Composite Membrane for High Water Flux and Antibacterial
Properties. Colloids and Surfaces A: Physicochemical and Engineering Aspects 651 (October):
129655.
Nakao, S. 1994. Determination of Pore Size and Pore Size Distribution. Journal of Membrane Science
96, no. 1–2 (November): 131–165.
Naresh-Kumar, G., B. Hourahine, P.R. Edwards, A.P. Day, A. Winkelmann, A.J. Wilkinson, P.J.
Parbrook, G. England, and C. Trager-Cowan. 2012. Rapid Nondestructive Analysis of Threading
Dislocations in Wurtzite Materials Using the Scanning Electron Microscope. Physical Review
Letters 108, no. 13 (March 30): 135503.
Niknejad, A. S., S. Bazgir, M. Ardjmand, and M.M.A. Shirazi. 2021. Spent Caustic Wastewater Treatment
Using Direct Contact Membrane Distillation with Electroblown Styrene-Acrylonitrile
Membrane. International Journal of Environmental Science and Technology 18, no. 8 (August
14): 2283–2294.
Niknejad, Ali Sallakh, S. Bazgir, and A. Kargari. 2021. Desalination by Direct Contact Membrane
Distillation Using a Superhydrophobic Nanofibrous Poly (Methyl Methacrylate) Membrane.
Desalination 511 (September): 115108.
Niknejad, Ali Sallakh, S. Bazgir, A. Sadeghzadeh, and M.M.A. Shirazi. 2021. Evaluation of a Novel and
Highly Hydrophobic Acrylonitrile-Butadiene-Styrene Membrane for Direct Contact Membrane
Distillation: Electroblowing/Air-Assisted Electrospraying Techniques. Desalination 500
(March): 114893.
Norby, P. 2006. In-Situ XRD as a Tool to Understanding Zeolite Crystallization. Current Opinion in
Colloid & Interface Science 11, no. 2–3 (June): 118–125.
Olatunji, S.O., and L.M. Camacho. 2018. Heat and Mass Transport in Modeling Membrane Distillation
Configurations: A Review. Frontiers in Energy Research 6 (December 4).
Omar, N.M.A., M.H.D. Othman, Z.S. Tai, T.A. Kurniawan, T. El-badawy, P.S. Goh, N.H. Othman, M.A.
Rahman, J. Jaafar, and A.F. Ismail. 2022. Bottlenecks and Recent Improvement Strategies of
Ceramic Membranes in Membrane Distillation Applications: A Review. Journal of the European
Ceramic Society 42, no. 13 (October): 5179–5194.
Pangarkar, B.L., and S.K. Deshmukh. 2015. Theoretical and Experimental Analysis of Multi-Effect
Air Gap Membrane Distillation Process (ME-AGMD). Journal of Environmental Chemical
Engineering 3, no. 3 (September 29): 2127–2135.
Park, S.H., J.H. Kim, S.J. Moon, J.T. Jung, H.H. Wang, A. Ali, C.A. Quist-Jensen, F. Macedonio, E. Drioli,
and Y.M. Lee. 2020. Lithium Recovery from Artificial Brine Using Energy-Efficient Membrane
Distillation and Nanofiltration. Journal of Membrane Science 598 (March): 117683.
Peng, N., N. Widjojo, P. Sukitpaneenit, M.M. Teoh, G.G. Lipscomb, T.-S. Chung, and J.-Y. Lai. 2012.
Evolution of Polymeric Hollow Fibers as Sustainable Technologies: Past, Present, and Future.
Progress in Polymer Science 37, no. 10 (October): 1401–1424.

