University of Southern California Dissertations and Theses

Optimizing biomembrane reactor systems for water reclamation and reuse applications

(USC Thesis Other)

Optimizing biomembrane reactor systems for water reclamation and reuse applications

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

OPTIMIZING BIOMEMBRANE REACTOR SYSTEMS
FOR WATER RECLAMATION AND REUSE
APPLICATIONS

by

Woonhoe Kim

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ENVIRONMENTAL ENGINEERING)

May 2016

Copyright 2016 Woonhoe Kim
i

Dedication

To my family

ii

Acknowledgements
I remember receiving the acceptance letter from the University of Southern California
(USC), Department of Civil and Environmental Engineering, a few years ago. I was thrilled at the
prospects of getting my higher education towards a doctorate degree at a world-class university
with internationally renowned professors. At that time I was working in Korea after having
completed my Master’s degree in Environmental Engineering at the Yonsei University. After a
few years, USC and Los Angeles gave my more than I could ask for in terms of education,
knowledge, friendships and collaborations.
First and foremost, I would like to express my sincere thanks and appreciation to my
advisor, Professor Massoud Pirbazari, for his constant inspiration, insightful thoughts, strong
motivation, and endless support throughout my journey at USC. His novel ideas and unique
approach to the field of environmental science and engineering were responsible for my academic
and professional development, and changed the way I thought. He taught me not only the art of
research but also the role of an educator. Additionally, I would like to thank Dr. Varadarajan
Ravindran for generously sharing his unique ideas with me and being with me always whenever I
needed any help in my research work.
I would like to extend my gratitude to the members of my dissertation committee, Professor
Thieo Hogen-Esch and Professor Patrick Lynett for their helpful suggestions and encouragement.
Additionally, I would like to acknowledge the financial support received in the form of graduate
teaching and research assistantships from the Viterbi School of Engineering at USC, and various
federal and other agencies that provided funding.
I owe a lot to my advisor’s research collaborator, Professor Thieo Hogen-Esch, Loker
Hydrocarbon Research Institute, USC Department of Chemistry, and his former doctoral advisee,
Dr. Hariye Merve Yurdacan. The work on the synthesis and development of novel membranes
with impregnation of nanomaterials such as graphene oxide into polymer matrices would not have
been possible without their substantial and valuable contributions.
I would like to express my special thanks to Dr. Ryan Thacher, a former advisee of
Professor Pirbazari, for helping me to start my work in the laboratory as a researcher. I am also
grateful to all the undergraduate and graduate members of the Safe Water for All Nations (SWAN)
Research Group of Professor Pirbazari.
iii

I wish to thank Professor Hyung-keun Chung, my former academic advisor at Yonsei
University in the Master’s degree program, who created in me a deep interest in the area of water
quality and analysis. My special thanks are due to the late Mr. Young J. Paik (R.I.P.), former
President at Paco Steel & Engineering Corp., Mrs. Kyeong-sook (Susan) Paik, and the Board
Members of Yonsei International Foundation, who initially supported me to start my life in the
United States.
My biggest thanks are due to my father Jae-sil, my mother Hee-sun, my sister Young-shin,
my father-in-law Soo-boong, my mother-in-law Yoon-kyung, my sister-in-law Suna, and my
brother-in-law Byung-woo, and other relatives, for their endless support, ceaseless encouragement
and constant love.
Lastly, my sincere gratitude goes to my wife Younga, who has been showing her infinite
and consistent patience and love since 2002. She always makes me smile.

iv

Table of Contents
Dedication ………………………………………………...……………………………………….i
Acknowledgements ……………………………………………………………………………….ii
List of Tables ……………………………………………………………...…………..….….… viii
List of Figures ……………………………………………………………………….…….……. ix
Nomenclatures ……...………………………………………………………………...….…..… xvi
Acronyms ……………………………………………………………….………….......…….. xviii
Greek Symbols ………………………………………………………….………...……....….… xx
Abstract ………………………………………………………………….…………….………. xxi
Chapter 1 Introduction, Background, and Research Scope …………………………...…………. 1
1.1 Introduction ………………………..…………………………………………………...… 1
1.2 Background …………………………………………………………………………..…… 4
1.2.1 Membrane Processes ……………………………………………………………..…… 4
1.2.2 Membrane Surface Modification …………………………………………...……….. 10
1.2.3 Membrane Bioreactor Processes ………………………………………….…...…….. 16
1.3 Research Scope …………….……………………………………...…………………….. 17
1.4 References ……………………………………………………………………….……… 18
Chapter 2 Models for Predicting the Performance of Membranes Used in Water Reclamation
and Reuse ……………………………………………..……………..………………………… 22
2.1 Introduction …………………………………………………………………………...… 22
2.2 Membrane Transport Model ………………………………………………….…………. 23
2.3 Computational Method …………………………………………………….……………. 27
2.4 Estimation of Model Parameters ……………………………………………..…………. 29
v

2.5 Modeling Approach for the MBR process ……………………………………………….. 29
2.6 Model Assumption and Development …………………………………….…………….. 34
2.7 Model Conceptualization and Formulation ………………………………...……………. 37
2.8 Modeling under Different MBR Process Conditions ……………………………………. 44
2.9 Model Parameter Estimation …………………………………………………………….. 45
2.10 Model Sensitivity Analysis ……………………………………………..……………… 49
2.11 References ……………………………………………………………….…………….. 50
Chapter 3 Investigating the Potential of Flat Sheet Ultrafiltration Membranes for Water
Reclamation and Reuse ………………………………..……………………………………… 53
3.1 Introduction …………………………………………………………...………………… 53
3.2 Materials and Methods ……………………………………………………….………….. 55
3.3 Ozonation Batch Studies ……………………………………………………...…………. 62
3.4 Adsorption Isotherm Studies …………………………………………………………….. 63
3.5 Results and Discussion …………………………………………………………...……… 67
3.6 Summary and Conclusions ………………………………………………………...…….. 86
3.7 Limited Study of Nanofiltration Membrane for Water Reuse Application …………..….. 87
3.8 References …………………………………………………………………………….… 94
Chapter 4 Investigating the Potential of Hollow Fiber Ultrafiltration Membranes for Water
Reclamation and Reuse ……………………………………………..……….…..……………. 99
4.1 Introduction …………………………………………………………………………..…. 99
4.2 Materials and Methods …………………………………………………………………. 104
4.3 Biokinetic Studies …………………………………………………………………...…. 108
4.3.1 Biokinetic Reactor Studies …………………………………………………………. 109
vi

4.3.2 Estimation of Biokinetic Parameters …………………………………………..…… 110
4.3.3 Biodegradation ………………………………………………………………….….. 113
4.3.4 Biofilm ………………………………………………………………………….….. 115
4.4 Results and Discussion …………………………………………………………………. 116
4.4.1 Micro-Scale UF Hollow Fiber Membrane ……………………………….……..….. 116
4.4.2 Mini-Pilot-Scale UF Hollow Fiber Membrane …………………………………..… 122
4.5 Summary and Conclusions ………………………………………………………...…… 153
4.6 References ………………………………………………………………………...…… 155
Chapter 5 Developing Polymeric Nanofiltration Membrane Impregnated with Graphene
Oxide Nanoparticles …………………………………………………...………………..…… 157
5.1 Introduction …………………………………………………………………………..... 157
5.2 Research Background ………………………………………………………………….. 160
5.3 Materials and Methods ………………………………………………………………… 166
5.4 Results and Discussion …………………………………………………………………. 171
5.5 Nanofiltration Membranes for Water Reclamation …………………………………….. 182
5.6 Application of Polymers for Membrane Fabrication …………..………………………. 184
5.7 Application of Nanomaterials in Membrane Matrices …………………………………. 185
5.8 Membrane Fouling Control and Cleaning Strategies ………………………………..…. 188
5.9 Development of Novel Polymeric Nanofiltration Membranes …………………………. 190
5.9.1 Results and Discussion ……………………………………………………………... 191
5.10 Summary and Conclusions ………………………………………………………..….. 197
5.11 Future Work …………………………………………………………………………... 198
5.12 References ………………………………………………………………………...….. 199
vii

Chapter 6 Summary, Conclusions, and Recommendation ………………………………..…… 211
6.1 Summary and Conclusions ………………………………………………………..……. 211
6.1.1 General ………………………………………………………………………..……. 211
6.1.2 Flat Sheet UF Membranes and Modeling ………………………………………….. 211
6.1.3 Hollow Fiber UF Membrane Bioreactor Experiments and Modeling …………..…. 213
6.1.4 Synthesizing New Membranes ………………………………………………..…… 213
6.2 Recommendation ………………………………………………………………………. 214
Bibliography ………………………………………………………………………………...… 216

viii

List of Tables
Table 1.1 - Effects of cleaning process on membrane fouling …………………………………. 10
Table 2.1 - Summary of estimated model parameters for plate-and-frame system ….….…….... 30
Table 2.2 - Summary of estimated model parameters for MBR system ……………….……….. 31
Table 2.3 - Mathematical equations of mass flux and reactions associated phenomenological
mass transfer mechanisms …………………………………………..………………………. 37
Table 3.1 - Wastewater characteristics at San Jose Creek Water Reclamation Plant (Aug.
2011 - Oct. 2014) and experimental permeate quality of UF membrane filtration ………….. 58
Table 3.2 - Adsorption isotherm parameters for Freundlich and Langmuir isotherms …….…... 65
Table 3.3 - Reuse of wastewater applications on a basis of different advanced wastewater
treatment …………………………………………………………………………………….. 92
Table 4.1 - Biokinetic parameters for secondary clarifier effluent …………………………….. 113
Table 5.1 - Composition of synthesized membranes and commercial nanofiltration
membrane …………………………………….…………………………....…………….… 168
Table 5.2 - Performance comparison of various membranes …………………………………. 172
Table 5.3 - Feed and permeate water quality for NF90 membrane filtration experiments …….. 196

ix

List of Figures
Figure 1.1 - Possible resistances against solvent transport: (a) permeate flow, (b) gel layer
formation, (c) internal pore fouling, and (d) surface fouling (source: ocw.tudelft.nl) .………. 8
Figure 1.2 - Filtration spectrum (source: ocw.tudelft.nl) ……..…………....………………….…. 8
Figure 2.1 - Schematic of solute transport through the membrane filtration system ………....… 23
Figure 2.2 - Schematic of concentration profile in the proposed model for ultrafiltration
membranes …………………………………………………………………………………... 24
Figure 2.3 - Flow chart for model parameter estimation and numerical algorithm for membrane
transport model simulation ………………………………………………...………………... 28
Figure 2.4 - Protocol for the modeling and design of the MBR for water reclamation ……..…. 35
Figure 2.5 - Schematic diagram of the bioflim on activated carbon …………………………… 37
Figure 2.6 - Schematics of (a) hollow fiber UF membrane, (b) solute transport through the
hollow fiber membrane, and (c) concentration profile …………………...………………….. 39
Figure 3.1 - Wastewater treatment plant at San Jose Creek Water Reclamation Plant in Los
Angeles County (adapted from San Jose Creek Water Reclamation Plant) …………...…..… 57
Figure 3.2 - Schematic of flat sheet cross-flow membrane filtration unit …………….……...… 60
Figure 3.3 - Schematic of experimental setup for ozonation system …………………..…….…. 63
Figure 3.4 - Determination of equilibration time with 100 mg L
-1
of PAC in SCE ……...……... 65
Figure 3.5 - Determination of adsorption patterns by fitting experimental data to (a) Freundlich
and (b) Langmuir isotherms ….………………………………………….…………………... 66
Figure 3.6 - Relationship between permeate flux and trans-membrane pressure using UF flat
sheet membrane using distilled-deionized water ……………………...…………………..….67
Figure 3.7 - Effect of trans-membrane pressure on permeate flux for SCE using UF flat sheet
membrane ……………………………………………….………………..…………………. 68
Figure 3.8 - Effect of trans-membrane pressure on TOC concentration of SCE using UF flat
sheet membrane (TOC concentration in feed reservoir = 7.2 mg L
-1
) ……………..……….. 69
Figure 3.9 - Effect of PAC addition on permeate flux of SCE using UF flat sheet membrane .... 70
Figure 3.10 - Effect of PAC addition on TOC concentration of SCE using UF flat sheet
membrane (TOC concentration in feed reservoir = 7.2 mg L
-1
) …………………………...... 70
Figure 3.11 - Effect of E.coli addition on permeate flux of SCE using UF flat sheet membrane . 71

x

Figure 3.12 - Effect of E.coli on TOC concentration of SCE using UF flat sheet membrane
(TOC concentration in feed reservoir = 9.5 mg L
-1
) ………………………….……………... 72
Figure 3.13 - Effect of E.coli and PAC addition on permeate flux of SCE using UF flat sheet
membrane ………………………………………………………………………….………....73
Figure 3.14 - Effect of E.coli and PAC addition on TOC concentration of SCE using UF flat
sheet membrane (TOC concentration in feed reservoir = 9.5 mg L
-1
) …………..…………... 73
Figure 3.15 - Effect of ozone (5 mg L
-1
) and a combination of ozone (5 mg L
-1
) and hydrogen
peroxide (5 mg L
-1
) on permeate flux for UF flat sheet membrane (Temperature = 25
o
C,
and time duration = 1 hour) ………….……….……….……….……….……….………….... 75
Figure 3.16 - Effect of ozone (5 mg L
-1
) and a combination of ozone (5 mg L
-1
) and hydrogen
peroxide (5 mg L
-1
) on TOC concentration for UF flat sheet membrane (Temperature = 25
o
C,
and time duration = 1 hour) ………………………………………………………………….. 75
Figure 3.17 - Flux recovery and filtration after membrane cleaning processes for SCE with UF
flat sheet membrane (TMP = 30 psi): (a) 1 × 10
-3
M of sodium hydroxide; (b) backwash
with DDI water and cleaning with Triton X-100 and RID-X ……………………………….. 77
Figure 3.18 - TOC concentration after membrane cleaning processes for UF flat sheet
membrane: (a) 1 × 10
-3
M of sodium hydroxide at 8 and 16 hours; (b) backwash with DDI
water at 6 hours, cleaning with 5 mg L
-1
of Triton X-100 at 12 hours, and 5 mg L
-1
of
RID-X at 18 hours …………….………….………….………….………….………………... 78
Figure 3.19 - Permeate flux pattern at different trans-membrane pressure for UF flat sheet
membrane filtration (feed TOC concentration = 7.2 mg L
-1
) …………………………..…..... 80
Figure 3.20 - TOC concentration profiles at different trans-membrane pressure for UF flat
sheet membrane filtration (feed TOC concentration = 7.2 mg L
-1
) ……………..…………... 81
Figure 3.21 - Effect of cross-flow rate on permeate flux for UF flat sheet membrane filtration
(feed TOC concentration = 7.2 mg L
-1
) …………………………………..……………….…. 82
Figure 3.22 - Effect of cross-flow rate on TOC concentration for UF flat sheet membrane
filtration (feed TOC concentration = 7.2 mg L
-1
) …………………….……………………... 83
Figure 3.23 - Effect of PAC addition on permeate flux for UF flat sheet membrane filtration ... 84
Figure 3.24 - Effect of PAC addition on TOC concentration for UF flat sheet membrane
filtration (trans-membrane pressure = 30 psi, and feed TOC concentration = 7.2 mg L
-1
) ….. 84

xi

Figure 3.25 - Experimental data and model profiles for model sensitivity analysis for flat sheet
UF membrane: (a) Effect of diffusion coefficient; (b) effect of gel layer …..……………….. 85
Figure 3.26 - Permeate flux pattern for NF flat sheet membranes ……………………………... 88
Figure 3.27 - TOC concentration for NF flat sheet membranes …………………….……..….... 88
Figure 3.28 - Effect of 15 mg L
-1
PAC addition on NF permeate flux pattern …..…….…….… 90
Figure 3.29 - Effect of 15 mg L
-1
PAC addition on NF permeate TOC removal ..…..…….…… 90
Figure 3.30 - Actual and proposed water reclamation schemes ………………....……..….…… 93
Figure 4.1 - Schematic of micro-scale hollow fiber membrane filtration unit ………….…….. 107
Figure 4.2 - Schematic of mini-pilot hollow fiber membrane filtration …………….……….... 107
Figure 4.3 - Schematic of experimental setup for biokinetic studies …………………..…….... 110
Figure 4.4 - Overall E.coli growth curve in secondary clarifier effluent at 25
o
C …………….... 112
Figure 4.5 - TOC reduction by E.coli growth at 25
o
C ………………………………..……..… 113
Figure 4.6 - Scanning electron microscope (SEM) showing the growth of E.coli (a) on the UF
membrane surface at 50,000 times magnification, and (b) on the activated carbon particle
surface at 10,000 times magnification …………………………………………………...…. 116
Figure 4.7 - Effect of trans-membrane pressure on permeate flux for SCE using micro-scale
UF hollow fiber membrane ………………………………………………………..……..… 117
Figure 4.8 - Effect of trans-membrane pressure on TOC concentration of SCE using micro-
scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
) …… 117
Figure 4.9 - Effect of PAC addition on permeate flux of SCE using micro-scale UF hollow
fiber membrane ………………………………………...………………………………...… 118
Figure 4.10 - Effect of PAC addition on TOC concentration of SCE using micro-scale UF
hollow fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
) …………….… 119
Figure 4.11 - Effect of E.coli addition on permeate flux of SCE using micro-scale UF hollow
fiber membrane ………………………………...………………………………………..…. 120
Figure 4.12 - Effect of E.coli on TOC concentration of SCE using micro-scale UF hollow
fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
) ……………..……….. 120
Figure 4.13 - Effect of E.coli and PAC addition on permeate flux of SCE using micro-scale
UF hollow fiber membrane ………………………………………..……………………….. 121
Figure 4.14 - Effect of E.coli and PAC addition on TOC concentration of SCE using micro-
scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
) …….. 122
xii

Figure 4.15 - Relationship between permeate flux and trans-membrane pressure using mini-
pilot-scale UF hollow fiber membrane using distilled deionized water …………………….. 124
Figure 4.16 - Effect of trans-membrane pressure on permeate flux for SCE using mini-pilot-
scale UF hollow fiber membrane …………………………….……………………………. 124
Figure 4.17 - Effect of trans-membrane pressure on TOC concentration of SCE using mini-
pilot-scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
) . 125
Figure 4.18 - Effect of hydraulic retention time on permeate flux for SCE using mini-pilot-
scale UF hollow fiber membrane …………..……………...……………………………..… 125
Figure 4.19 - Effect of hydraulic retention time on TOC concentration of SCE using mini-
pilot-scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
) . 126
Figure 4.20 - Effect of PAC addition on permeate flux of SCE using mini-pilot-scale UF
hollow fiber membrane …………………………………………………………………….. 127
Figure 4.21 - Effect of PAC addition on TOC concentration of SCE using mini-pilot-scale
UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
) …..……… 127
Figure 4.22 - Effect of E.coli addition on permeate flux of SCE using mini-pilot-scale UF
hollow fiber membrane …………………………………………………………………….. 128
Figure 4.23 - Effect of E.coli on TOC concentration of SCE using mini-pilot-scale UF hollow
fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
) ……………..……..… 129
Figure 4.24 - Effect of E.coli and PAC addition on permeate flux of SCE using mini-pilot-
scale UF hollow fiber membrane ………..……………………………………...………..… 130
Figure 4.25 - Effect of E.coli and PAC addition on TOC concentration of SCE using mini-
pilot-scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
) . 130
Figure 4.26 - Effect of ozonated SCE on permeate flux using mini-pilot-scale UF hollow
fiber membrane ….……………………………………...………………………………….. 131
Figure 4.27 - Effect of ozonated SCE on TOC concentration using mini-pilot-scale UF hollow
fiber membrane: TOC concentrations for 10% TOC reduction, 20%, and 30% were 5.3, 4.8,
and 4.2 mg L
-1
, respectively …………...………………………………………………...…. 132
Figure 4.28 - Effect of “relaxation” on permeate flux recovery of SCE using mini-pilot-scale
UF hollow fiber membrane: (a) SCE at 4 psi of TMP; (b) SCE with 40 mg L
-1
of PAC
addition at 4 psi of TMP …………………………………………………..………………... 134

xiii

Figure 4.29 - Effect of “relaxation” on TOC concentration using mini-pilot-scale UF hollow
fiber membrane: (a) SCE at 4 psi of TMP; (b) SCE with 40 mg L
-1
of PAC addition at 4 psi
of TMP (TOC concentration in feed reservoir = 6.5 mg L
-1
) …………………………….…. 135
Figure 4.30 - Flux recovery and filtration after membrane cleaning processes for SCE with
mini-pilot-scale UF hollow fiber membrane (TMP = 4 psi): Backwash with DDI water at
6 hours, cleaning with 5 mg L
-1
of RID-X at 12 hours, cleaning with 50 mg L
-1
of RID-X at
18 hours, “relaxation” at 24 hours, cleaning with 1 × 10
-3
M of sodium hydroxide at 30
hours, cleaning with 1000 mg L
-1
of RID-X at 36 hours ……………….……………….….. 136
Figure 4.31 - TOC concentration after membrane cleaning processes for mini-pilot-scale UF
hollow fiber membrane: Backwash with DDI water at 6 hours, cleaning with 5 mg L
-1
of
RID-X at 12 hours, cleaning with 50 mg L
-1
of RID-X at 18 hours, “relaxation” at 24 hours,
cleaning with 1 × 10
-3
M of sodium hydroxide at 30 hours, cleaning with 1000 mg L
-1
of
RID-X at 36 hours (TOC concentration in feed reservoir = 6.5 mg L
-1
) …...………….…… 136
Figure 4.32 - Experimental data and theoretical predictions of the TOC concentration profiles
in the mini-pilot MBR reactor in the absence of PAC adsorbent and presence of 10
8
CFU
per 100 mL E.coli ….……………….……………….……………….…………………...… 139
Figure 4.33 - Experimental data and theoretical predictions of the TOC concentration profiles
in the MBR using 40 mg L
-1
PAC adsorbent and no microorganisms ..……………….…… 140
Figure 4.34 - Experimental data and theoretical predictions of the TOC concentration profiles
in the MBR using 100 mg L
-1
PAC adsorbent and no microorganisms …..….……………... 140
Figure 4.35 - Experimental data and theoretical predictions of the TOC concentration profiles
in the MBR reactor system using 40mg L
-1
PAC adsorbent and 10
8
CFU per 100 mL

Ecoli
………….……….……….……….……….……….……….……….……….……….…….. 141
Figure 4.36 - Sensitivity of MBR TOC removal to (a) the reactor flow rate Q and (b) the
influent TOC concentration C0 …………………………………...……………………..….. 144
Figure 4.37 - Sensitivity of MBR TOC removal to (a) the Freundlich capacity coefficient KF ;
(b) the Freundlich intensity coefficient 1/n; (c) the surface diffusion coefficient Ds; and (d)
the film transfer coefficient kf ………………...…………………………………………….. 146
Figure 4.38 - Sensitivity of MBR TOC removal to (a) the maximum substrate utilization rate µ max; (b)
the half saturation coefficient K s; and (c) the microbial yield coefficient Y ………..……………..…. 148

xiv

Figure 4.39 - Sensitivity of MBR TOC removal to (a) the microorganism concentration X; (b)
the maximum biofilm thickness Lfmax; and (c) the biofilm diffusion coefficient Df ….…….. 150
Figure 4.40 - Permeate flux pattern at different trans-membrane pressure for UF hollow fiber
membrane filtration (feed TOC concentration = 6.5 mg L
-1
) ………………..……………… 152
Figure 4.41 - Effect of PAC addition on permeate flux for UF hollow fiber membrane filtration
(feed TOC concentration = 6.5 mg L
-1
) …………..………………………………………... 153
Figure 5.1 - Synthesis of polyamide copolymers for membrane fabrication …………………. 164
Figure 5.2 - Partially sulfonated polyamides for membrane applications ……………………… 164
Figure 5.3 - Synthesis of graphene oxide-modified polyamides (adopted from Yurdacan, 2015)
…………………..………………………………………………………………………..… 165
Figure 5.4 - Chemical modification of graphene oxide for infusion into polymeric matrices …. 166
Figure 5.5 - Schematic of the configuration of synthesized membrane filtration ……………… 169
Figure 5.6 - Comparison of permeate flux for different membranes: (a) Membranes A, B, and
C; (b) Membranes D and E ……………………………………..…………………………... 174
Figure 5.7 - Comparison of TOC removal for different membranes: (a) Membranes A, B, and
C; (b) Membranes D and E ………………………………..……………………………...… 175
Figure 5.8 - Effect of PAC on permeate flux for Membranes A and C ……..………………… 176
Figure 5.9 - Effect of PAC on TOC concentration for Membranes A and C …….…………… 176
Figure 5.10 - Flux recovery and filtration after membrane cleaning processes for Membranes
A, C, D, and E ……………………………….....………………...………………………… 180
Figure 5.11 - Sulfonated membranes annealed at 165
o
C: (a) Membrane D, 30% PSSA; (b)
Membrane E, 35% PSSA ……………………...…………………………………………… 181
Figure 5.12 - TOC concentration after membrane cleaning processes for Membranes A, C, D,
and E ….………………………………………………………….……………………….... 182
Figure 5.13 - Permeate flux pattern for NF flat sheet membranes (same as Figure 3.26) ….…. 192
Figure 5.14 - TOC concentration for NF flat sheet membranes (same as Figure 3.27) ………. 193
Figure 5.15 - Effect of 15 mg L
-1
PAC addition on NF permeate flux pattern (same as Figure
3.28) ……………………………..…………………………………………………………. 194
Figure 5.16 - Effect of 15 mg L
-1
PAC addition on NF permeate TOC removal (same as Figure
3.29) ……………………………………………………………..…………………...…..… 195

xv

Figure 5.17 - Flux recovery and filtration after membrane cleaning processes for UF permeate
water with NF flat sheet membrane: 1 × 10
-3
M of sodium hydroxide at 12 hours ……..….. 197

xvi

Nomenclatures
1/n Freundlich intensity coefficient or exponent constant
Bw adsorbent particle weight difference
C0 influent TOC concentrion
Cb concentration in bulk region
Ce solute concentration at equilibrium
C f concentration in feed
Cg concentration of gel layer
CL concentration at the membrane surface on bulk side
Cpac concentration of powdered activated carbon
Cper concentration of permeate
CP concentration at the membrane surface on permeate side
dp diameter of adsorbent particle
D diffusion coefficient
Df diffusion coefficient in biofilm
Dl free liquid diffusivity
Dr dilution rate
Ds surface diffusion in activated carbon
DT diameter of inner pipe
f Fanning friction factor
J permeate flux
k mass transfer coefficient
k ’ maximum substrate utilization rate per unit mass of microorganism
kd specific death rate
kf substrate mass transfer coefficient
km Monod specific substrate utilization rate
ks partition coefficient of the solute in the membrane
Kf liquid film transfer coefficient
KF Freundlich isotherm constant
KL Langmuir isotherm constant
Ks Monod half saturation constant
xvii

Lf biofilm thickness
q solid-phase concentration of adsorbed substrate
qe solute mass adsorbed per unit adsorbent mass at equilibrium
qfs solute concentration at the biofilm-adsorbent interface
qmax maximum adsorption capacity
Q flow rate
rf radical coordinate in biofilm
rg net rate of biomass growth
rs radical coordinate in particle
rsu substrate utilization rate
R radius of particle
Re Reynolds number
S concentration of substrate in a reactor
S0 concentration of substrate in influent
Sb substrate concentration in bulk liquid
Sf substrate concentration in biofilm
Sfs substrate concentration at the biofilm-adsorbent interface
Slf substrate concentration at the liquid-biofilm interface
Sin substrate concentration in feed solution
Sp substrate concentration in intrapore void space
Sc Schmidt number
Sh Sherwood number
t time
Uw average fluid velocity
v cross-flow velocity
V volume of a reactor
We weight of evaporated water
Wp dry weight of retained particle
X concentration of biomass in a reactor
X0 concentration of biomass in influent
Xf biomass density in biofilm
xviii

Xs suspended biomass concentration
Xw particle weight
y b distance from the membrane surface toward bulk region
y m distance from the membrane surface toward permeate region
Y cell yield coefficient

Acronyms
AOP advanced oxidation process
ATCC American Type Culture Collection
CA cellulose acetate
CE cellulose esters
CMB completely mixed batch
CN nitrocellulose
CNT carbon nano tube
CSA camphor sulfonic acid
DDI distilled-deionized
DMF N,N-Dimethylformamide
DOC dissolved organic carbon
E.coli Escherichia coli
EDC endocrine disrupting chemical
ESP extracellular polymeric substances
GO graphene oxide
HAA haloacetic acid
HRT hydraulic retention time
MBR membrane bioreactor
MD molecular dynamics
MF microfiltration
MPD m-phenylenediamine
MWCO molecular weight cutoff
NF nanofiltration
xix

NPC-TF nanomaterial-polymer composite thin-film
NPOC non-purgeable organic carbon
PA polyamide
PAC powdered activated carbon
PAN polyacrylonitrile
P(BASS-S-CMS) poly (tetrabutyl ammonium styrene sulfonate-co-styrene-co-4-
chloromethyl styrene)
PE polyethylene
PEG polyethylene glycol
PEGM polyethylene glycol methacrylate
PES polyethersulfone
POSS polyhedral oligomeric silsesquioxanes
PP polypropylene
PPCP pharmaceuticals and personal care product
PS polysulfone
PSSA polystyrene sulfonic acid
PTFE polytetra fluoro ethylene
PVC polyvinyl chloride
PVDF polyvinylidene fluoride
RO reverse osmosis
SCE secondary clarifier effluent
SJCWRP San Jose Creek Water Reclamation Plant
TEA triethanol amine
TFC thin-film composite
TMC trimesoyl chloride
TMP trans-membrane pressure
TOC total organic carbon
UF ultrafiltration
UV ultraviolet
ZNC zeolite nanocrystal

xx

Greek Symbols
α compactness of gel layer
β ratio of particle surface area to particle volume
ΔP trans-membrane pressure
δcp concentration polarization layer
δg gel layer thickness
δm membrane thickness
ε experimental correlation
θ hydraulic retention time
θp intrapore void space in particle
κ resistance
κ ac resistance of the activated carbon
κ cp resistance of the concentration polarization
κ g resistance of the gel layer
κ ir resistance of the irreversible fouling
κ in resistance of the internal pore fouling
κ m resistance of the membrane
κ r resistance of the reversible fouling
μ dynamic viscosity
μmax maximum specific growth rate
ν kinetic viscosity
ρ density
ρp activated carbon particle density
ρw water density

xxi

Abstract
Membrane separations promise to yield substantial environmental and economic benefits
leading to enhanced global competitiveness by significantly reducing energy consumption,
increasing industrial productivity, lowering waste generation, and addressing global water
shortage problems. Membrane technologies face several scientific and technological challenges
that must be overcome before witnessing widespread use in environmental, industrial and
commercial applications. Environmental applications include wastewater treatment, water
reclamation and reuse, water treatment, water purification, and water desalination. The specific
challenges include membrane fouling and permeate flux decline, poor rejection or selectivity, and
large energy footprint. The research presented here was intended to address most of these critical
issues.
An important aspect of this study was the examination of water reclamation processes using
combination of ultrafiltration membranes and activated carbon adsorption, and to develop novel,
high-performance nanofiltration membranes superior to existing commercial membranes. Firstly,
flat-sheet ultrafiltration membranes were tested to evaluate the fundamental performance criteria
including permeate fluxes and total organic carbon (TOC) removals in water reclamation
applications using real wastewaters. Secondly, ultrafiltration membranes were employed in
continuous flow hybrid membrane bioreactor (MBR) systems for treating wastewaters after
secondary treatment. Thirdly, the novel membranes were synthesized using impregnation of
graphene oxide (GO) nanoparticles in polymeric matrices. These membranes were intended to be
superior to existing commercial membranes for water reclamation and reuse as well as other uses
regarding various criteria: aqueous transport and permeability properties, anti-fouling potential
and fouling resistance, rejection and separation characteristics, cleanability and flux recovery,
xxii