131
Experimental Methods for Membrane Applications

Petkov, V. 2008. Nanostructure by High-Energy X-Ray Diffraction. Materials Today 11, no. 11
(November): 28–38.
Peydayesh, M., P. Kazemi, A. Bandegi, T. Mohammadi, and O. Bakhtiari. 2015. Treatment of Bentazon
Herbicide Solutions by Vacuum Membrane Distillation. Journal of Water Process Engineering 8
(December): e17–e22.
Phattaranawik, J., R. Jiraratananon, and A.G. Fane. 2003. Heat Transport and Membrane Distillation
Coefficients in Direct Contact Membrane Distillation. Journal of Membrane Science 212, no. 1–2
(February): 177–193.
Pouya, Z.A., M.A. Tofighy, and T. Mohammadi. 2021. Synthesis and Characterization of
Polytetrafluoroethylene/Oleic Acid-Functionalized Carbon Nanotubes Composite Membrane
for Desalination by Vacuum Membrane Distillation. Desalination 503 (May): 114931.
Prince, J.A., G. Singh, D. Rana, T. Matsuura, V. Anbharasi, and T.S. Shanmugasundaram. 2012.
Preparation and Characterization of Highly Hydrophobic Poly(Vinylidene Fluoride) – Clay
Nanocomposite Nanofiber Membranes (PVDF–Clay NNMs) for Desalination Using Direct
Contact Membrane Distillation. Journal of Membrane Science 397–398 (April): 80–86.
Qasim, M., I.U. Samad, N.A. Darwish, and N. Hilal. 2021. Comprehensive Review of Membrane Design
and Synthesis for Membrane Distillation. Desalination 518 (December): 115168.
Qin, L., I.A. Mergos, and H. Verweij. 2015. Obtaining Accurate Cross-Section Images of Supported
Polymeric and Inorganic Membrane Structures. Journal of Membrane Science 476 (February):
194–199.
Qtaishat, M., T. Matsuura, B. Kruczek, and M. Khayet. 2008. Heat and Mass Transfer Analysis in Direct
Contact Membrane Distillation. Desalination 219, no. 1–3 (January): 272–292.
Quist-Jensen, C. A., J.M. Sorensen, A. Svenstrup, L. Scarpa, T.S. Carlsen, H.C. Jensen, L. Wybrandt, and
M.L. Christensen. 2018a. Membrane Crystallization for Phosphorus Recovery and Ammonia
Stripping from Reject Water from Sludge Dewatering Process. Desalination 440 (August 15):
156–160.
Quist-Jensen, C.A., F. Macedonio, C. Conidi, A. Cassano, S. Aljlil, O.A. Alharbi, and E. Drioli. 2016.
Direct Contact Membrane Distillation for the Concentration of Clarified Orange Juice. Journal of
Food Engineering 187 (October): 37–43.
Quist-Jensen, Cejna Anna, A. Ali, S. Mondal, F. Macedonio, and E. Drioli. 2016a. A Study of Membrane
Distillation and Crystallization for Lithium Recovery from High-Concentrated Aqueous
Solutions. Journal of Membrane Science 505 (May): 167–173.
Quist-Jensen, Cejna Anna, F. Macedonio, D. Horbez, and E. Drioli. 2017. Reclamation of Sodium Sulfate
from Industrial Wastewater by Using Membrane Distillation and Membrane Crystallization.
Desalination 401 (January): 112–119.
Rácz, G., S. Kerker, Z. Kovács, G. Vatai, M. Ebrahimi, and P. Czermak. 2014. Theoretical and
Experimental Approaches of Liquid Entry Pressure Determination in Membrane Distillation
Processes. Periodica Polytechnica Chemical Engineering 58, no. 2: 81–91.
Rahman, M.M., S. Al-Sulaimi, and A.M. Farooque. 2018. Characterization of New and Fouled SWRO
Membranes by ATR/FTIR Spectroscopy. Applied Water Science 8, no. 6 (October 26): 183.
Rahman, S.N., H. Saleem, and S.J. Zaidi. 2023. Progress in Membranes for Pressure Retarded Osmosis
Application. Desalination 549 (March): 116347.
Ramlow, H., R.K.M. Ferreira, C. Marangoni, and R.A.F. Machado. 2019. Ceramic Membranes Applied
to Membrane Distillation: A Comprehensive Review. International Journal of Applied Ceramic
Technology 16, no. 6 (November 13): 2161–2172.
Ravi, J., M.H.D. Othman, T. Matsuura, M. Ro’il Bilad, T.H. El-badawy, F. Aziz, A.F. Ismail, M.A. Rahman,
and J. Jaafar. 2020. Polymeric Membranes for Desalination Using Membrane Distillation: A
Review. Desalination 490 (September): 114530.
132
 Chapter 5