chemical tolerance, mechanical strength, and overall durability. These membranes were tested in
batch systems to evaluate their feasibility for the above applications. The finished water quality
was intended to meet the necessary treatment standards for water reclamation and reuse
applications regarding chemical and biological purity. Different types of cleaning agents such as
caustic solution (NaOH), surfactant (Triton X-100), and biological enzyme (RID-X) were
evaluated for foulants removal and permeate flux recovery. Lastly, modeling approaches were
employed to predict permeate fluxes and TOC removals for MBR systems in various
configurations.
The study further included laboratory-scale flat-sheet plate-and-frame membrane filtration
tests for investigating the permeate flux patterns and the TOC removals with secondary clarifier
effluent obtained from Los Angeles County. As the TOC removals were not satisfactory with the
ultrafiltration membrane itself, additional processes were required such as powder activated carbon
(PAC) adsorption, microbial degradation (microorganisms including E.coli), and oxidation (ozone,
and peroxone). In all these case, the permeate fluxes were also evaluated. Bench-scale studies were
conducted to determine parameters used for prediction of permeate flux and TOC concentration
using mathematical models.
A micro-scale hollow fiber ultrafiltration membrane bench setup was designed and tested
to evaluate membrane performances regarding permeate fluxes and TOC removals. The micro-
scale tests were intended to provide a guideline for the design of hollow fiber membrane modules
used in a mini-pilot-scale system. The mini-pilot-scale system represented a continuous flow
hybrid MBR process using a hollow-fiber ultrafiltration membrane module, and it was used to
assess the feasibility of water reclamation applications using permeate fluxes and TOC removals
as criteria. The uniqueness of the MBR unit was the special design for controlling membrane
xxiii

fouling and permeate flux decline, combining powdered activated carbon (PAC) sorption and fluid
management techniques. The membrane module was operated in “outside-in” dead-end fluid-
dynamic regime, and equipped with structural features to promote local vortex and turbulence for
fouling control. The mini pilot-scale MBR studies evaluated permeate flux decline patterns,
membrane fouling, and organic rejection as TOC and UV254 (ultraviolet absorbance at 254 nm
wavelength). The feed and effluent streams were also analyzed to a limited extent for biochemical
oxygen demand (BOD), chemical oxygen demand (COD), biomass, and other relevant water
quality parameters.
A transport model using the resistances of various layers constituting concentration
polarization and gel layer besides adsorbent and biofilm layer was employed for predicting the
permeate fluxes and TOC removals for flat-sheet and hollow-fiber membrane configurations. The
model considered membrane surface fouling, internal pore fouling, and membrane rejection in its
formulation as well. The necessary model parameters were obtained from bench-scale membrane
filtration tests.
This investigation involved a feasibility study of the laboratory-scale MBR process for the
purification of potable water sources. Reliable predictions of process performance were obtained
regarding organic removal efficiency reflected by the effluent concentration profiles as functions
of time. Such predictions were made on the basis of easily determined laboratory experiments,
pilot-plant scale studies would be minimized and considerable savings in cost and time can be
achieved. This objective was achieved by developing and employing a mathematical modeling
approach. Simultaneously, a modeling protocol was observed for the process design and upscaling
using dimensional analysis and similitude (although it was not the focus of this research).
xxiv

The model for performance prediction of organic removals (TOC removals) in the mini-
pilot-scale MBR system incorporated the following phenomenological aspects including
adsorption mass transfer resistance, adsorption equilibrium, biological reaction due to suspended
microorganisms in bulk solution, and biological reaction within biofilms. In this regard adsorption
equilibrium studies was performed to evaluate the adsorption equilibria for dissolved organic
matter (DOM) using total organic carbon (TOC) as a surrogate parameter. Adsorption rate studies
were conducted in batch reactors for determining the adsorption kinetics and the associated mass-
transfer parameters. Batch biokinetic studies were undertaken for the estimation of biological
parameters pertaining to TOC removal using an indigenous population of microorganisms.
Laboratory scale MBR experiments were performed to evaluate organic removals (TOC removals)
and membrane permeate fluxes as functions of operating time for a variety of process conditions.
The experimental determination of TOC removal efficiencies in MBR systems provided the
necessary feedback for MBR model verification and refinement. This adsorption and
biodegradation model provided excellent predictions of TOC removals in MBR systems under a
variety of operating scenarios including the following: (i) using PAC adsorbent alone (PAC at 40
mg L
-1
); (ii) using microorganisms alone (E.coli at 10
8
CFU per 100 mL), and (iii) combination;
and (iii) using PAC adsorbent with microorganisms (PAC at 40 mg L
-1
and E.coli at 10
8
CFU per
100 mL. The TOC removals were the highest for the combination of PAC and E.coli demonstrating
the synergistic effects of adsorption with microbial degradation for organic removal. The model
was employed under these conditions by suppressing the effects of adsorption or biodegradation
wherever necessary. The model predictions were in good agreement with the experimental results
for all the three scenarios. Model sensitivity analyses was performed with respect to various reactor
flow, biological and adsorption parameters for the following reasons: (i) to obtain a priori estimates
xxv

of the accuracy required for the determination of each parameter, and (ii) to predict/simulate the
behavioral patterns of process dynamics under a variety of process and operating conditions.
Limited work was directed at development and production of novel high-performance
membranes for water reclamation and other applications. This research in an ongoing collaborative
effort between the groups of Professor Massoud Pirbazari and Professor Thieo Hogen-Esch with
their expertise in the areas of membrane processes and polymer science, respectively. The initial
work included the development of polymer synthesis protocols with appropriate reaction schemes,
free-radical processes, syntheses conditions such as reaction times, curing procedures, and
quantitatively controlled incorporation of graphene oxide (GO) into the polymers. Superior
membranes were fabricated by adjusting these conditions. The membranes used in the series of
preliminary tests were prepared by interfacial polymerization by sequential addition of MPD and
TMC on a commercial polyether sulfone (PES) ultrafiltration membrane base with a nominal pore
size of 4 – 10 nm and molecular weight cutoff off (MWCO) of 10,000 Daltons. The monomers
used in the preparation of polyamide membrane were m-phenylene diamine (MPD) and 1,3,5-
benzene tricarbonyl chloride or trimesoyl chloride (TMC). Another set of membranes were cast
using these monomers MPD and TMC, but with the addition of camphor sulfonic acid (CSA) and
triethanol amine (TEA) to make the membranes material more solvophilic in nature, and to observe
their hydrophilicity, aqueous transport and rejection characteristics. These novel membranes were
synthesized by the impregnation of GO nanoparticles into polymeric matrices at different
concentrations and various conditions. The fundamental idea was to enhance the aqueous transport,
fouling resistance, rejection characteristics, chemical cleanability, and mechanical durability of the
membranes to make them far superior to existing commercially available membranes such as the
nanofiltration NF90 membranes in performance levels. The preliminary results were promising in
xxvi

so far as manifesting increased aqueous permeability and permeate fluxes by the impregnation of
GO nanoparticles. The future work would involve polymer and nanomaterial combinations to
introduce fine levels of tunability of membrane characteristics including pore sizes, pore-size
distributions, charge effects, rejection characteristics and fouling resistance besides chemical
tolerance. The novel membranes would possibly be superior to existing commercial membranes
for nanofiltration applications in the realm of wastewater treatment, water reclamation and reuse.
It is strongly believed that these concepts could be extrapolated to produce superior microfiltration,
ultrafiltration and reverse osmosis membranes as well.

1

Chapter 1
Introduction, Background, and Research Scope
1.1 Introduction
Water scarcity is ironical terminology in a watery planet yet it is true. Water is essential
natural resources on earth and it covers approximately 70% of the surface of the planet. It delivers
life, and life evolves in the water. The amount of water on earth has never changed, but simply
changed its location and form. U.S. Geological Survey (USGS) estimated the global water
distribution to be, 96.5% of the water (1.34 billion km
3
) is in oceans, 1.7% in glaciers, 1.7% in
groundwater, the rest of the water is in surface and atmosphere, and about 400 million gallons (1.5
million m
3
) per day of water was withdrawn for use in the United States in 2005. U.S.
Environmental Protection Agency (USEPA) reported that American residents generally use about
100 gallons of water per day, while Europeans use about half of the water that Americans use, and
residents of sub-Saharan African use only 2-5 gallons of water per day. From 1950 to 1980, trends
in water withdraw in the United States have significantly increased yet have been fairly steady
since 1980 (Kenny et al., 2009). The United States, however, is facing with water shortage over
the Nation due to population growth, more food consumption, more industrialization, droughts,
etc.
Water security is one of the important political and environmental issues in the world. UN-
Water defines water security as following (UN-Water, 2013):
“Water security is defined as the capacity of a population to safeguard sustainable access
to adequate quantities of acceptable quality water for sustaining livelihoods, human well-
being, and socio-economic development, for ensuring protection against water-borne
2

pollution and water-related disasters, and for preserving ecosystems in a climate of peace
and political stability. ”
The solution to water security is not just the availability of water. Individuals have reliable access
to enough quantity and safe water quality. As well as, water should be sustainable with aquatic
system when discharging it to nature. There are several methods to resolve them: using high-
efficiency toilet and appliances, charging those who overuse or waste water more money, treating
wastewater with adequate processes, and recycling highly treated wastewater for households,
industries, and agriculture rather than returning it to nature (Convention on Biological Diversity,
2013).
Water reclamation is a process that treats municipal wastewater with advanced wastewater
treatment processes in order to recycle water resources for direct or indirect potable reuse. Direct
potable reuse can be defined as introducing directly treated wastewater into a water supply system,
while indirect potable reuse can be referred to reusing treated wastewater for beneficial purposes
such as agricultural irrigation, industrial cooling water, residential landscaping and toilet flushing,
recreational water, and replenishing a groundwater basin (California Department of Public Health,
2014). Currently, there are several innovative methods to acquire safe recycling water such as
advanced oxidation processes, integrated membrane filtration systems, and sustainable reverse
osmosis operations. These technologies already exist, but need to be further developed, vastly
improved, and made economically sustainable. In order to improve the sustainability of facilities
to meet the regulation, innovative processes should be effectively implemented in the conventional
wastewater treatment.
The purpose of membrane filtration for recycled water is an alternative method of
sustainability and water conservation. Reclaimed water can decrease diversion of freshwater from
3

sensitive ecosystems, irrigation with recycled water can reduce the need for chemical fertilizers,
production of reuse water can reduce the amount of imported water and can help replenish the
groundwater resources by recharging. Reclaimed water can help us meet the domestic, industrial
and environmental water demands (The Royal Academy of Engineering, 2010).
One of the important goals of water reclamation treatment is to improve the efficiency of
the membrane technologies and membrane bioreactor (MBR) processes utilizing an integrated
membrane processes for water reclamation and groundwater recharge applications across the
nation. The purpose of membrane filtration for recycled water is an alternative method of
sustainability and water conservation. Membrane filtration has two major important functions such
as rejection and disinfection.
This research aims to investigate an integrated membrane bioreactor processes for water
reclamation using ultrafiltration (UF) membranes. Additionally, the effect of powdered activated
carbon (PAC) and microorganisms on permeability and total organic compound (TOC) removal
will be investigated by semi-batch and mini-pilot-scale experiments.
This dissertation is structured to facilitate the understanding of physicochemical
mechanisms and the feasibility of integrated membrane filtration processes. Chapter 1 provides
the general introduction, the background of membrane filtration processes, and the scope of
research. Chapter 2 discusses mathematical equations of predictive models for membrane
filtrations. Chapter 3 through 5 are stand-alone chapters and discuss experimental results. Chapter
3 presents the flat sheet UF membrane was experimental data for permeate flux and TOC removal
for different test conditions such as presence of microorganisms, PAC adsorbent , and combination
of those. Chapter 4 discusses the results of a novel micro-scale UF hollow fiber membrane system
as well as those of a mini-pilot-scale system. Micro-scale experiments were conducted to study
4

parameters that provide information for scale-up. Mini-pilot-scale UF membrane filtration system
evaluates the viability of integrated membrane bioreactor system. Chapter 5 investigates the
development of the novel polymeric membranes with impregnated nanoparticles. Chapter 6
presents the summary of research and recommendation for further studies.

1.2 Background
1.2.1 Membrane Processes
In order to achieve the high water quality for recycled water, wastewater should be treated
using the most suitable technologies in terms of purification and disinfection. Membrane
technologies are currently in the limelight to satisfy the high quality standards (Drioli et al., 2004).
The fundamental understanding of membrane process includes the knowledge of characteristics of
membrane surface, mechanisms for retaining, mechanisms for fouling, membrane cleaning process,
and driving force for transport due to operation by pressure differential (MHW, 2005; Ho et al.,
1992). In addition, membranes are usually classified into two categories based on their pore sizes
and filtration mechanisms: microfiltration (MF) and ultrafiltration (UF); and nanofiltration (NF)
and reverse osmosis (RO). MF and UF remove suspended solids including colloidal particles and
microorganisms by a sieve mechanism and a partial pore diffusion mechanism, while NF and RO
remove dissolved substances including dissolved organic carbons (DOCs) and ions by a reverse
osmotic pressure and adiffusion mechanisms.
In general, membrane processes can remove microorganisms including bacteria, viruses,
and pathogens. Typical sizes of Giardia Lamblia and Cryptosporidium Parvum are 5-12 µm and
4-7 µm, respectively. The removal of these protozoa by UF membrane is 6 log units (99.9999%).
For bacteria such as E.coli and Pseudomonas, the removal by UF membrane is 5 log units
5

(99.999%), and the general sizes are 0.5-2 µm and 0.5-1.5 µm, respectively. Viruses such as
Enterovirus (0.02 µm) and MS2 (0.025 µm) will be removed by UF membrane up to 4 log units.
Several membrane characteristics are important in determining membrane suitability for
separation applications including porosity, morphology, surface properties, mechanical strength
and chemical resistance. These characteristics depend on the membrane material. The most
important membrane properties are the membrane productivity, flux, and rejections (Ho et al.,
1992). Commercial ultrafiltration membrane processes are widely used in water and wastewater
treatment due to the reliable uniform water quality and stable high productivity.
Membrane fouling, however, is the most negative issue when operating membrane
processes because it results in flux decline and low efficiency. Therefore, membrane fouling
should be overcome with sufficient knowledge of foulant types and characteristics of membrane
surfaces.

Membrane Permeate Flux
The evaluation of the performance of membrane is the comparison to permeate flux. The
flow of water through microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse
osmosis (RO) membranes follows the fundamental law for flow through porous media known as
Darcy’s law:
𝑣 = 𝑘 𝑃 ℎ
𝐿 𝐿
(1)
where 𝑣 = superficial fluid velocity, m/s
𝑘 𝑃 = hydraulic permeability coefficient, m/s
ℎ
𝐿 = head loss across porous media, m
𝐿 = thickness of porous media, m
6

The hydraulic permeability coefficient in Darcy’s law is an empirical parameter used to
describe the proportionality between head loss and fluid velocity and is dependent on media
characteristics such as porosity and specific surface area. Flow is expressed in terms of volumetric
flux 𝐽 rather than superficial velocity, the driving force is expressed as transmembrane pressure
(TMP) ∆𝑃 rather than head loss, and media characteristics are expressed as a resistance coefficient,
𝑅 𝑀 . In addition, the membrane flow equation includes the fluid viscosity explicitly because
viscosity has a significant impact on flux. Finally, the membrane flux equation incorporates the
membrane thickness into the resistance coefficient. The equation for membrane flux is as follows:
𝐽 =
∆𝑃 𝜇 𝑅 𝑀 (2)
where 𝐽 = volumetric water flux through membrane, L/(m
2
· h)
∆𝑃 = differential pressure across membrane, bar
𝜇 = dynamic viscosity of water, kg/(m· s)
𝑅 𝑀 = membrane resistance coefficient, m
-1

Membrane Fouling
Fouling is a major problem encountered in membrane filtration processes for water or
wastewater treatment applications. Membranes themselves represent a relevant and significant
fraction of capital costs, so that everything that can reduce the operational lifespan of membranes
would directly affect the process economics. Furthermore, membrane cleaning affects the process
owing to stoppage in operation, production of chemicals that require disposal, and impact on
membrane properties as well as lifetimes. Numerous factors have been identified as promoters of
permeate flux decline including biomass, colloids, natural organic matter (NOM), inorganic
precipitates or scalants, and extracellular polymers resulting in increase of membrane resistance,
7

pore blocking, adsorption in the pores, gel layer formation, etc. The relative importance of each
factor shall be dependent upon the operating conditions. Generally these foulants block the
membrane pores or cover the membrane surfaces due to cake or gel formation, and hence causing
permeate flux decline (Figure 1.1). However, membrane fouling is a more complicated process
and therefore its mechanisms should be studied for each specific application.
A decrease in the permeate flux or increase in trans-membrane pressure (TMP) during a
membrane process is generally understood by the term “fouling.” Fouling occurs as a consequence
of interactions between the membrane and the mixed liquor, and is one of the principal limitations
of the membrane bioreactor (MBR) process. Fouling of membranes in MBR systems is a very
complex phenomenon with diverse relationships among its causes, and it is very difficult to
localize and define membrane fouling clearly.
Membrane fouling can be characterized under several categories; irreversible or reversible,
external or internal, particulate, organic, inorganic, and biological fouling. Irreversible fouling
refers to permeate flux loss that the performance of membrane cannot be recovered by
backwashing or chemical cleaning, while reversible fouling refers to the flux losses that can be
recovered by backwashing or chemical cleaning. External fouling refers to foulants blocking pores
or covering membrane surfaces, while internal fouling refers to membrane pores being coated by
adsorption of small foulants. The terms particulate, organic, inorganic, and biological fouling refer
to fouling caused by different foulants. The fouling is attributed to cake or a gel layer formation
on the membrane surfaces or adsorption of foulants within the pores.
The mechanisms of pathogenic microorganism rejection by membrane processes are still
ambiguous. Escherichia coli (E.coli) can be used as an indicator for pathogenic microbes (Kamiko
et al., 1993). UF is effective in complete removal of bacteria. The filtration spectrum shown in
8

Figure 1.2 provides brief idea of rejection ranges of various constituents in different membrane
processes.