Rezaei, M., D.M. Warsinger, J.H. Lienhard V, M.C. Duke, T. Matsuura, and W.M. Samhaber. 2018.
Wetting Phenomena in Membrane Distillation: Mechanisms, Reversal, and Prevention. Water
Research 139 (August): 329–352.
Sadeghi Ghari, H., Z. Shakouri, and M.M.A. Shirazi. 2014. Evaluation of Microstructure of Natural
Rubber/Nano-Calcium Carbonate Nanocomposites by Solvent Transport Properties. Plastics,
Rubber and Composites 43, no. 6.
Sadeghzadeh, A., S. Bazgir, and M.M.A. Shirazi. 2020. Fabrication and Characterization of a Novel
Hydrophobic Polystyrene Membrane Using Electroblowing Technique for Desalination by
Direct Contact Membrane Distillation. Separation and Purification Technology 239.
Sadeghzadeh, Amirhossein, S. Bazgir, and M.M.A. Shirazi. 2020. Fabrication and Characterization of a
Novel Hydrophobic Polystyrene Membrane Using Electroblowing Technique for Desalination
by Direct Contact Membrane Distillation. Separation and Purification Technology 239 (May):
116498.
Saffarini, R.B., B. Mansoor, R. Thomas, and H.A. Arafat. 2013. Effect of Temperature-Dependent
Microstructure Evolution on Pore Wetting in PTFE Membranes under Membrane Distillation
Conditions. Journal of Membrane Science 429 (February): 282–294.
Said, I.A., T. Chomiak, J. Floyd, and Q. Li. 2020. Sweeping Gas Membrane Distillation (SGMD) for
Wastewater Treatment, Concentration, and Desalination: A Comprehensive Review. Chemical
Engineering and Processing - Process Intensification 153 (July): 107960.
Salinas Rodríguez, S. G., Amy, G. L., Schippers, J. C. & Kennedy, M. D. (2015). The Modified Fouling
Index Ultrafiltration Constant Flux for assessing particulate/colloidal fouling of RO systems.
Desalination, 365, 79-91.
Sanaeepur, H., A. Ebadi Amooghin, M.M.A. Shirazi, M. Pishnamazi, and S. Shirazian. 2022. Water
Desalination and Ion Removal Using Mixed Matrix Electrospun Nanofibrous Membranes: A
Critical Review. Desalination 521 (January): 115350.
Schofield, R.W., A.G. Fane, and C.J.D. Fell. 1987. Heat and Mass Transfer in Membrane Distillation.
Journal of Membrane Science 33, no. 3 (October): 299–313.
Schofield, R.W., A.G. Fane, C.J.D. Fell, and R. Macoun. 1990. Factors Affecting Flux in Membrane
Distillation. Desalination 77 (March): 279–294.
Shahu, V.T., and S.B. Thombre. 2019a. Air Gap Membrane Distillation: A Review. Journal of Renewable
and Sustainable Energy 11, no. 4 (July): 045901.
———. 2021. Experimental Analysis of a Novel Helical Air Gap Membrane Distillation System. Water
Science and Technology: Water Supply 21, no. 4 (June 1): 1450–1463.
Shalaby, S.M., A.E. Kabeel, H.F. Abosheiasha, M.K. Elfakharany, E. El-Bialy, A. Shama, and R.D. Vidic.
2022. Membrane Distillation Driven by Solar Energy: A Review. Journal of Cleaner Production
366 (September 15): 132949.
Sharma, S., C.N. Shyam Kumar, J.G. Korvink, and C. Kübel. 2018. Evolution of Glassy Carbon
Microstructure: In Situ Transmission Electron Microscopy of the Pyrolysis Process. Scientific
Reports 8, no. 1 (November 2): 16282.
Sharma, V., and A. Bhardwaj. 2019. Scanning Electron Microscopy (SEM) in Food Quality Evaluation.
In Evaluation Technologies for Food Quality, 743–761. Elsevier.
Shirazi, M. M.A., A. Kargari, S. Bazgir, M. Tabatabaei, M.J.A. Shirazi, M.S. Abdullah, T. Matsuura, and
A.F. Ismail. 2013. Characterization of Electrospun Polystyrene Membrane for Treatment
of Biodiesel’s Water-Washing Effluent Using Atomic Force Microscopy. Desalination 329
(November 15): 1–8.
Shirazi, A., M., and A. Kargari. 2019. Concentrating of Sugar Syrup in Bioethanol Production Using
Sweeping Gas Membrane Distillation. Membranes 9, no. 5 (May 1): 59. Shirazi, M.J.A., S.

133
Experimental Methods for Membrane Applications

Bazgir, M.M.A. Shirazi, and S. Ramakrishna. 2013. Coalescing Filtration of Oily Wastewaters:
Characterization and Application of Thermal Treated, Electrospun Polystyrene Filters.
Desalination and Water Treatment 51, no. 31–33 (September): 5974–5986.
Shirazi, M.M.A., A. Kargari, S. Bazgir, M. Tabatabaei, M.J.A. Shirazi, M.S. Abdullah, T. Matsuura, and
A.F. Ismail. 2013. Characterization of Electrospun Polystyrene Membrane for Treatment
of Biodiesel’s Water-Washing Effluent Using Atomic Force Microscopy. Desalination 329
(November): 1–8.
Shirazi, M.M.A., A. Kargari, and M.J.A. Shirazi. 2012. Direct Contact Membrane Distillation for Seawater
Desalination. Desalination and Water Treatment 49, no. 1–3.
Shirazi, M.M.A., A. Kargari, and M. Tabatabaei. 2014. Evaluation of Commercial PTFE Membranes in
Desalination by Direct Contact Membrane Distillation. Chemical Engineering and Processing:
Process Intensification 76.
Shirazi, Mohammad Mahdi A., D. Bastani, A. Kargari, and M. Tabatabaei. 2013. Characterization of
Polymeric Membranes for Membrane Distillation Using Atomic Force Microscopy. Desalination
and Water Treatment 51, no. 31–33 (September): 6003–6008.
Shirazi, Mohammad Mahdi A., S. Bazgir, and F. Meshkani. 2020a. A Novel Dual-Layer, Gas-Assisted
Electrospun, Nanofibrous SAN4-HIPS Membrane for Industrial Textile Wastewater Treatment
by Direct Contact Membrane Distillation (DCMD). Journal of Water Process Engineering 36
(August): 101315.
———. 2020b. A Dual-Layer, Nanofibrous Styrene-Acrylonitrile Membrane with Hydrophobic/
Hydrophilic Composite Structure for Treating the Hot Dyeing Effluent by Direct Contact
Membrane Distillation. Chemical Engineering Research and Design 164 (December): 125–146.
Shirazi, Mohammad Mahdi A., and L.F. Dumée. 2022. Membrane Distillation for Sustainable
Wastewater Treatment. Journal of Water Process Engineering 47 (June 1): 102670.
Shirazi, Mohammad Mahdi A., A. Kargari, A.F. Ismail, and T. Matsuura. 2016. Computational Fluid
Dynamic (CFD) Opportunities Applied to the Membrane Distillation Process: State-of-the-Art
and Perspectives. Desalination 377 (January 1): 73–90.
Shirazi, Mohammad Mahdi A., A. Kargari, M. Tabatabaei, A.F. Ismail, and T. Matsuura. 2014.
Concentration of Glycerol from Dilute Glycerol Wastewater Using Sweeping Gas Membrane
Distillation. Chemical Engineering and Processing: Process Intensification 78 (April): 58–66.
Shyam Kumar, C.N., V.S.K. Chakravadhanula, A. Riaz, S. Dehm, D. Wang, X. Mu, B. Flavel, R.
Krupke, and C. Kübel. 2017. Understanding the Graphitization and Growth of Free-Standing
Nanocrystalline Graphene Using in Situ Transmission Electron Microscopy. Nanoscale 9, no. 35:
12835–12842.
Silva, R. de S., H. Ramlow, B. de C. Santos, H.B. Madalosso, R.A.F. Machado, and C. Marangoni. 2021.
Membrane Distillation: Experimental Evaluation of Liquid Entry Pressure in Commercial
Membranes with Textile Dye Solutions. Journal of Water Process Engineering 44 (December):
102339.
Simoni, G., B.S. Kirkebæk, C.A. Quist-Jensen, M.L. Christensen, and A. Ali. 2021. A Comparison of
Vacuum and Direct Contact Membrane Distillation for Phosphorus and Ammonia Recovery
from Wastewater. Journal of Water Process Engineering 44 (December 1): 102350.
Suk, D., T. Matsuura, H. Park, and Y. Lee. 2006. Synthesis of a New Type of Surface Modifying
Macromolecules (NSMM) and Characterization and Testing of NSMM Blended Membranes for
Membrane Distillation. Journal of Membrane Science 277, no. 1–2 (June 1): 177–185.
Susanto, H. 2011a. Towards Practical Implementations of Membrane Distillation. Chemical
Engineering and Processing: Process Intensification 50, no. 2 (February): 139–150.

134
 Chapter 5

Tai, Z.S., M.H.A. Aziz, M.H.D. Othman, A.F. Ismail, M.A. Rahman, and J. Jaafar. 2019. An Overview of
Membrane Distillation. In Membrane Separation Principles and Applications, 251–281. Elsevier.
Talukder, M.E., Md.N. Pervez, W. Jianming, G.K. Stylios, M.M. Hassan, H. Song, V. Naddeo, and A.
Figoli. 2022. Ag Nanoparticles Immobilized Sulfonated Polyethersulfone/Polyethersulfone
Electrospun Nanofiber Membrane for the Removal of Heavy Metals. Scientific Reports 12, no.
1 (April 6): 5814.
Tanis-Kanbur, M.B., R.I. Peinador, J.I. Calvo, A. Hernández, and J.W. Chew. 2021. Porosimetric
Membrane Characterization Techniques: A Review. Journal of Membrane Science 619
(February): 118750.
Tibi, F., A. Charfi, J. Cho, and J. Kim. 2020. Fabrication of Polymeric Membranes for Membrane
Distillation Process and Application for Wastewater Treatment: Critical Review. Process Safety
and Environmental Protection 141 (September): 190–201.
Tijing, L.D., J.-S. Choi, S. Lee, S.-H. Kim, and H.K. Shon. 2014. Recent Progress of Membrane Distillation
Using Electrospun Nanofibrous Membrane. Journal of Membrane Science 453 (March): 435–
462.
Tijing, L.D., W. Choi, Z. Jiang, A. Amarjargal, C.-H. Park, H.R. Pant, I.-T. Im, and C.S. Kim. 2013. Two-
Nozzle Electrospinning of (MWNT/PU)/PU Nanofibrous Composite Mat with Improved
Mechanical and Thermal Properties. Current Applied Physics 13, no. 7 (September): 1247–1255.
Tijing, L.D., J.R.C. Dizon, I. Ibrahim, A.R.N. Nisay, H.K. Shon, and R.C. Advincula. 2020. 3D Printing
for Membrane Separation, Desalination and Water Treatment. Applied Materials Today 18
(March 1): 100486.
Tsai, J.H., C. Quist-Jensen, and A. Ali. 2023. Multipass Hollow Fiber Membrane Modules for Membrane
Distillation. Desalination 548 (February 15).
Tung, K.-L., K.-S. Chang, T.-T. Wu, N.-J. Lin, K.-R. Lee, and J.-Y. Lai. 2014. Recent Advances in the
Characterization of Membrane Morphology. Current Opinion in Chemical Engineering 4 (May):
121–127.
Vicente, J., Y. Wyart, and P. Moulin. 2013. Characterization (two-dimensional - three-dimensional)
of ceramic microfiltration membrane by synchrotron radiation: new and abraded membranes.
Journal of Porous Media 16, no. 6: 537–545.
Vladár, A.E., and V.-D. Hodoroaba. 2020. Characterization of Nanoparticles by Scanning Electron
Microscopy. In Characterization of Nanoparticles, 7–27. Elsevier.
Wang, K., A.A. Abdalla, M.A. Khaleel, N. Hilal, and M.K. Khraisheh. 2017. Mechanical Properties of
Water Desalination and Wastewater Treatment Membranes. Desalination 401 (January): 190–
205.
Wang, P., and T.-S. Chung. 2015. Recent Advances in Membrane Distillation Processes: Membrane
Development, Configuration Design and Application Exploring. Journal of Membrane Science
474 (January): 39–56.
Wang, X., Y. Liu, K. Fan, P. Cheng, and S. Xia. 2023. Nano-Striped Polyamide Membranes Enabled by
Vacuum-Assisted Incorporation of Hierarchical Flower-like MoS2 for Enhanced Nanofiltration
Performance. Journal of Membrane Science 668 (February): 121250.
Wei, X., B. Zhao, X.-M. Li, Z. Wang, B.-Q. He, T. He, and B. Jiang. 2012. CF4 Plasma Surface Modification
of Asymmetric Hydrophilic Polyethersulfone Membranes for Direct Contact Membrane
Distillation. Journal of Membrane Science 407–408 (July): 164–175.
Wiktor, C., M. Meledina, S. Turner, O.I. Lebedev, and R.A. Fischer. 2017. Transmission Electron
Microscopy on Metal–Organic Frameworks – a Review. Journal of Materials Chemistry A 5, no.
29: 14969–14989.