Figure 1.1 - Possible resistances against solvent transport: (a) permeate flow, (b) gel layer
formation, (c) internal pore fouling, and (d) surface fouling (source: ocw.tudelft.nl)

Figure 1.2 - Filtration spectrum (source: ocw.tudelft.nl)
(a) (b)
(c) (d)
9

Membrane Cleaning
Membrane cleaning is a necessary operation to maintain the performance such as
permeability and selectivity, that is, membrane permeate flux recovery. The recovery of membrane
flux can be represented by Eq. (3) (Lindau and Jönsson, 1994; Kennedy et al., 1998):
Recovery ( %) =
Flux after cleaned
Flux of initial
× 100
(3)

The frequency of membrane cleaning and selection of membrane cleaning agents are important
depending on the type and degree of membrane fouling. More importantly, membrane cleaning
should be performed without causing any damage to the membrane material and the pore structure.
There are several cleaning methods including back flush, air scouring, forward flush, chemical
flush, and chemical soaking. Cleaning procedures typically take several hours, involving
circulation of cleaning solutions such as acids, bases, and surfactants. Caustic solutions at high pH
are generally effective for removing organic foulants, while acidic solutions at low pH are used
for removing inorganic foulants (MWH, 2005; Song et al., 2004; Ang et al., 2006).
Some fouling materials can be removed by physical processes such as backwashing or
relaxation (stopping the permeate flow and continuing to scour the membrane with air bubbles).
The term “backwashing” refers to reversing flow across the membrane to detach the deposit on the
membrane surface. Generally, backwashing is performed using the permeate or deionized water
to reduce fouling due to concentration polarization and cake formation. In practice, backwashing
is set to occur automatically at timed intervals ranging from 30 to 90 minutes when the TMP
typically rises by 20% (0.2 to 1 psi) due to membrane fouling (MWH, 2005). Chlorinated
backwash water is effective in controlling microbial and biological fouling, and is referred to as
enhanced chemical backwash (Li et al., 2008).
10

Some membrane fouling caused by organic or inorganic foulants can be cleaned by acidic
and caustic solutions. Chemical cleaning is a more effective method for removing strongly
adsorbed deposits. Table 1.1 shows proper selection of chemical cleaning agents (Li et al., 2008).
Cleaning is performed by soaking the membrane in the cleaning solution or by adding the cleaning
agent into the back-flush water. Most full-scale MBR systems employ chemical maintenance
cleaning on a weekly basis, and permeate flux recovery cleaning when filtration is no longer
sustainable. Deposits that cannot be removed by available methods of cleaning are called
“irreversible fouling”. This fouling builds up over years of operation and eventually determines
the membrane life-time.

Table 1.1 - Effects of cleaning process on membrane fouling
Type of Foulants
Operating Cleaning Process
Backwashing Acidic Solution Caustic Solution
Particulate Very effective Not effective Not effective
Organic Not effective Not effective Very effective
Inorganic Not effective Very effective Not effective
Biological Effective Effective Effective

1.2.2 Membrane Surface Modification
Membranes can be classified into polymeric or organic membranes, and ceramic or
inorganic membranes in terms of their material characteristics. Organic membranes are referred to
polymer membranes such as cellulose acetate (CA), polyamide (PA), polysulfone (PS),
polyethersulfone (PES), polyvinylidene fluoride (PVDF), polypropylene (PP), and others.
Polymeric organic membranes are relatively inexpensive and easily manufactured in a wide range
11

of pore sizes, although they have certain limitations such as pH, temperature, pressure, or chlorine
tolerance (Cui et al., 2010). On the other hand, inorganic membranes have pore size distributions
in MF, UF, and NF but not in RO because of the technical difficulties involved in synthesizing
fine pore sizes (Yu, 2006). Inorganic membranes have advantages such as high mechanical
strength, chemical and thermal stability, and extended lifetime over conventional polymeric
membranes (Cui et al., 2010). Hydrophobic polymers such as PES or PVDF require material
modification for achieving higher fluxes and less fouling. Various methods of altering the surface
chemistries are as follows (Ho et al., 1992; Geise et al., 2010):
1. Reacting base polymer with hydrophilic pendent groups and then casting the membrane
2. Surface grafting of hydrophilic species on a previously made membrane
3. Blending of polymers

Nanomaterials in Polymeric Membranes
Application of nanoparticles in the manufacturing process of polymeric membranes has
received much attention during the last few years with reference to their ability to improve and
increase aqueous permeate transport, and produce desired structure and functionalities.
Nanoparticles-based membranes can be developed by assembling engineered nanoparticles into
porous membranes or blending them with polymeric or inorganic membranes. Nano-sized
inorganic-material-blended composite membranes are attractive candidates owing to their
enhanced properties such as perm-selectivity, hydrophilicity, and fouling resistance. In these
nanocomposite materials, the concentration of the filler can be very high without any loss of
physical properties of a polymeric membrane. These membranes are denoted as ‘mixed matrix’
membranes, wherein both the phases manifest a positive mutual influence.
12

In polymeric membranes, carbon nanotubes (CNTs) have been used as additives for the
fabrication of carbon nanotubes mixed matrix membranes. Among different nanomaterials, CNTs
have received increasing attention in the scientific community due to their unique structural and
extraordinary physical, chemical as well as biological properties. A remarkable property of CNTs,
which makes them attractive for transport applications is their unique combination of significantly
high aspect ratios with small dimensions. The second important property of CNTs critical for
transport applications is the remarkable atomic scale smoothness and chemical inertness of their
graphitic walls. With diameters in the nanometer range and atomically smooth surfaces, CNTs
offer a unique system for applications of molecular transport. A study by Holt et al (2006)
demonstrated that the membranes also transported water across CNT channels at rates that cannot
be accounted for by continuum flow models. Membrane permeability provides a figure of merit
for membrane performance for practical applications. Despite having an order of magnitude
smaller pore size, the enhanced flow rate per pore and the higher pore density render the sub–2-
nm membranes superior to conventional membranes in both air and water permeability. Choi et
al (2006, 2007) studied the use of multi-walled carbon nanotubes (MWCNTs) incorporated in
polyethersulfone ultrafiltration membranes for performance improvement. These researchers
prepared these membranes with different MWCNT content ranging from 2% to 4% using a
conventional phase inversion process, and used them for water-treatment applications. They
further observed that the permeate fluxes were higher and that organic fouling was reduced by the
addition of the MWCNTs to a polyethersulfone matrix.
Other nanoparticles commonly used in membrane applications include oxides such as
silicon dioxide (SiO2) and silicates such as zeolites. Several researchers have fabricated
membranes with metal oxide nanoparticles besides carbon nanotubes to increase the novelty of
13

membrane materials, permeability and fouling-resistance as well as permeate quality. Owing to
excellent catalytic properties of some metal oxide nanoparticles (mainly TiO2), combining
chemical oxidation with nanoparticles-based membranes can mitigate membrane fouling and can
provide a built-in oxidative functionality. Permeate quality should also be improved due to
decomposition of organic compounds on the catalytic membrane surface. Inactivation of bacteria
and viruses with carbon-based nanomaterials has also been demonstrated. Although much effort
has been made to develop low-fouling or functional membranes using various nanoparticles,
further research is still needed to better understand design and operation of nanoparticles-based
membranes. The development of low-fouling membranes that would cover a wide range of
potential foulants fabricated with functional nanoparticles is indeed important.

Reverse Osmosis Membranes Using Thin Film Composites
Thin-film composite (TFC) polyamide RO membranes in spiral wound configuration are
commonly used in seawater desalting, brackish water treatment, and water reclamation on a
worldwide basis because of their superior solute-rejecting properties and attractive water
permeation rates. These TFC membranes generally consist of three layers, and are prepared in two
stages. Conventional phase separation and interfacial polymerization techniques are usually
combined to give an ultra-thin separation barrier deposited over a fabric-supported macro-porous
membrane in spiral-wound configuration. It must be noted that preparation of composite
polyamide membranes in commercial dimensions requires control of chemical, mechanical and
environmental factors to ensure uniformity, reproducibility, and performance at desired levels.
Indeed, the specification of raw materials, choice of solvents/reactive chemicals and their purity
requirements, and ambient humidity during membrane preparation require careful monitoring and
14

control. It is reported that the superiority in properties of these membranes will be greatly enhanced
by the inclusion of a combination of nanomaterials. The concept is based on the formation of
nanomaterial-polymer composite thin-film (NPC-TF) membrane using an interfacial
polymerization. The new type of NPC-TF RO membranes will dramatically improve permeability
and interfacial properties when compared to similarly formed pure polymer TFC membranes.
These thin film nanocomposite materials represent a breakthrough in the design of RO membranes.
In general, the basic choice of polymer material for the reverse osmosis membrane is
always polyamide. The use of zeolites-carbon nanotubes combination can be used with the
polyamide membrane fabrication process in certain proportions ranging from 2 – 4% of each with
polyamide in the membrane production scheme (Lind et al., 2009). It is reported that modified
zeolites will enhance the aqueous permeate flux through the membranes (Lind et al., 2009).
Membranes that have CNTs in the core material could be used in desalination and demineralization.
Salt removal from water, commonly performed through RO, uses less permeable membranes,
requires large trans-membrane pressure, and is quite expensive. One of the main problems with
RO desalination is the high energy demand and cost associated with pumping and forcing water
through the membranes at high pressure. These energy (primarily electricity) costs account for
perhaps 44% of the cost of RO desalination. Holt and coworkers (2006) observed that the more
permeable nanotube membranes could reduce the energy costs of desalination by up to 75%
compared to conventional membranes used in RO applications. The general notion is that the super
smooth inside of the nanotubes allow liquids and gases to rapidly flow through, while the tiny pore
size can block larger molecules, and by filtering out larger molecules the membrane would achieve
the water purification objective.
The main concept is that the introduction of nanomaterials and polymers would lead to the
15

development of superior membranes with increased aqueous permeability and greater resistance
to organic, biological and inorganic fouling. Such membranes would greatly reduce the energy
demands for various membrane applications including seawater desalination, water reclamation,
and brackish water treatment.

Anti-Fouling Coating
La and coworkers (2012) have developed novel anti-fouling coatings comprised of a
hydrophilic monomer, polyethylene glycol methacrylate (PEGM), and a multi-functional cross-
linker, methacryl-polyhedral oligomeric silsesquioxanes (POSS). The PEG moiety affords a
fouling-resistant, inert surface to prevent the deposition of organic compounds and bio-
contaminants during water purification. The POSS molecules are comprised of ultraviolet-
radiation-curable functional groups attached to the molecular vertices (8–12 depending on their
cage structures) of a rigid silica core and are used here as a cross-linking agent for the mono-
functional PEGM, which, if polymerized by itself, would lead to a linear, water-soluble polymer.
When cross-linked by itself or other organic materials, POSS molecules form well-dispersed
nanocomposites with precisely size-controlled hard blocks within soft organic matrix. La et al.
(2012) demonstrated a new approach to make controlled molecular scale-pores, as characterized
by molecular weight cutoff (MWCO) experiments and dye filtration, within POSS–PEGM
hydrogels. The addition of a water-soluble additive, such a non-cross-linkable polyethylene glycol
(PEG), into a pre-polymerization solution containing POSS and PEGM, and the subsequent
extraction of the additive from the UV-cured films generate molecular scale pores (or channels)
within the polymer films, resulting in significant increase in water permeability. In order to find
the optimum nanocomposite hydrogel composition for efficient antifouling activity, free-standing
16

hydrogel films were prepared by varying the weight ratio between the methacryl-POSS and PEGM,
and the amount of sacrificial additive PEG (La et al., 2012). The intrinsic properties of the POSS–
PEGM films, including water uptake, water permeability, and MWCO, were thoroughly
characterized, and the antifouling efficiency of the nano-porous POSS–PEGM-coated UF
membranes was evaluated by cross-flow filtration of a synthetic oil–water emulsion and bovine
serum albumin (BSA) solution. Molecular dye filtration was performed with a composite
membrane coated with a POSS–PEGM film to evaluate the membrane’s molecular separation
capability.

1.2.3 Membrane Bioreactor Processes
In the realm of wastewater treatment, water treatment and water reclamation, the
membrane bioreactor (MBR) process employing microfiltration and/or ultrafiltration membranes
has found wide application. The technology has been successfully employed to conventional
activated sludge treatment with microbial growth (Ravindran et al., 2009).
The membrane bioreactor technology was employed by Fane and coworkers (1980) for
wastewater treatment applications in Australia. These researchers combined activated sludge
process with an ultrafiltration unit and were able to obtain an effluent of high quality by membrane
exclusion of suspended solids including silt, clay and microorganisms. Chang et al. (1993)
employed a similar technology using a bioreactor and an ultrafiltration module for the
denitrification of drinking water. Cicek et al. (1998) employed a pilot-scale version in the
reclamation of municipal wastewater for non-potable uses. Nowadays, MBRs also employ
submerged membrane configuration for applications such as municipal wastewater treatment
(Rosenberger et al., 2002; Witzig et al., 2002). The use of submerged membranes has certain
17

advantages including a significant reduction in power consumption, a factor that has increased the
potential for membranes in wastewater treatment. However, submerged membrane configuration
also has certain disadvantages that offset the advantages of energy saving, and they include severe
membrane fouling, frequent membrane backwashing and cleaning requirements, and susceptibility
to material damage. The relative merits and demerits of submerged membrane reactors have been
discussed elsewhere (Rosenberger et al., 2002; Witzig et al., 2002).
Some organic matter may lead to surface fouling or internal pore fouling, particularly in
UF membranes. These organics can be simply removed by caustic cleaning agents. Therefore, the
MBR systems are ideal for reducing organic fouling through biological activity (Tsai et al., 2004).

1.3 Research scope
The scope of this research and its associated tasks are summarized as follows:
 Conduct membrane filtration experiments using a plate-and-frame membrane cell and a
membrane bioreactor with hollow fiber membranes to evaluate membrane performances
such as permeability and TOC removal under different operating conditions using
biologically treated wastewater.
 Use mathematical models to predict the permeate flux pattern and TOC concentrations by
implementing the concepts of pore diffusion and gel layer formation, and compare the
results with experimental works.
 Evaluate the effect of activated carbon on the membrane permeability and TOC removal,
and compare the results with the model prediction.
 Determine the effect of biofilm on the membrane permeate flux pattern and TOC rejection,
and compare the results with predicted model.
18

 Determine the biokinetic parameters from the batch biokinetic studies.
 Perform the effect of combination of activated carbon and microorganisms on permeate
flux and TOC removal.
 Investigate the membrane permeate flux and TOC removal for the novel membranes
synthesized in this research. Furthermore, determine the effects of addition of activated
carbon and microorganisms on membrane performances.

1.4 References
Ang, W. S., Lee, S., & Menachem Elimelech (2006). Chemical and physical aspects of cleaning
of organic-fouled reverse osmosis membranes, Journal of Membrane Science, 272, 198-210.

California Department of Public Health (2014). Title 17 and 22 California Code of Regulations.
Retrieved from http://www.waterboards.ca.gov/drinking_water/certlic/drinkingwater/documents
/lawbook/RW regulations_20140618.pdf.

Chang, J., Manem, J., & Beaubien, A. (1993). Membrane bioprocesses for the denitrification of
drinking water supplies. Journal of Membrane Science, 80(1-3), 233-239.

Choi, J. H., Jegal, J., & Kim, W. N. (2006). Fabrication and characterization of multi-walled carbon
nanotubes/polymer blend membranes. Journal of Membrane Science, 284 (1-2), 406-415.

Choi, J. H., Jegal, J., & Kim, W. N. (2007). Modification of performances of various membranes
using MWNTs as modifier. Macromolecular Symposia, Advanced Polymers for Emerging
Technologies, 249-250, 610-617 (Special topic issue).

Cicek, N., Franco, J. P., Suidan, M. T., & Urbain, V. (1998). Using a membrane bioreactor to
reclaim wastewater. Journal of American Water Works Association, 90(11), 105-113.

19

Convention on Biological Diversity (2013). Water and Biodiversity - Natural Solutions for Water
Security. Montreal.

Cui, Z. F. & Muralidhara, H. S., Membrane technology: A practical guide to membrane technology
and applications in food and bioprocessing, Elsevier Ltd., 2010.

Drioli, E. & Fontananova, E. (2004). Membrane technology and sustainable growth, Chemical
Engineering Research and Design, 82(A12): 1557-1562.

Fane, A. G., Fell, C. J. D., & Nor, M. T. (1980). Ultrafiltration/activated sludge system:
development of a predictive model. In Ultrafiltration Membranes and Applications, Cooper, A.R.
(editor), pp. 631-658, Plenum Press, New York.

Geise, G. M., Lee, H. S., Miller, D. J., Freeman, B. D., McGrath, J. E., & Paul, D. R. (2010). Water
purification by membranes: the role of polymer science, Journal of Polymer Science: Part B:
Polymer Physics, 48, 1685-1718.

Ho, W. S. W. & Sirkar, K. K. Membrane Handbook, Van Nostrand Reinhold, 1992.

Holt, J. K., Park, H. G., Wang, Y., Stradermann, M., Artyukhin, A. B., Grigoropoulos, C. P., Noy,
A., & Bakajin, O. (2006). Fast mass transport through sub–2-nanometer carbon nanotubes.
Science, 312, 1034-1037.

Kamiko, N. & Ohgaki, S. (1993). Multiplication characteristics of FRNA phage and its utility as
an indicator for pathogenic viruses, Water Science and Technology, 27(3-4), 133-136.

Kenndy, M., Kim, S., Mutenyo, I., Broens, L., & Schippers, J. (1998). Intermittent crossflushing
of hollow fiber ultrafiltration systems, Desalination, 118(1-3), 175-187.

Kenny, J. F., Barber, N. L., Hutson, S. S., Linsey, k. S., Lovelace, J. K., & Maupin, M. A. (2009).
Estimated Use of Water in the United States in 2005, U.S. Geological Survey Circular: 1344.
20

La, Y. H., Sooriyakumaran, R., McCloskey, B. D., Allen, R. D., Freeman, B. D., & Al-Rasheed,
R. (2012). Enhancing water permeability of fouling-resistant POSS-PEGM hydrogels using
‘addition-extraction’ of sacrificial additives. Journal of Membrane Science, 401-402, 306-312.

Li, N. N., Fane, A. G., Winston Ho, W. S., & Matsuura, T., Advanced membrane technology and
applications, Wiley, 2008.

Lind, M. L., Ghosh, A. K., Jawor, A., Huang, X., Hou, W., Yang, Y., & Hoek, E. M. V. (2009).
Influence of crystal size on zeolites-polyamide thin film nanocomposite membranes. Langmuir,
25(17), 10139-10145.

Lindau, J. & Jönsson, A. S. (1994). Cleaning of ultrafiltraion membranes after treatment of oily
wastewater, Journal of Membrane Science, 87(1-2), 71-78.

MWH, Water Treatment: Principles and Design, 2nd ed., John Wiley and Sons, 2005.

Ravindran, V., Tsai, H. H., Williams, M. D., & Pirbazari, M. (2009). Hybrid membrane bioreactor
technology for small water treatment utilities: process evaluation and primordial considerations,
Journal of Membrane Science, 344(1-2), 39-54.

Rosenberger, S., Krüger, U., Witzig, R., Manz, W., Szewzyk, U., & Kraume, M. (2002).
Performance of a bioreactor with submerged membranes for aerobic treatment of municipal waste
water. Water Research, 36, 413-420.

Song, W., Ravindran, V., Koel, B. E., & Pirbazari, M (2004). Nanofiltration of natural organic
matter with H2O2/UV pretreatment: fouling mitigation and membrane surface characterization,
Journal of Membrane Science, 241, 143-160.

The Royal Academy of Engineering, Global Water Security – an engineering perspective, London,
United Kingdom, April 2010.
21

Tsai, H. H., Ravindran, V., Williams, M. D., & Pirbazari, M. (2004). Forecasting the performance
of membrane bioreactor process for groundwater denitrification. Journal of Environmental
Engineering and Science, 3(6), 507-521.

UN-Water (2013). Water security and the global water agenda, A UN-Water Analytical Brief.
United Nations University.

Water treatment: Micro- and ultrafiltration. (2016, March 8), Retrieved from https://ocw.tudelft.nl/
wp-content/uploads/Micro-and-ultrafiltration-1.pdf

Witzig, R., Manz, W., Rosenberger, S., Krüger, U., Kraume, M. & Szewzyk, U. (2002).
Microbiological aspects of a bioreactor with submerged membranes for aerobic treatment of
municipal wastewater. Water Research, 36, 394-402

Yu, D. (2006). Inorganic mesoporous membranes for water purification applications: synthesis,
testing and modeling, (Doctoral Dissertation), The Ohio State University.

22

Chapter 2
Models for Predicting the Performance of Membranes Used in
Water Reclamation and Reuse
2.1 Introduction
A mathematical model is a practical, reliable, and useful tool or package for the
performance prediction, analysis, design, scale-up, and optimization of processes under a variety
of design and operating conditions. More importantly such models encompass process conditions
beyond normal operations to predict the dynamics in situations that include deviations from
specifications, operational turndowns and shutdowns. Membrane transport models are generally
based on the fundamental principles of process engineering and fluid mechanics involving the
concepts of equilibria, surface chemistry and transport phenomena. A close approach to full-scale
system is not always possible with a laboratory-scale system, and so a pilot-scale system may be
required. A reliable model is generally used to minimize the effort, time, and cost associated with
pilot-scale studies. The most significant parameters such as those related to temperature, pH,
biological kinetics, and mass-transfer, and most of these factors can be obtained from laboratory-
scale experiments, and some parameters may be estimated and verified by correlation techniques.
In general, the systematic modeling protocol provides an excellent approach for upscaling or
downscaling of reactor system designs in various applications including water treatment,
wastewater treatment, and water reclamation and reuse.

23

2.2 Membrane Transport Model
Membrane transport and permeate flux models are based on pore diffusion transport
phenomena between the membrane internal pore and the membrane surface wall. Schematics of
solute transport membrane filtration system are presented in Figures 2.1 and 2.2.

Figure 2.1 - Schematic of solute transport through the membrane filtration system

24

Figure 2.2 - Schematic of concentration profile in the proposed model for ultrafiltration
membranes

In the membrane transport model, pore diffusion, internal pore fouling, concentration
polarization, gel layer formation, and surface fouling are considered. The mathematical
representation of solute (aqueous) transport through the membrane system is illustrated in Figure
2.1. The region between the membrane and the cell wall as well as within the membrane pores are
used for formulating the solute constitutive relationships within the contact volume of interest,
represented in a certain coordinate system.
The model assumptions can be briefly summarized as follows:
 Pore diffusion and pore sorption of solute are important mechanisms
 Concentration polarization and surface fouling are considered significant
 Micro-pores in the membrane region in the pore diffusion domain are cylindrical in
shapes
 Pore lengths significantly exceed the pore radii
Bulk solution (C
b
)
Concentration
polarization layer
(C
cp
, δ
cp
)
Gel layer (C
g
, δ
g
)
Permeate
concentration (C
per
)
Membrane surface
(Boundary layer)
Distance (y)
Concentration (C) Bulk region
Membrane thickness (δ
m
)
Permeate region
Concentration on
boundary (C
L
)
Concentration on
boundary (C
P
)
Distance (y)
25

 Gel layer formation arises due to the saturation of solute layer on the membrane surface
The membrane transport model is described by the species conservation concept based on
the advection-diffusion phenomena tailored to a solute (foulant) and solvent (water) that can be
represented by the following equations (Eq. 1 and 2).
∂𝐶 𝐿 ∂𝑡 − 𝐽 ∂𝐶 𝐿 ∂𝑦 𝑏 = 𝐷 ∂
2
𝐶 𝐿 ∂𝑦 𝑏 2
Eq. (1)

∂𝐶 𝑃 ∂𝑡 + 𝐽 ∂𝐶 𝑃 ∂𝑦 𝑚 = 𝐷 ∂
2
𝐶 𝑃 ∂𝑦 𝑚 2
Eq. (2)

At the boundary of membrane surface, the solute concentrations on both sides should be
adjusted by partition coefficient, ks because the concentration on the permeate side, CP is much
lower than the concentration on the bulk side, CL. Eq. (3) shows the ratio of solute concentrations,
wherein the partition coefficient (determined by membrane isotherm tests) represents the
equilibrium between the solute and the membrane surface.
𝑘 𝑠 =
𝐶 𝑃 𝐶 𝐿 Eq. (3)
Both Eqs. (1) and (2) can be written as follows under various conditions - prior to and after
the formation of the saturated gel layer.
Prior to the formation of the gel layer, the solute concentration on the membrane surface,
CL is less than the gel layer concentration, Cg, that is CL < Cg and δ g = 0.
At t = 0, C = CL,0 = Cb,0; at yb = δ cp, C = Cb; and at yb = 0, 𝐶 𝐿 𝐽 ( 1 − 𝑘 𝑠 ) = −𝐷 (
∂𝐶 ∂𝑦 )
𝑦 =0

At t = 0, CP = Cper,0 = ks Cb,0; at y m = δ m, CP = Cper; and at y m = 0, CP = ksCb
After the formation of the saturated gel layer; the solute concentration on the membrane
surface CL is the same as Cg in the gel layer, so that CL = Cg.
At y b = δ g + δ cp, C = Cb, and for 0 ≤ y b ≤ δ g, C = Cg
26

At y m = δ m, CP = Cper, and at y m = 0, CP = ksCg
Under steady-state conditions, Eq. (1) can be written in the form of Eq. (4). As the gel layer
formation occurs Eq. (4) can be transformed to Eq. (5), wherein equilibrium is established for the
solute between the gel layer and the membrane surface:
𝐽 =
𝐷 𝛿 𝑐𝑝
ln
𝐶 𝐿 − 𝐶 𝑃 𝐶 𝑏 − 𝐶 𝑃 = 𝑘 ln
𝐶 𝐿 − 𝑘 𝑠 𝐶 𝐿 𝐶 𝑏 − 𝑘 𝑠 𝐶 𝐿 Eq. (4)

𝐽 = 𝑘 ln
𝐶 𝐿 − 𝑘 𝑠 𝐶 𝐿 𝐶 𝑏 − 𝑘 𝑠 𝐶 𝐿 = 𝑘 ln
𝐶 𝑔 − 𝑘 𝑠 𝐶 𝑔 𝐶 𝑏 − 𝑘 𝑠 𝐶 𝑔 Eq. (5)

The permeate flux can be described by Darcy’s law establishing the relationship between
trans-membrane pressure and the sum of the transport resistances in series, Eq. (6). The equation
can be written in differential form as Eq. (7). In this relationship, the intrinsic resistance of the
membrane, κ m is attributed to the membrane material and the surface roughness. The resistance
due to concentration polarization, κ cp is expressed in terms of the trans-membrane pressure, fluid
velocity, and solute concentrations. Internal pore fouling also offers transport resistance due to
solute sorption and pore blocking. These two resistances can be combined and represented by Eqs.
(8) and (9), and the total resistance is time dependent. The resistance due to the gel layer, κ g can
be denoted by Eq. (10) and is also time dependent.
𝐽 =
𝛥 𝑃 𝜇 ( 𝜅 𝑚 + 𝜅 𝑟 + 𝜅 𝑖𝑟
)
=
𝛥 𝑃 𝜇 ( 𝜅 𝑚 + 𝜅 𝑔 + 𝜅 𝑐𝑝
+ 𝜅 𝑖𝑛
)

Eq. (6)

𝛥 𝑃 d𝐽 d𝑡 + 𝜇 𝐽 2
(
d𝜅 𝑔 d𝑡 +
d𝜅 𝑐𝑝
d𝑡 +
d𝜅 𝑖𝑛
d𝑡 ) = 0 Eq. (7)

𝜅 𝑐𝑝 +𝑖𝑛
= 𝜅 𝑐𝑝
+ 𝜅 𝑖𝑛
= 𝑎𝑣
𝑏 𝛥 𝑃 𝑐 𝐶 𝑏 𝑑 𝐶 𝑝𝑒𝑟 𝑒 𝐶 𝑓 𝑓 Eq. (8)

27

d𝜅 𝑐𝑝 +𝑖𝑛
d𝑡 = 𝑎𝑑 𝑣 𝑏 𝛥 𝑃 𝑐 𝐶 𝑏 𝑑 −1
𝐶 𝑝𝑒𝑟 𝑒 𝐶 𝑓 𝑓 d𝐶 𝑏 d𝑡 + 𝑎𝑒 𝑣 𝑏 𝛥 𝑃 𝑐 𝐶 𝑏 𝑑 𝐶 𝑝𝑒𝑟 𝑒 −1
𝐶 𝑓 𝑓 d𝐶 𝑝𝑒𝑟 d𝑡
Eq. (9)

d𝜅 𝑔 d𝑡 = 𝛼 ( 𝐶 𝑏 − 𝑘 𝑠 𝐶 𝑔 ) 𝐽 − 𝛼 𝑘 ( 𝐶 𝑏 − 𝑘 𝑠 𝐶 𝑔 ) ln
𝐶 𝑔 − 𝑘 𝑠 𝐶 𝑔 𝐶 𝑏 − 𝑘 𝑠 𝐶 𝑔 Eq. (10)

2.3 Computational Method
The membrane transport and flux model are composed of two differential equations (Eq. 9
and 10) for the resistances and other equations for flux, and trans-membrane pressure (Eq. 6, 7,
and 8). The approximations for the differential coefficients of all the independent variables with
respect to time and space are represented by an implicit finite difference scheme that is inherently
convergent. The computational scheme was tested and checked for the following criteria:
numerical consistency, conservative property, transportive property, computational stability,
computational accuracy, and convergence conditions. A MATLAB R2015b computer program
was used for solving the model equations. The general flow chart for the model simulation protocol
is illustrated in Figure 2.3.
28

Figure 2.3 - Flow chart for model parameter estimation and numerical algorithm for membrane
transport model simulation
29

2.4 Estimation of Model Parameters
Tables 2.1 and 2.2 show the estimated model input parameters used in membrane filtration
tests in the plate-and-frame system and MBR system, respectively. Some of the parameters
including the solute diffusivity D and mass transfer coefficient k were estimated by correlation
techniques. The Stokes-Einstein equation (Eq. 11) was employed to estimate the diffusion
coefficient D.
𝐷 =
𝑘 ̅
𝑇 6𝜋𝜇𝑟 =
𝑘 ̅
𝑇 6𝜋𝜇
(
3
4𝜋 𝑀𝑊
𝜌 𝐴 𝑁 )
−
1
3
=
𝑘 ̅
𝑇 3𝜋 (6𝜋 2
𝑀𝑊
𝜌 𝐴 𝑁 )
−
1
3

Eq. (11)

The Sherwood number correlation (Eq. 12) was used to determine k.
𝐿 < 𝐿 ∗
; 𝑆 ℎ =
𝑘 𝑑 ℎ
𝐷 = 𝑃 𝑅 𝑒 0.5
𝑆 𝑐 0.33
(
𝑑 ℎ
𝐿 )
0.33

Eq. (12)

2.5 Modeling Approach for the MBR process
Mathematical models consider significant parameters such as temperature, pH, biological
kinetics, and mass-transfer phenomena, and most of these factors can be obtained from laboratory-
scale experiments and some may be estimated and verified by correlation techniques. In general,
the systematic modeling protocol provides means for upscaling or downscaling process reactor in
water treatment, wastewater treatment, and water reclamation and reuse.
A close approach to full-scale systems is often not possible with a laboratory-scale unit,
and therefore an extensive pilot-scale campaign is usually undertaken. A reliable or robust model
can minimize, if not eliminate, the cost and effort associated with such pilot-scale studies. The
laboratory-scale experimental studies provide the best approach for model verification and
validation, as well as model refinement whenever necessary. Additionally, scale-up projection and
30

optimal design criteria are established taking into consideration, process efficiency as well as
operational control and flexibility. Furthermore, minor design modifications can be employed
based on specific process requirements for different applications.