135
Experimental Methods for Membrane Applications

Winter, D., J. Koschikowski, and S. Ripperger. 2012. Desalination Using Membrane Distillation: Flux
Enhancement by Feed Water Deaeration on Spiral-Wound Modules. Journal of Membrane
Science 423–424 (December): 215–224. http://linkinghub.elsevier.com/retrieve/pii/
S0376738812006175.
Winter, D., J. Koschikowski, and M. Wieghaus. 2011. Desalination Using Membrane Distillation:
Experimental Studies on Full Scale Spiral Wound Modules. Journal of Membrane Science 375,
no. 1–2: 104–112. http://dx.doi.org/10.1016/j.memsci.2011.03.030.
Wyart, Y., G. Georges, C. Deumié, C. Amra, and P. Moulin. 2008. Membrane Characterization by Microscopic
Methods: Multiscale Structure. Journal of Membrane Science 315, no. 1–2 (May): 82–92.
Xing Yang, Hui Yu, Rong Wang, A.G.F., X. Yang, H. Yu, R. Wang, and A.G. Fane. 2012. Optimization
of Microstructured Hollow Fiber Design for Membrane Distillation Applications Using CFD
Modeling. Journal of Membrane Science 421–422 (December): 258–270. http://dx.doi.
org/10.1016/j.memsci.2012.07.022.
Yang, F., J.E. Efome, D. Rana, T. Matsuura, and C. Lan. 2018. Metal–Organic Frameworks Supported on
Nanofiber for Desalination by Direct Contact Membrane Distillation. ACS Applied Materials &
Interfaces 10, no. 13 (April 4): 11251–11260.
Yang, X., R. Wang, A.G. Fane, C.Y. Tang, and I.G. Wenten. 2013. Membrane Module Design and
Dynamic Shear-Induced Techniques to Enhance Liquid Separation by Hollow Fiber Modules: A
Review. Desalination and Water Treatment 51, no. 16–18: 3604–3627. https://doi.org/10.10
80/19443994.2012.751146.
Zhang, W., Y. Li, J. Liu, B. Li, and S. Wang. 2017. Fabrication of Hierarchical Poly (Vinylidene Fluoride)
Micro/Nano-Composite Membrane with Anti-Fouling Property for Membrane Distillation.
Journal of Membrane Science 535 (August): 258–267.
Zhao, D., J. Zuo, K.-J. Lu, and T.-S. Chung. 2017. Fluorographite Modified PVDF Membranes for
Seawater Desalination via Direct Contact Membrane Distillation. Desalination 413 (July): 119–
126.
Zhao, K., W. Heinzl, M. Wenzel, S. Büttner, F. Bollen, G. Lange, S. Heinzl, and N. Sarda. 2013.
Experimental Study of the Memsys Vacuum-Multi-Effect-Membrane-Distillation (V-MEMD)
Module. Desalination 323 (August): 150–160. http://linkinghub.elsevier.com/retrieve/pii/
S0011916412006509.
Zhu, Y., H. Inada, A. Hartschuh, L. Shi, A. Della Pia, G. Costantini, A.L. Vázquez de Parga, et al. 2012.
Scanning Electron Microscopy. In Encyclopedia of Nanotechnology, 2273–2280. Dordrecht:
Springer Netherlands.
Ziel, R., A. Haus, and A. Tulke. 2008. Quantification of the Pore Size Distribution (Porosity Profiles) in
Microfiltration Membranes by SEM, TEM and Computer Image Analysis. Journal of Membrane
Science 323, no. 2 (October): 241–246.
Zou, D., and Y.M. Lee. 2022. Design Strategy of Poly(Vinylidene Fluoride) Membranes for Water
Treatment. Progress in Polymer Science 128 (May): 101535.
Zuo, J., and T.-S. Chung. 2016. Metal–Organic Framework-Functionalized Alumina Membranes for
Vacuum Membrane Distillation. Water 8, no. 12 (December 8): 586.
Zuo, J., T.-S. Chung, G.S. O’Brien, and W. Kosar. 2017. Hydrophobic/Hydrophilic PVDF/Ultem®
Dual-Layer Hollow Fiber Membranes with Enhanced Mechanical Properties for Vacuum
Membrane Distillation. Journal of Membrane Science 523 (February): 103–110.