Table 2.1 - Summary of estimated model parameters for plate-and-frame system
Model parameter Symbol (unit) Flat sheet UF membrane
Trans-membrane pressure ΔP (Pa) 2.07×10
5
, 3.10×10
5
, & 4.14×10
5

Cross-flow velocity v (m s
-1
) 0.02, 0.20, & 0.50
Feed concentration Cf (kg m
-3
) 7.2 ×10
-3

Gel layer concentration Cg (kg m
-3
) 1.11 ×10
-2

Membrane resistance κ m (m
-1
) 2.12 ×10
12

Activated carbon
resistance
κ ac (m
-1
) 1.90 ×10
12

Partition coefficient ks 7.301 ×10
-1

Gel layer compactness α (m kg
-1
) 1.32 ×10
16

Diffusion coefficient D (m
2
s
-1
) 1.68 ×10
-10

Mass-transfer coefficient k (m s
-1
) 1.11 ×10
-5

Dynamic viscosity μ (Pa s) 1.00 ×10
-3

Membrane surface area Am (m
2
) 1.55 ×10
-2

Membrane cell cross-
section area
Ac (m
2
) 1.25 ×10
-4

Coefficient for κ cp+in
a
b
c
d
e
f
1.661
7.09 ×10
-3

2.207
1
1.13 ×10
-7

1

31

Table 2.2 - Summary of estimated model parameters for MBR system
Model parameter Symbol (unit)
Hollow fiber UF
membrane
Reactor and influent characteristics
Reactor operating volume V (L) 93
Influent TOC concentration Cf (mg L
-1
) 6.7
Influent reactor flow rate Q (mL min
-1
) 56
Reactor hydraulic retention time HRT (hr) 27.7
Adsorbent concentration (PAC) Cpac (mg L
-1
) 40
Apparent density of activated carbon ρp (g cm
-3
) 0.8
Adsorption equilibrium and mass-transfer parameters
Freundlich capacity coefficient KF (mg/g)(L/mg)
(1/n)
1.7486
Freundlich intensity coefficient 1/n 0.2161
Surface diffusion coefficient Ds (cm
2
s
-1
) 2.0 ×10
-10

Film transfer coefficient kf (cm s
-1
) 4.0 ×10
-3

Biological and biofilm parameters
Monod maximum utilization rate µ max (hr
-1
) 0.8
Monod half saturation coefficient Ks (mg L
-1
) 11.13
Monod microbial yield coefficient Y (mg mg
-1
) 0.754
Monod microbial decay coefficient kd (hr
-1
) 0.074
Biomass concentration in reactor X0 (CFU mL
-1
) 1.0 ×10
-6

Maximum biofilm thickness Lfmax (µm) 10
Free liquid diffusion coefficient for TOC Dl (cm
2
s
-1
) 2.0 ×10
-6

Biofilm diffusion coefficient for TOC Df (cm
2
s
-1
) 1.6 ×10
-6

32

An important technology that has found wide application in water and wastewater
treatment including removal of organic contaminants from water, treatment of different types of
industrial wastewaters, and water reclamation and reuse, is the bioactive sorption process.
Bioactive adsorbers involve the growth of a biofilm on activated carbon surface for degrading the
adsorbed pollutants. Pioneering work on bioactive adsorbers could be attributed to Ying and
Weber (1979), and Andrews and Tien (1981), for the development of mathematical models to
describe the process dynamics. The use of bioactive adsorbers for degrading several chlorinated
organic compounds at trace levels was demonstrated by Bouwer and McCarty (1982). Ying and
Weber (1979) considered the variation of biofilm thickness as a function of time and position
(depth) in the adsorber bed. They further assumed that the biofilm thickness represented less than
that of a mono-layer of bacteria because the biofilm layer was maintained thin by backwashing
and air-scouring the adsorber bed. Hence, the model did not account for additional mass-transfer
resistance to adsorption attributed to intra-biolayer diffusion. However, the model was applicable
to both fixed-bed and fluidized-bed adsorbers, and showed good predictive capability for glucose
and sucrose. The model of Andrews and Tien (1981) considered substrate diffusion through the
biofilm, but did not account for the substrate diffusion into the adsorbent. The model exhibited
good predictive capability for the adsorption and biodegradation of valeric acid in fluidized-bed
adsorbers.
These efforts were followed by those of Chang and Rittmann (1987), and Speitel et al.
(1987), whose models were also based on the fundamental mechanisms of film transport, biofilm
growth, biofilm degradation of the substrate, and substrate adsorption on activated carbon. The
former model was applied for completely mixed stirred-tank reactors (Chang and Rittmann, 1987),
while the latter model was used for fixed-bed reactors (Speitel et al., 1987). These models were
33

tested for single-substrate systems containing easily biodegradable compounds such as phenol or
p-nitrophenol. Pirbazari and coworkers (Kim and Pirbazari, 1989; Ravindran et al., 1997)
developed models for fluidized-bed and stationary-bed bioactive adsorbers for the treatment of
dairy wastewater. More recently, Pirbazari and coworkers (Pirbazari et al., 1991) developed a thin-
biofilm model for expanded-bed adsorbers employed for the removal of alachlor from potable
water supplies by a combination of sorption and biodegradation (Badriyha et al., 2003).
Furthermore, the above models were developed for different reactor configurations ranging from
fixed-bed to fluidized-bed adsorbers but for granular activated carbon adsorbers.
The traditional approach for process design involves feasibility studies followed by pilot-
scale studies and eventually full-scale design. However, this procedure has proven expensive in
view of the time, effort and cost associated with pilot-scale studies. The modeling and design
protocol outlined for MBR process modeling and design in Figure 2.4 would be a useful approach
in economizing time, effort, and costs. The MBR model utilizes biodegradation parameters and
flow conditions as inputs to predict the process dynamics under a variety of operating conditions.
Subsequently, the results from laboratory scale MBR experiments provide the feed-back to
evaluate the predictive capability of the model and to implement any potential model refinement,
if necessary. The next few steps involve procedures for process up-scaling from laboratory to pilot-
scale and eventually to full-scale operation. The concepts of dimensional analysis and similitude
used in the up-scaling procedure are elaborately discussed by Pirbazari and coworkers (Ravindran
et al., 1996; Den and Pirbazari, 2002; Badriyha et al., 2003). The pilot-scale design conventionally
requires verification by limited experimental testing prior to full-scale implementation. The
protocol addresses the steps leading to modeling for predicting process dynamics and model
verification by laboratory-scale experiments.
34

The model discussed herein is intended for powdered activated carbon particles with
biodegradation occurring in the bulk liquid phase as well as in the biofilm immobilized on the
carbon particle. The modeling milieu and rationale discussed here are based on the earlier work by
Pirbazari and coworkers (Pirbazari and Weber, 1984; Pirbazari et al., 1992; and Pirbazari et al.
1996) for the use of MBR systems in various water treatment and wastewater treatment
applications. These include the removals of nitrate and organic contaminants from contaminated
ground waters (Tsai et al., 2004, 2005), as well as the removal of biodegradable organic matter
from natural waters (Williams and Pirbazari, 2007; Williams et al., 2012).

2.6 Model Assumptions and Development
The present research focuses on the application of a mathematical model for description
and performance forecasting of the MBR process dynamics under a variety of operating and
process conditions. This involves the formulation of differential equations with appropriate initial
and boundary conditions describing adsorption equilibrium and kinetics, biodegradation kinetics,
and associated transport phenomena. In order to effectively simulate the system dynamics, the
model must adequately represent the phenomenological aspects of the actual process. Therefore,
reasonable assumptions were made to account for all the physicochemical, biological, and
transport sub-processes relevant to the MBR process, and they can be delineated as follows:
 The MBR system is homogeneous mixing with high velocity of flow rate and recycling.
 The powdered activated carbon (PAC) particles are uniform in size and shape.
 The biofilm on the surface of activated carbon is homogeneous with respect to porosity,
composition, and density.
 The adsorption is considered as a reversible reaction, adsorption and desorption.
35

Figure 2.4 - Protocol for the modeling and design of the MBR for water reclamation

Membrane Bioreactor Modeling Approach
2. Batch Biokinetic Studies
Evaluation of Biodegradability of Organic Matter
3. Chemostat Studies
Determination of Biological Parameters
K s, µ m, Y and k d
6. Modeling the MBR
Process Dynamics
4. Correlation Techniques
Estimation of D, D f, D l and k f
5. Biofilm Parameters
Estimation of X f, L f0 and L fmax
7. Laboratory Scale MBR Studies
8. Model Simulation, Calibration
and Verification
9. Upscaling Using Dimensional
Analysis and Similitude
10. Pilot-Scale Design and Testing
11. Full-Scale System Design and
Implementation
12. Process Evaluation and Cost
Estimation
1. Batch Adsorption Studies
Evaluation of Adsorption Equilibrium and Rate Parameters
36

 The adsorption isotherm describes the equilibrium between the surface and liquid
concentration at the biofilm-carbon interface, that is, Freundlich isotherm model.
 Biodegradation occurs in both bulk liquid phase and the biofilm, while it does not occur
within activated carbon particles.
 Biodegradation and biofilm growth are represented by Monod equation.
 The loss of biomass is adjusted and balanced by new biomass because biofilm thickness
increases with time, and will eventually reach a maximum value. The biomass loss from
the activated carbon surface is directly transferred to the bulk liquid phase in suspended
form.
The schematic of biofilm developed on activated carbon particle is shown in Figure 2.5.
The mathematical equations describing substrate transport, utilization, adsorption, as well as
biofilm degradation and suspended biomass/biofilm growth, with the appropriate boundary are
presented. The equation terms pertaining to various phenomenological aspects are listed in Table
2.3 The processes include the following: (a) diffusion within the activated carbon particles, (b)
adsorption equilibrium occurring at the biofilm-carbon interface, (c) diffusion with biological
reaction inside biofilm, (d) film transfer from bulk liquid phase to the biofilm-liquid interface, and
(e) biological reaction in bulk liquid solution.

37

Figure 2.5 - Schematic diagram of the bioflim on activated carbon

Table 2.3 - Mathematical equations of mass flux and reactions associated phenomenological mass
transfer mechanisms

In activated
carbon particle
Biofilm layer
Dynamic liquid
film layer
Bulk region
Mass flux −𝐷 𝑠 𝜌 𝜕𝑞
𝜕 𝑟 𝑠 −𝐷 𝑓 𝜕 𝑆 𝑓 𝜕 𝑟 𝑓 𝐾 𝑓 ( 𝑆 𝑏 − 𝑆 𝑙𝑓
)
Reaction rate −
𝑘 𝑚 𝑆 𝑓 𝐾 𝑠 + 𝑆 𝑓 𝑋 𝑓 −
𝑘 𝑚 𝑆 𝑏 𝐾 𝑠 + 𝑆 𝑏 𝑋 𝑠

2.7 Model Conceptualization and Formulation
The MBR process should be considered as two non-steady state processes: adsorption and
biodegradation. The MBR model involves adsorption and biodegradation, and accommodates for
microbial degradation in the biofilm as well as in the bulk liquid phase comprised of
Activated
carbon particle
Biofilm
layer
Bulk region
Dynamic
liquid film
38

microorganisms in suspension. The fundamental transportation of substrate occurs by three steps:
liquid film mass transfer, diffusion into biofilm, and biodegradation in biofilm.

Permeate Flux
The mathematical approach of permeate flux is discussed in this chapter and the schematics
of hollow fiber membrane is described in Figure 2.6. In a similar manner, the permeate flux for
hollow fiber membranes can be also described by Darcy’s law. Eq. (13) represents the relationship
between the trans-membrane pressure and the sum of the resistances in series. The differential
form can be written as Eq. (14). The resistance of the membrane, κ m is attributed to the membrane
material and the surface roughness, and this characteristic may not change during membrane
filtration except a limiting case such as extremely high trans-membrane pressure. The
concentration polarization resistance, κ cp is expressed in terms of the trans-membrane pressure,
fluid velocity, and solute concentrations. The internal pore fouling also offers transport resistance
due to solute sorption and pore blocking. These two resistances can be combined and represented
by Eq. (15) and (16). The resistance due to the gel layer, κ g can be denoted by Eq. (17) and is time
dependent. The activated carbon particles may have contributed to particulate fouling. The
resistance of activated carbon, κ cp has not changed basis on time.
𝐽 =
𝛥 𝑃 𝜇 𝜅
=
𝛥 𝑃 𝜇 ( 𝜅 𝑚 + 𝜅 𝑔 + 𝜅 𝑐𝑝
+ 𝜅 𝑖𝑛
+ 𝜅 𝑎𝑐
)

Eq. (13)

𝛥 𝑃 d𝐽 d𝑡 + 𝜇 𝐽 2
(
d𝜅 𝑔 d𝑡 +
d𝜅 𝑐𝑝
d𝑡 +
d𝜅 𝑖𝑛
d𝑡 ) = 0 Eq. (14)

𝜅 𝑐𝑝 +𝑖𝑛
= 𝜅 𝑐𝑝
+ 𝜅 𝑖𝑛
= 𝑎𝑣
𝑏 𝛥 𝑃 𝑐 ( 𝑆 𝑏 + 𝑋 𝑠 )
𝑑 𝐶 𝑝𝑒𝑟 𝑒 𝑆 𝑖𝑛
𝑓 Eq. (15)

39

d𝜅 𝑐𝑝 +𝑖𝑛
d𝑡 = 𝑎𝑑 𝑣 𝑏 𝛥 𝑃 𝑐 ( 𝑆 𝑏 + 𝑋 𝑠 )
𝑑 −1
𝐶 𝑝𝑒𝑟 𝑒 𝑆 𝑖𝑛
𝑓 d( 𝑆 𝑏 + 𝑋 𝑠 )
d𝑡 + 𝑎𝑒 𝑣 𝑏 𝛥 𝑃 𝑐 ( 𝑆 𝑏 + 𝑋 𝑠 )
𝑑 𝐶 𝑝𝑒𝑟 𝑒 −1
𝑆 𝑖𝑛
𝑓 d𝐶 𝑝𝑒𝑟 d𝑡
Eq. (16)

d𝜅 𝑔 d𝑡 = 𝛼 ( ( 𝑆 𝑏 + X
s
)− 𝑘 𝑠 𝐶 𝑔 ) 𝐽 − 𝛼 𝑘 ( ( 𝑆 𝑏 + X
s
)− 𝑘 𝑠 𝐶 𝑔 ) ln
𝐶 𝑔 − 𝑘 𝑠 𝐶 𝑔 ( 𝑆 𝑏 + X
s
)− 𝑘 𝑠 𝐶 𝑔
Eq. (17)

Figure 2.6 - Schematics of (a) hollow fiber UF membrane, (b) solute transport through the hollow
fiber membrane, and (c) concentration profile

Membrane
thickness
Bulk phase
Concentration
polarization
layer
Permeate
phase
Gel layer
Permeate flow
Filtration
flow
Membrane pore
Internal pore
fouling
Organics,
biomass,
impurities
Cross flow
Permeate flow
Filtration
flow
Cross flow
Hollow
fiber
membrane
Membrane pore
Bulk
solution
Concentration
polarization
layer
Permeate
Concentration
Gel layer
Distance
Concentration
Membrane surface
(Boundary layer)
(a)
(b)
(c)
40