136
 Part 2
Particulate fouling
 doi: 10.2166/9781789062977_0139

Chapter 6

Silt Density Index

Steven J. Duranceau, University of Central Florida, USA

The learning objectives of this chapter are the following:

• Define the silt density index (SDI) and explain its significance

• Present a method that can be used to characterize the particulate fouling potential of
reverse osmosis feedwater

• Understand the theory behind the SDI and discuss the basic equations that are used
to calculate fouling indices

• To learn how to perform a SDI test using the appropriate equipment and procedures

• Identify and explain the limitations and deficiencies of the SDI as a measure of
particulate fouling in synthetic membrane processes.

6.0 ABSTRACT

The most widely accepted and historically used predictor of the fouling potential of reverse
osmosis feedwater is the plugging factor (PF), now commonly known as the Silt Density
Index (SDI). The SDI procedure was standardized by ASTM International and is intended
to be used as a measure of the fouling capacity of feedwater to reverse osmosis systems. The
SDI is an index calculated from a test that measures the rate at which a 0.45-micrometer
(μm) filter is plugged when subjected to a constant water pressure of 206.8 kPa (30 psi).
The SDI gives the percent drop per minute in the flow rate of the water through the filter,
averaged over a specified time-period, typically 15 minutes. Because the SDI is a relatively
simple procedure and inexpensive to implement, it has been universally applied since the
1960s to assess the particulate fouling tendency of a feedwater intended for treatment by
reverse osmosis (RO) membrane processes. Many facilities in the United States rely on
automated SDI systems that are microprocessor controlled and fully automatic to allow
operators to regularly monitor the feedwater. Care must be taken when employing the
SDI with regards to accuracy and reproducibility, as the index is not based on any filtration

© 2024 The Authors. This is an Open Access book chapter distributed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0),
(https://creativecommons.org/licenses/by-nc-nd/4.0/). The chapter is from the book Experimental
Methods for Membrane Applications in Desalination and Water Treatment, Sergio G. Salinas-Rodriguez,
Loreen O. Villacorte (Eds).
Experimental Methods for Membrane Applications

mechanism and is not proportional with colloidal or particle concentration, in addition to

the fact that there is no correlation factor for temperature nor for variations in membrane
resistance. Even though there are legitimate concerns regarding the predictive value of the
SDI, the index continues to be successfully used in the planning of RO facilities to guide
the selection of pretreatment processes, and is often the basis of membrane guarantees and
other plant performance contracts.

6.1 DEVELOPMENT OF THE FOULING INDEX

Fouling is a major obstacle to the widespread use of membrane technology. Membrane

fouling has a direct impact on process performance because it decreases productivity
(permeate flux) over time, increases the amount of energy required to meet water demand,
and accelerates the need for membrane cleaning and replacement. Fouling is simply defined
as the accumulation of undesired materials, via deposition or adsorption of soluble and
particulate matter, onto the surface of the membrane. Membrane fouling can consist of
either non-living (inorganic and organic matter) or living organisms (bacteria) and is the
primary cause of permeate decline and loss of overall productivity in reverse osmosis and
nanofiltration processes.

Early in the development of semi-permeable membranes for water treatment, the need to
estimate membrane fouling potential of the raw water was found to be essential to identify
pretreatment requirement to prepare the feedwater prior to processing. This is important
because effective pretreatment can lower the number of required cleaning events and
extend the life of membrane elements. Attempts to correlate fouling propensity of water
with turbidity was only slightly successful.

To help solve these issues, the Du Pont de Nemours & Company (Du Pont) introduced
its first reverse osmosis permeators for water desalination in 1969 under the trade name
‘Permasep’ an outcome of the company’s research in polymer chemistry and synthetic fibers
(Hagley Library, 2022). By 1997, Du Pont had sold over 1.5 billion gallons of desalting
capacity, dominating the seawater desalination market for many decades. Since the first
hollow-fine fiber membranes were sold by DuPont it was initially believe that performance
was hampered by suspended and colloidal matter in the feedwater. Consequently, DuPont
developed the Fouling Index, which was later denoted as the Silt Density Index (Schippers
et al. 2014). Despite Permasep’s success, DuPont decided to discontinue the production
of hollow-fiber permeators in 1997, primarily attributed to the rise of the spiral-wound
membrane configuration’s success in the global desalination market (Hagley Library, 2022).