Reactor Mass Balance for Substrate in Bulk Solution
The reactor mass balance considering total organic carbon (TOC) as substrate (Sb) can be
written as follows:
𝑑 𝑆 𝑏 𝑑𝑡 =
𝑄 𝑉 ( 𝑆 𝑖𝑛
− 𝑆 𝑏 )−
3𝑋 𝑤 ( 𝑅 + 𝐿 𝑓 )
2
𝑅 3
𝜌 𝑝 𝑉 𝑘 𝑓 ( 𝑆 𝑏 − 𝑆 𝑙𝑓
)−
𝑘 𝑆 𝑏 𝐾 𝑠 + 𝑆 𝑏 𝑋 𝑠 for 𝑡 ≥ 𝑡 0

Eq. (18)

Reactor Mass Balance for Suspended Biomass in Bulk Solution
𝑑 𝑋 𝑠 𝑑𝑡 = (
𝑌 𝑘 𝑚 𝑆 𝑏 𝐾 𝑠 + 𝑆 𝑏 − 𝑘 𝑑 )𝑋 𝑠 for 𝑡 ≥ 𝑡 0
Eq. (19)
In the above relation, Q is the feed flow rate (L s
-1
), p is the apparent particle density (g
cm
-3
), Sin is the substrate concentration in feed solution (mg L
-1
), Xw is particle weight in the system
(mg), Sb is the substrate concentration in bulk liquid (mg L
-1
), and kf is the substrate mass transfer
coefficient (cm s
-1
) through liquid film. km is the Monod specific substrate utilization rate (mg s
-1

mg
-1
), Ks is the Monod half-saturation constant (mg L
-1
), Y is the growth yield coefficient (mg mg
-
1
), kd is overall biomass decay coefficient (s
-1
), Slf is the substrate concentration at the liquid-
biofilm interface (mg L
-1
), R is radius of particle ( m), and V denotes the effective volume of
reactor system (L). Xs is suspended biomass concentration (mg L
-1
), which should be corrected by
the term Xsf (mg L
-1
) which denotes the concentration of biomass from biofilm due to shear loss if
biofilm thickness is larger than maximum thickness after each time-step calculation. Their initial
conditions are as follows:
𝑆 𝑏 = 𝑆 𝑏 0
and 𝑋 𝑠 = 𝑋 𝑠 0
for 𝑡 = 𝑡 0
Eq. (20)

41

Substrate Intraparticle Diffusion
The general partial differential equation representing the pore and surface diffusion
transport in a porous and spherical particle can be written as follows:



























s
p
2
s
s
s
2
s
s
2
s
r
S
D
ρ
θ
r
r
q
D r
r
r
1
S
ρ
θ
q
t
p
p
p
p
p
p
at R r
s
  0 for
0
t t  Eq. (21)
In the above equation, q represents the solid-phase concentration of adsorbed substrate (mg
g
-1
), p denotes the intrapore void space in the particle (dimensionless), Sp is the substrate
concentration in the intrapore void space (mg L
-1
), t is time (s), rs is radial coordinate in particle
( m), and Ds and Dp are the surface diffusion (cm
2
s
-1
) and pore diffusion coefficient (cm
2
s
-1
),
respectively. The MBR model essentially considers surface diffusion as dominating adsorption
kinetics, so that the generalized equation can be rewritten as follows:















s
s
s s
s
r
q
r
r r
D
t
q
2
2
at R r
s
  0 for
0
t t  Eq. (22)
The first boundary condition pertains to the symmetry of contaminant (or substrate)
concentration profile with respect to the adsorbent particle center, and can be written as
0 


s
r
q
at 0 
s
r for
0
t t  Eq. (23)
The second boundary condition pertains to the particle-biofilm interface where the rate of
substrate transport through biofilm-particle interface must balance the rate of change in the amount
of adsorbed substrate in the particle, and can be written as follows:







R

s
2
s r
f
f
f
2
qdr r 4
t r
S
D R 4
f
0
p 0
π π  at
0
for t t R r
s
  Eq. (24)
In the above relation, Sf is the substrate concentration in biofilm (mg L
-1
), Df is the diffusion
coefficient in biofilm (cm
2
s
-1
), and rf is radial coordinate in biofilm ( m). The left hand side
42

represents the substrate mass flux at the biofilm-particle interface, and can be equated to the
product of the radial concentration gradient, biofilm diffusivity of substrate, and surface area of
adsorbent particle. The right hand side represents the mass rate of accumulation of the substrate
over the entire adsorbent particle. The initial condition for the above equation can be stated as
follows:
0
q q  at R r
s
  0 for
0
t t  Eq. (25)

Substrate Adsorption Equilibrium
A local adsorption equilibrium exists at the biofilm-adsorbent interface between the
substrate concentrations in the liquid and solid phases, described by the Freundlich equation
n
fs F fs
S K q
1
) (  at
0
for t t R r
s
  Eq. (26)
where KF and 1/n denote the Freundlich adsorption capacity constant (mg g
-1
)(mg L
-1
)
(-1/n)
and the
Freundlich adsorption intensity constant (dimensionless), respectively.

Substrate Diffusion with Reaction in Biofilm
The model assumes mass transfer across the liquid-biofilm interface as well as diffusion
(Fick’s law) and biological reaction (Monod kinetics) within the biofilm. The non-steady state
relationship for diffusion with Monod type reaction within the biofilm can be expressed as
f
f s
f m
f
f
f
f
X
S K
S k
r
S
D
t
S







2
2
at
0
for 0 t t L r
f f
   Eq. (27)
where Xf is the biofilm cell density (mg L
-1
).
Since the differential equations (22) and (27) share the same boundary at the biofilm-
particle interface, the boundary condition, Eq. (12) must be satisfied for Eq. (24) as well. It must
43

also be noted that at the liquid-biofilm interface, the flux across the interface from the liquid phase
must be balanced by the flux across the interface into biofilm, so that the second boundary equation
can be written as follows:
f f
f f
L r
f
f
f
L r
f b f
r
S
D S S k













0
or f at t t r r
f
  Eq. (28)
The initial condition for Eq. (28) can be written as
0 f f
S S  at
f f
L r   0 for
0
t t  Eq. (29)

Biofilm Growth Kinetics
As the substrate diffuses into the biofilm, the microbial biomass utilizes the substrate for
biosynthesis and respiration. The biomass can increase with time until the growth rate reaches a
steady state value. The biofilm thickness is assumed to increase with time during its growth phase
because biomass density within the biofilm is assumed constant. It will reach a maximum due to
the balance between microbial growth, decay and shear loss. Its temporal variation can be
described by the following equation and boundary conditions:
f
L
d
f s
f m f
dr k
S K
S Yk
dt
dL
f













0
for
0
t t  Eq. (30)
V R L L X X X L L
fmax f f w sf fmax f p
/ ) ( 3 and     if Lf > Lfmax Eq. (31)
where Lf denotes the biofilm thickness ( m), and Xsf is the biomass (mg L
-1
) lost from biofilm to
bulk solution. In the above boundary condition, it must be noted that once the biofilm thickness
tends to exceed the maximum level Lfmax, transfer of excess biomass from the biofilm to the bulk
fluid occurs. The initial condition corresponding to the above equation can be written as follows:
0 f f
L L  for
0
t t  Eq. (32)
44

The complete model for the MBR process is assembled from the model components
described in the above equations.

2.8 Modeling under Different MBR Process Conditions
The MBR model is presented in its most general form, incorporating all mechanisms for
contaminant removal operative in the MBR process, including, adsorption, biofilm degradation,
and liquid suspension biodegradation. However, in the present study, a phenomenological
decomposition approach will be employed to determine the relative contributions of different
phenomena to contaminant removal. Model simulation studies can be performed under different
process scenarios: (a) adsorption alone occurs; (b) biodegradation is operative in biofilm and liquid
phase suspension, but adsorption is absent; (c) adsorption and biofilm degradation are operative,
but liquid phase biodegradation is nonexistent; (d) adsorption and biodegradation in biofilm and
liquid phase suspension are operative, conforming to the assumptions of the generalized MBR
model. The modifications in the model equations, boundary conditions, or parameter values
associated with different scenarios are discussed herein. If adsorption alone were to be operative,
the biodegradation effects in the biofilm and suspension will be suppressed by setting the Monod
specific substrate utilization rate km, and the biomass density in biofilm Xf , to zero. On the other
hand, if degradation in biofilm and suspension phase alone are to be operative, then adsorption
effects will be suppressed by setting the Freundlich capacity constant KF to a low value (say 10
-5

or so), practically approaching zero. Furthermore, if adsorption and biofilm degradation were to
be operative but not suspension phase biodegradation, the boundary condition corresponding to
the Eq. (18) describing biofilm growth can be modified, ignoring the shearing or sloughing of
biofilm from the adsorbent particle to the suspension phase. The condition Lf = Lfmax shall remain
45

in the overall setting, but the condition Xsf = 3XwXf (Lf – Lf,max)/R pV for Lf > Lfmax will be
excluded.

2.9 Model Parameter Estimation
Adsorption Isotherm Parameters: KF and 1/n
Completely mixed batch reactor technique is generally employed for estimating the
adsorption equilibrium parameters (Pirbazari et al., 1992). The linearized Freundlich equation is
shown below from which the Freundlich parameters KF and 1/n were determined by regression
analyses:
e F e
S K q
10 10 10
log
n
1
log log   Eq. (33)
In the above relation, qe and Se represent the contaminant equilibrium concentrations in the
liquid and solid phases, respectively.

Estimation of adsorption mass transfer parameters Ds and kf
The adsorption mass transfer or rate parameters, Ds and kf shall be estimated from
adsorption batch reactor rate studies. The adsorption rate data for various adsorption scenarios
pertaining to TOC or other organic constituetns will be fitted by simulations using the
homogeneous surface diffusion model (HSDM) for batch reactor systems, as described by
Pirbazari et al (1991) and Badriyha et al (2003). In this procedure, the model simulations will be
performed using a parameter search techniques with respect to Ds and kf so as to minimize the
summation of normalized least-square differences between the experimental data and the
simulated values. These values will be compared with those obtained using correlation techniques.

46

Intraparticle Surface Diffusion Coefficient: Ds
Completely mixed batch (CMB) reactors experiments are usually conducted to determine
the surface diffusion coefficient Ds for adsorbates (Pirbazari et al., 1992). The homogenous surface
diffusion model for the batch reactor system is commonly employed for estimating Ds for using
the rate-parameter-search program by minimizing the variation between the model output and the
experimental data (Crittenden and Weber, 1978a; Crittenden and Weber, 1978b; Kim and Pirbazari,
1989; Pirbazari et al., 1992).

Biological Kinetic Parameters: k, Ks, Y and kd
The biokinetic studies discussed in Chapter 4 was used to determine the biokinetic
parameters for TOC removal. These biokinetic parameters will include the Monod coefficients, km
and Ks, the yield coefficient, Y, and the decay coefficient kd for the model compound. The details
of chemostat studies for determining the biokinetic parameters are discussed by Kim and Pirbazari
(1989) and Ersever et al (2007).

Free Liquid Diffusivity: Dl
The correlation technique proposed by Wilke and Chang (1955) was employed for
estimating Dl, as shown below:
6 . 0
5 . 0
8 2
) (
) (
10 4 . 7 sec) / cm (
A b B
B
AB l
V
T MW
D D


  

Eq. (34)
In the above equation, DAB represents the free liquid diffusivity of solute A in solvent B,
(Vb)A is modal volume of solute A at normal boiling point expressed in cm
3
g
-1
mol
-1
, T is the
absolute temperature in K, MWB is the molecular weight of component B,  is viscosity of solvent
expressed in centipoise,  is an association factor of solvent B, dimensionless.
47

Biofilm Diffusion Coefficient: Df
The biofilm substrate diffusion coefficients Df for the organic substrate was estimated from
the corresponding free liquid diffusivity using the relation Df = 0.8 Dl, employed for performance
predictions of bioactive adsorbers by Pirbazari and coworkers (Kim and Pirbazari, 1989;
Ravindran et al., 1996, 1997).

Liquid Film Transfer Coefficient: kf
In the MBR process, the bioactive PAC particles, microbial suspension, and bulk liquid
phase are continuously re-circulated through the filter tube assembly line and by-pass piping
system, analogous to a two-phase flow reactor. The liquid-film mass transfer rate between the
adsorbent particle and surrounding fluid is important from a modeling or design standpoint,
because it could be rate-limiting in comparison with particle surface or pore diffusion transport.
Some investigators have observed that in stirred tank reactors, a correlation exists between the
mass transfer coefficient and energy dissipation rate (Brain et al., 1969; Ohashi et al., 1979). The
energy dissipation rate is a measure of fluid turbulence and vorticity, factors that affect the
boundary layer thickness surrounding the adsorbent particle, and consequently influence substrate
transport rate from the bulk fluid to the adsorbent particle. Among several correlations reported in
literature, each distinguished by a particular relationship between Reynolds, Schmidt and
Sherwood numbers, that attributed to Ohashi and coworkers (1979) was found most appropriate.
The correlation considers the effects of such parameters as liquid flow rate, tube diameter, particle
diameter, particle density, and temperature, and is given by
Sh = 2 + 0.44Re
0.63
Sc
1/3
(12< Re <190, 250< Sc <2300) Eq. (35)
where Sh = kf dp /Dl
48

The other relevant dimensionless numbers are defined by the relations
Re =
v
d
p
3 / 4 3 / 1

and
l
D
v
Sc  Eq. (36)
where dp is the diameter of adsorbent particle, Dl the diffusivity of model compound,  the kinetic
viscosity of the fluid, and  the energy dissipation rate per unit mass of liquid. In a pipe flow
system,  is experimentally correlated to
T w
D U f / ) ( 2
3
, wherein the empirical form of the Fanning
friction factor f is given by the relation
1/4
w T
/v) U 0.082(D f

 in the range of (DTUw/v)>10
4
,
where DT represents the inner pipe diameter, and Uw denotes the average fluid velocity.

Biofilm Thickness: Lf
The biofilm is comprised approximately 99% water, and so an average value of Lf was
estimated from the relation employed by Kim and Pirbazari (1989):
 

p w
e p
f
W
W
L
) 99 . 0 (
 Eq. (37)
where We is weight of evaporated water and Wp is dry weight of retained particles,  is the ratio of
particle surface area to particle volume and w represents water density. This technique is
commonly employed for estimating the initial and maximum biofilm thicknesses, Lf0 and Lfmax,
respectively.

Biofilm Density: Xf
The biomass density in the biofilm is generally estimated from the following equation:
f p
w p
f
L W
B
X


 Eq. (38)
49

where Bw represents the difference between the weights of adsorbent particles with biofilm before
drying and the weight of virgin adsorbent particles.

2.10 Model Sensititvity Analysis
Sensitivity analyses in mathematical modeling confers certain advantages to process desgin
and upscaling, that can be briefly summarized as follows: (i) It facilitates gaining insight into the
process dynamics, as well as quantification of changes in magnitude and direction in observed
variables expected in response to variations in certain parameters and/or dependent variables. Thus,
it provides relevant information on process conditions that can be altered by modifying certain
parameters to enhance process efficiency. (ii) It offers the modus operandi for possible model
expansion, reduction, or refinement, taking into consideration the relative importance of various
phenomenological model components. (iii) It enables improvement in methodologies for
parameter estimation with reference to uncertainities of observable variables, also indicating
certain conditions or features not suitable for such estimations.
There are two basic methods of sensitvity analyses, namely, the global method and the
local method. Global methods establish the domains of parameteric importance over entire regions
in parameter space, and the sensitivity informaion represents an ensemble or temporal average.
Local methods, on the other hand, seek to find variations in observable variables at fixed points
in parameter space. Local methods have more applicability in process modeling than global
methods because they are amenable to more straightforward interpretations. Although sensitivity
investigations were conducted for several model parameters, only selected parameters that
profoundly influenced process dynamics are discussed.

50

2.11 References
Andrews, G. F. & Tien, C. (1981). Bacterial film growth in adsorbent surfaces. American Institute
of Chemical Engineers Journal, 27(3), 396-403.

Badriyha, B. N. Ravindran, V., Den, W., & Pirbazari, M. (2003). Bioadsorber efficiency, design,
and performance forecasting for alachlor removal. Water Research, 37(17), 4051-4072.

Bouwer, E. J. & McCarty, P. L. (1982). Removal of trace chlorinated organic compounds by
activated carbon and fixed-film bacteria. Environmental Science and Technology, 16(12), 836-843.