6.2 SILT AS A COMPONENT OF MEMBRANE FOULING

Assalay and colleages (1998) described silt as a solid granular material that is comprised of
suspended rock and mineral particles with a size between sand or clay that can accumulate
on the membrane surface. Although the SDI is termed as a silt index, this does not mean that
the measurement is for silt considerations alone. The SDI is a parameter used to determine
the fouling propensity of a source water intended to be processed using reverse osmosis
membranes. Sources of membrane fouling can be divided into four principal categories:

140
 Chapter 6

• Particulate (silt, inorganic colloids, oxidized iron and manganese, algae, aluminosilicates)
• Microbiological (bacteria)
• Organic fouling (natural or synthetic compounds, oils, grease)
• Scale (limiting salt chemistries)

Since reverse osmosis synthetic membrane processes were first introduced for the
treatment of water supplies, it was found that in most instances plugging of the elements
was due to blocking filtration by suspended particulate matter (Comstock, 1982). Fouling
by particulates (silt) generally impacts the lead membrane elements of any pressure vessel
process configuration unlike scale that concentrates in the flow channels of the tail-end
or last membrane elements located in a pressure vessel. Scaling is of greater concern with
more concentrated feed solutions, therefore the last modules in the process pressure vessel
configurations are most affected because they are exposed to the most concentrated feed
water. Microbiological and organic-type fouling can occur anywhere within the membrane
configuration depending on feedwater quality, pretreatment methods and process
operating conditions. Consequently, the SDI is a measurement that can determine the
fouling potential of a feedwater for particulate fouling, and may not be as predictive for
microbiological, organic or scale type conditions.

6.3 STANDARDIZATON OF THE SILT DENSITY INDEX

In 1982, ASTM International (West Conshohocken, PA), formerly known as the American
Society for Testing and Materials, developed the ‘Standard Test Method for Silt Density
Index (SDI) of Water’ designated as D4189-14 by ASTM International (2014) which was
first revised on January 30, 1987. According to ASTM International (2014), the SDI test
method can be used to ‘indicate the quantity of particulate matter in water and is applicable
to relatively low (<1.0 NTU) turbidity waters such as well water, filtered water, or clarified
effluent samples.’ The test is not applicable to RO, NF or ultrafiltration (UF) permeate. The
test essentially consists of filtering water through a 47-mm diameter cellulose-based filter
that possesses 0.45-μm at a constant pressure of 30 psi (210 kPa). The standard ASTM SDI
test does not contain any correction for testing parameters such as membrane resistance,
membrane area, feed temperature and applied pressure. The SDI increases with increasing
temperature since the water viscosity is affected; additionally, an increase in pressure and
decrease in membrane resistance will increase the measurements result. The SDI test is not
an absolute measurement of the quantity of particulate matter.

The method has essentially remained the same procedure since that time and has been
proven to be useful from an operating perspective for membrane plant operators (Ruiz-
Garcia et al. 2015). According to Harn R/O Systems, Inc. (2022), the SDI test gives a
calculated number in the range 0 – 6, where 0 is excellent and 6 denotes a very high fouling
potential. Most membrane manufacturers require the feed SDI to be below 3.0 to indicate
control of colloidal and particulate fouling. SDI values above 3.0 typically indicate periodic
cleaning will be required and values above 5.0 could indicate rapid fouling can indicate
the need to provide additional pretreatment to protect the membranes during process
operation. Although in practice a high SDI typically indicates that fouling may occur, a low
value does not guarantee that fouling will not occur.

141
Experimental Methods for Membrane Applications

6.4 METHODS AND PROCEDURES

Manual SDI testing typically requires the following components and items, portions of
which are illustrated in Figure 1 (Hydranautics, 2022) and shown in Figure 2:
• SDI test assembly made of high-quality stainless steel or plastic.
• Filter holder to withstand 50 psi (350 kPa) pressure and hold.
• 47 mm nominal, plain filter papers115 to 180 micrometer thickness rated to 0.45 μm,
typically white hydrophilic cellulose triacetate or mixed cellulose nitrate type materials.
• Pressure regulator and gauge able to measure 30 psi.
• Feedwater ball valve, plastic.
• Graduated cylinder, 500 mL capacity
• Stopwatch, graduated in hundredths of a minute.
• Thermometer, liquid-in-glass, suitable for measuring the temperature of the water
sample; capable of being read to within ±1°C.
• Dull tweezers.

Feed

Ball valve or 1st stage regulator

Pressure regulator 30 psi

Pressure gauge

Bleed
O-ring
0.45 micron filter

Figure 1 Typical assembly of the apparatus used for SDI measurements

Source: https://membranes.com/wp-content/uploads/Documents/TSB/TSB113.pdf

Several private corporations and original equipment manufacturer’s provide in-house

procedures, based on ASTM D4189-07, such as the information provided by AquaPhoenix
Scientific (2022), Hydranautics (2022) and Harn R/O Systems, Inc. (2022). Note that a
booster pump may be required for use in a manual SDI kit if insufficient pressure exists for
testing to proceed. To perform an SDI test, the following general procedure can be followed,
and is provided in more detail in ASTM International SDI testing method (2014). On-
line instruments are also available that automatically and consistently monitor the SDI of
RO feedwater that are typically controlled by a microprocessor and rely on a typical 4-20
mA interface. These devices contain side-stream or automatic flush sequences, and often
extrapolate the SDI to 75% of the filter plugging condition.

142
 Chapter 6

Figure 2 Photo depicting the components of the SDI apparatus: storage box for equipment,
500-mL graduated cylinder, Teflon tape, dull tweezers, SDI filter pads, stopwatch, cell
assembly and flexible tubing, pressure regulator and gauge, booster pump.
Source: Courtesy of Harn R/O Systems, Inc., a Division of Komline-Sanderson Corp.

Typically, most reverse osmosis and nanofiltration membrane water treatment facilities
make available a well flush valve on the raw water line upstream of the process building. The
flush valve allows the raw water to be flushed to waste at process start-up for an operational
pre-determined time period to reduce the SDI reading to below 3.0 prior to allowing water
to be transferred to the pretreatment system ahead of the membranes. An SDI sample point
can easily be installed on the raw water line upstream of the plant inlet valve so that SDI
tests can be performed during well flush events. Taking these SDI tests during plant start-up
allows operators to determine the length of time needed for individual and collective (or
intake) flushing cycles as each facility may have differing source water supply transmission
line configurations. It is recommended that an SDI test should be performed at least once a
day on the raw water when the process is in operation and the results should be recorded in
the operator’s daily log.

Step 1: Measure the time required to filter a fixed volume of water through a standard space:
0.45 μm pore size microfiltration membrane at a constant pressure of 30 psi (2.07 bar) per
the following procedure. Record this as Ti, or initial T.
a. Connect the test kit less filter paper for pretest flush.
b. Flush the test kit and supply line for 3 to 5 minutes to remove any possible contaminants.
c. Measure the temperature of the water and record the reading.
d. Make sure the O-ring on the filter is in good condition and properly placed. Set the
pressure regulator to 30 psig (210 kPa). The set screw on the regulator should be
adjusted while there is a small flow. Supply pressure to the regulator should be greater
than 40 psig (276 kPa).
e. Open the 47 mm in diameter filter holder and carefully place a 0.45 μm membrane
filter into the filter holder using the dull tweezers to avoid damage and touching the
filter paper. Screw loosely together.

143
Experimental Methods for Membrane Applications

f. Open the feed valve slightly and adjust the filter housing to overflow, displacing any
trapped air. Residual air trapped in the housing can be flushed by opening the small
‘bleed’ screw; care should be taken else this part can come loose and be easily lost.
Tighten the filter housing, open the feed valve completely and make final adjustments
to the pressure regulator as required; close the valve and tighten the filter holder the
remainder of the way without overtightening.
g. Prepare to take measurements. Open the ball valve and simultaneously, using the
stopwatch, begin measuring the time required to fill the 500 mL measuring cylinder.
Record the time (ti). Leave the valve open for continued flow; do not stop the watch or
the water flow.

Step 2: Take additional time measurements, normally after 5, 10 and 15 minutes (after silt
build up). Measure and record the times to collect additional 500 mL volumes of sample,
starting the collection at 5, 10, 15 minutes of total elapsed flow time. This value is recorded
as (tf) with f being the time used. Measure the water temperature and check the pressure
as each sample is collected. The pressure must remain constant at 30 psig (± 1 psig) and
the temperature must remain constant to 1˚C. After completion of the test, the membrane
filter may be retained for future reference or additional chemical evaluations of the filtered
deposit matter. It is recommended that the date, time, sample location, operator name, SDI
value and notes or comments be collected along with the filter pad.

Step 3: Calculate the Plugging Factor (PF) after 5, 10 and 15 minutes as determined as
shown in Equations 1, 2 and 3, respectively:

Ti
PF5 min = (1 ) ×100 Eq. 1
T5

Ti
PF10 = (1 ) ×100 Eq. 2
min
T10

Ti Eq. 3
PF15 min = (1 ) ×100
T15

Step 4: The SDI value is then determined at each interval as SDI = PF/T. Calculate the Silt
Density Index (SDI) as follows using Equation (4):

Ti
1 ×100
%P30 Tf
SDIT = = Eq. 4
T T
where SDIT is the Silt Density Index (%/min) at time T, ti is the initial time required to
collect 500 mL of sample, tf is the elapsed filtration time (min) required to collect 500 mL of
sample after a test time (typically 15 minutes after the initial measurement), and %P is the
percent at 30 psi feed pressure.

144
 Chapter 6

The ASTM method recommends that if the %P30 exceeds 75% after 5 min then other test
methods should be used to analyze for particulate matter. Considering that the %P30 is
essentially the percentage of plugging factor, Equation 4 can be rewritten as Equation 5:

T1
1 ×100
%PF T2 Eq. 5
SDIT = =
T T

SDI measures the percentage of the filtrate flow rate decline per minute and is expressed as
a percentage per minute but typically is reported without units. As an example, a SDI of 2.5
would indicate that the SDI filtrate flow was reduced by 2.5 percent per minute during the
test. This concept is illustrated graphically in Figure 3 where the filtration flow is presented
as a function of time, and V1 and V2 are the volumes of the first and second sample:

“Percentage flux decline per minute”

V1

V2=V1

Flow
t1 t2 Time (minutes)
T

Figure 3 Representation of the filtration flow as a function of time per the SDI test method.
(Adapted from Alhadidi and colleagues 2011)

Calculation Examples
1. Calculate the PF and SDI for a test where the time measurements indicated a Ti of one
minute and T15 of 4.0 minutes.
Solution: The plugging factor is calculated as PF = 1-(1/4)×100 = 75. On the other hand,
the SDI is then calculated as SDI = 75/15 = 5 as a percentage of flux decline per minute.
These results indicated that flow had decayed by a factor of four times, indicating that
75% of the 0.45-micron filter has been plugged. As the SD