Brian, P. L. T., Hales, H. B., & Sherwood, T. K. (1969). Transport of heat and mass between
liquids and spherical particles in an agitated tank. American Institute of Chemical Engineers
Journal, 15(5), 727-733.

Chang, H. T. & Rittmann, B. E. (1987). Mathematical modeling of biofilm on activated carbon.
Environmental Science and Technology, 21(3), 273-280.

Crittende, J. C. & Weber, Jr., W. J. (1978a). Predictive model for design of fixed-bed adsorbers:
Parameter estimation and model development. Journal of the Environmental Engineering Division,
ASCE, 104(2), 185-197.

Crittende, J. C. & Weber, Jr., W. J. (1978b). Predictive model for design of fixed-bed adsorbers:
Single-component model verification. Journal of the Environmental Engineering Division, ASCE,
104(3), 433-443.

Den, W. & Pirbazari, M. (2002). Modeling and design of vapor-phase biofiltration for chlorinated
volarile organic compounds. American Institute of Chemical Engineers Journal, 48(9), 2084-2103.

Ersever, I., Ravindran, V., & Pirbazari, M. (2007). Biological denitrification of reverse osmosis
brine concentrates: I. Batch reactor and chemostat studies. Journal of Environmental Engineering
and Science, 6(5), 503-518.
51

Kim, S. H. & Pirbazari, M. (1989). Bioactive adsorber model for industrial wastewater treatment.
Journal of Environmental Engineering, 115(6), 1235-1256.

Ohashi, H., Sugawara, T., Kikuchi, K., Henmi, T. (1979). Mass transfer between particles and
liquid in solid-liquid two-phase upflow in vertical tubes. Journal of Chemical Engineering of
Japan, 12(3), 190-195.

Pirbazari, M., Badriyha, B. N., & Miltner, R. J. (1991). GAC adsorber design for removal of
chlorinated pesticides. Journal of Environmental Engineering, 117(1), 80-100.

Pirbazari, M., Badriyha, B. N., & Ravindran, V. (1992). MF-PAC for treating waters contaminated
with natural and synthetic organics. Journal American Water Works Association, 84(12), 95-103.

Pirbazari, M., Ravindran, V., Badriyha, B. N., & Kim, S. H. (1996). Hybrid membrane filtration
process for leachate treatment. Water Research, 30(11). 2691-2706.

Pirbazari, M. & Weber, Jr., W. J. (1984). Removal of dieldrin from water by activated carbon.
Journal of Environmental Engineering, 110(3), 656-669.

Ravindran, V., Badriyha, B. N., Pirbazari, M., & Kim, S. H. (1996). Modeling of bioactive carbon
adsorbers: A hybrid weighted residual-finite difference numerical technique. Applied Mathematics
and Computation, 76(2-3), 99-131.

Ravindran, V., Kim, S. H., Badriyha, B. N. & Pirbazari, M. (1997). Predictive modeling for
bioactive fluidized bed and stationary bed reactors: Application to dairy wastewater.
Environmental Technology, 18(11), 861-881.

Speitel, G. E. J., Dovantzis, K., & DiGiano, F. A. (1987). Mathematical modeling of
bioregeneration in GAC columns. Journal of Environmental Engineering, 113(3), 32-48.

52

Tsai, H. H., Ravindran, V., Williams, M. D., & Pirbazari, M. (2004). Forecasting the performance
of membrane bioreactor process for groundwater denitrification. Journal of Environmental
Engineering and Science, 3(6), 507-521.

Tsai, H. H., Ravindran, V. & Pirbazari, M. (2005). Model for predicting the performance of
membrane bioadsorber reactor process in water treatment applications. Chemical Engineering
Science, 60(20), 5620-5636.

Wilke, C. R. & Chang, P. (1955). Correlation of diffusion coefficients in dilute solutions. American
Institute of Chemical Engineers Journal, 1(2), 264-270.

Williams, M. D. & Pirbazari, M. (2007). Membrane bioreactor process for removing
biodegradable organic matter from water. Water Research, 41(17), 3880-3893.

Williams, M. D., Ravindran, V., & Pirbazari, M. (2012). Modeling and process evaluation of
membrane bioreactor for removing biodegradable organic matter from water. Chemical
Engineering Science, 84(20), 494-511.

Ying, W. & Weber, Jr., W. J. (1979). Bio-physicochemical adsorption model systems for
wastewater treatment. Water Pollution Control Federation Journal, 51(11), 2661-2677.

53

Chapter 3
Investigating the Potential of Flat Sheet Ultrafiltration Membranes
for Water Reclamation and Reuse
3.1 Introduction
Ultrafiltration (UF) membranes are designed to remove particulates, colloids,
microorganisms, and organic matter in water and wastewater applications. These membranes are
typically known to have pore size range of 0.01 to 0.1 micrometer (µm) and a molecular weight
cutoff (MWCO) range of 10,000 to 100,000 Daltons (Da). The UF membrane filtration system can
be operated at relatively low trans-membrane pressures ranging from 200 to 400 kPa (that is from
30 tod 60 psi).
Polymeric flat sheet UF membranes have been generally used in laboratory-scale tests to
evaluate membrane performances and intrinsic mechanisms because of their flexible material
characteristics (Maartens et al., 1996). These membranes are made of a variety of polymers and/or
polymer blends, typically cellulose acetate (CA), polysulfone (PS), polyethersulfone (PES),
polyvinylidene fluoride (PVDF), polypropylene (PP), etc. (Ulbricht, 2006). In addition, flat sheet
polymeric membranes have a flexible customization of surface modification (Pezeshk et al., 2012).
Flat sheet membrane laboratory experiments provide preliminary data under a variety of operating
condition regarding permeate flux pattern, total organic carbon (TOC) removal, and effect of
concentration polarization and gel formation on membrane surfaces (Kwon et al., 2006). The data
can be used to calibrate the developed transport models for prediction/simulation of membrane
process dynamics. Other applications of flat sheet membranes are tangential-flow modules and
single-layer spiral-wound and track-etched configurations (Mosqueda-Jimenez et al., 2004).
54

Activated carbon has been widely used as an adsorbent in water and wastewater treatment
applications because of its effectiveness in removal of broad spectrum of organic matter. It can
effectively remove color, odor, taste compounds, endocrine disrupting chemicals (EDCs), and
pharmaceuticals and personal care products (PPCPs) in water phase (Gao et al., 2011). The
combination of powdered activated carbon (PAC) and UF processes could significantly increase
the efficiency of membrane filtration performances and improve water quality as regards TOC
concentration (Campos et al., 2000). Nevertheless, PAC may lower permeate flux to some degree.
The addition of microorganisms to feed reservoir may reduce the organic fouling because
organics in wastewater can be digested by microorganisms as food resources and can be eventually
mineralized to carbon dioxide and water (Volk et al., 2000; Zhou et al., 2000). However, Motlagh
et al. (2013) stated that excessive biomass can generate biological fouling on membrane surfaces
resulting in permeate flux decline. Escherichia coli (E.coli), is a commonly used biological strain
in biological research, and serves as one of the important indicators used in the assessment of
treated wastewater quality (Jin et al., 2004). The E.coli was used in these studies to investigate
permeate flux decline, TOC removal, and the overall permeate quality.
Ozone is a widely used oxidant in water and wastewater treatment applications. Ozone
typically reacts with unsaturated double bonds and hence organic compounds may not be
completely mineralized during ozonation. Pre-ozonated wastewater may reduce the surface fouling
such as organic and biological fouling (Lehman et al., 2009), but increase the internal pore fouling
due to incomplete mineralization. Furthermore, ozonation can change the characteristics of
wastewater in terms of hydrophobicity or hydrophilicity (Chiang et al., 2009) and project the
significance to membrane filtration.
55

The oxidizing power of combination of ozone and hydrogen peroxide in decomposition of
wastewater organic matter is reported to be more than that of ozonation alone (Andreozzi et al.,
1999). This phenomenon can be attributed to the formation free radicals that possess more
oxidizing power.
This chapter investigates the following aspects:
i: Effect of trans-membrane pressure on permeate flux
ii: Effect of PAC on permeate flux and TOC
iii: Effect of E.coli on permeate flux and TOC
iv: Effect of combination of PAC and E.coli on permeate flux and TOC
v: Effect of ozonation on permeate flux and TOC
vi: Effect of ozonation plus hydrogen peroxide on permeate flux and TOC
vii: Application of a pore diffusion mathematical model for prediction/simulation of flat
sheet membrane processes
viii: Determining a most appropriate computational approach to solve the model
equations

3.2 Materials and Methods
Materials
Membrane: Polyethersulfone (PES) UF flat sheet membrane, YMPWSP3001, manufactured by
GE Osmonics, Inc. (Minnetonka, MN) was used in this study. It was characterized by a nominal
pore size of 4-10 nm and a 10,000 molecular weight cutoff (MWCO). The tolerance of pH ranged
from 1 to 11 at 25
o
C. These membranes have an asymmetric structure and are moderately
hydrophobic. The effective surface area for each sheet, as purchased, was 150 cm
2
(24 inch
2
).
56

Feed Solution: Secondary clarifier effluent (SCE) was obtained from the San Jose Creek Water
Reclamation Plant (SJCWRP) in Los Angeles County. Figure 3.1 shows the wastewater treatment
processes at the SJCWRP. Samples for this study were taken from the secondary clarifier effluent
just before the chemical addition step. The samples were filtered by using a 5 µm pore filter paper
in order to remove the larger particulate. The characteristics of wastewater are shown in Table 3.1.
The analyses of wastewater were periodically performed in accordance with EPA standard
methods (See Analytical Methods in this section).

Microorganism: Escherichia coli (E.coli) type ATCC 8739, characterized and cultured by
BioMérieux Inc. (Durham, NC), was used to examine the extent of SCE biodegradation and to
investigate the effect of microorganisms on permeate flux decline. The E.coli was cultivated for
two days to reach a steady phase. Appropriate quantities of cultivated E.coli were transferred into
the feed reservoir followed by complete mixing.

Adsorbent: Powdered activated carbon (PAC) was used as adsorbent. WPH-M PAC was
manufactured by Calgon Carbon Corporation (Pittsburgh, PA).

Oxidants: Ozone gas used in this study was generated by a bench scale ozone generator, (LAB2B;
Triogen, UK). Hydrogen peroxide of ACS reagent grade was purchased from Sigma-Aldrich (St.
Louis, MO).

57

Figure 3.1 - Wastewater treatment plant at San Jose Creek Water Reclamation Plant in Los Angeles
County (adapted from San Jose Creek Water Reclamation Plant)
58

Table 3.1 - Wastewater characteristics at San Jose Creek Water Reclamation Plant (Aug. 2011 -
Oct. 2014) and experimental permeate quality of UF membrane filtration
Items Influent
Primary
effluent
Secondary
effluent
Tertiary
effluent
Permeate
using UF
**

BOD5
(mg/L)
150 – 250 180 – 210 3 – 10 3 – 8 1 – 5
COD
(mg/L)
660 – 950 310 – 570 10 – 25 10 – 20 3 – 10
TOC
(mg/L)
90 – 98* 90 – 97* 7 – 8 7 – 8 4.5 – 5.5
UV254
(Abs)
0.85 – 0.89* 0.80 – 0.85* 0.20 – 0.25 0.02 – 0.25 0.15 – 0.18
Total Coliform
(MPN/100mL)
10
10
10
10
10
6
< 10 0
Turbidity
(NTU)
140 – 160 40 – 50 0.8 – 1.0 0.7 – 1.0 0.08 – 0.1
TSS
(mg/L)
600 – 1500 85 – 100 3 – 5 1 – 2 0
Alkalinity
(mg/L as CaCO 3)
450 – 550 300 – 360 200 – 240 200 – 220 200 – 250
Cl
-
*
(mg/L)
140 – 145 140 – 145 170 – 175 170 – 175 132 – 175
NO3
-
- N *
(mg/L)
0.2 – 1.0 0.2 – 1.0 3.0 – 4.5 3.0 – 3.5 2.3 – 3.5
PO4
3-
- P *
(mg/L)
2.0 – 2.5 2.0 – 2.5 0.5 – 0.7 1.0 – 1.5 0.19 – 0.8
SO4
2-
*
(mg/L)
150 – 160 150 – 160 140 – 145 145 – 150 108 – 145

* filtered by 0.2 µm syringe filter
** UF membrane filtration from secondary clarifier effluent
59

Chemical Cleaning Agents: In these experiments, three types of chemical cleaning agents were
used to defoul the membranes; a caustic solution (sodium hydroxide), a surfactant solution, and an
enzyme solution. An ACS reagent grade sodium hydroxide was used for caustic cleaning. A
laboratory grade surfactant, Triton X-100 polymer, (Dow Chemical Company, Midland, MI) was
employed for surfactant cleaning. An RID-X (Reckitt Benckiser, NJ) enzyme cleaner was used for
organic and biological cleaning.

Methods
Membrane Filtration: Stainless steel plate-and-flame membrane cell (SEPA CF Membrane Cell;
Osmonics, Minnetonka, MN) was employed in these experiments (Figure 3.2). The SEPA CF
Membrane Cell is widely used to evaluate the membrane separation techniques (Ates et al., 2009;
Dang et al., 2010; Rudie et al., 1993).
Virgin flat sheet membranes were submerged in DDI water for one hour prior to membrane
filtration tests in order to remove surface impurities. Feed spacers and permeate carriers were
similarly cleaned to remove any foulants. A membrane was installed between the feed spacer and
the permeate carrier. An assembled membrane cell body was inserted into the membrane cell
holders, and this provided constant pressures at desired levels.
Membranes were tested in the Osmonics cross flow membrane system using different types
of feed solutions: (1) effluent of secondary clarifier under different trans-membrane pressures
(TMP) of 20, 30, 40, 50, and 60 psi, (2) secondary clarifier effluent with addition of E.coli (10
6

and 10
8
CFU per 100 mL) and a TMP of 30 psi, (3) addition of 40 mg L
-1
PAC to secondary
clarifier effluent with a TMP of 30 psi, (4) secondary clarifier effluent with addition of both E.coli
and PAC with a TMP of 30 psi, and (5) ozonated secondary clarifier effluent with a TMP of 30
60

psi using (i) 5 mg L
-1
of ozone dosage, and (ii) combination of 5 mg L
-1
of ozone and 5 mg L
-1
of
hydrogen peroxide. The temperature of feed waters was maintained at 25
o
C by using a water bath
equipped with a heat exchanger unit. The TMP was maintained at 30 psi (2.07 bar) and the cross-
flow rate was kept at 1.5 liters per minute (L min
-1
).

Figure 3.2 - Schematic of flat sheet cross-flow membrane filtration unit

Membrane Backwashing and Chemical Cleaning: Fouled membranes were backwashed with
DDI water by employing an in situ method by reversing the flow direction and applying a TMP of
60 psi for 30 minutes. Chemical cleaning was also performed using a similar procedure for 1 hour
followed by rinsing with DDI water for 30 minutes several times (3-5 times) until the residuals of
chemical agents were completely removed.

Analytical Methods
Total Organic Carbon: The total organic carbon (TOC) was measured by non-purgeable organic
carbon (NPOC) method. The samples were acidified with phosphoric acid and purged with carbon
free air to remove inorganic carbon species. A Shimadzu TOC-V CSH analyzer (Shimadzu
Concentrate
Permeate
Flowmeter
Pressure gauge
Flat sheet membrane
Cell body
Cell holder
Feed pump
Influent Feed reservoir
61

Corporation, Kyoto, Japan) was employed in this study, and the analytical procedure used was in
accordance with EPA 415.1, and standard methods 5310B (APHA et al., 2005).

Inorganic Anions: Inorganic anions such as chloride (Cl
-
), nitrate (NO3
-
), phosphate (PO4
3-
), and
sulfate (SO4
2-
) were measured by Ion Chromatography (ICS-1100; Dionex Corporation,
Sunnyvale, CA) using an IonPac
®
AS4A-SC 4 mm anion-exchange column. The performance of
column met the standards specified in US EPA Method 300.0 (A). The detection limits for chloride,
nitrate, phosphate, and sulfate were 0.02 mg L
-1
, 0.008 mg L
-1
, 0.009 mg L
-1
, and 0.02 mg L
-1
,
respectively.

Biological Oxygen Demand: The BOD was determined by 5-day BOD and measured by Winkler
Titration Method in accordance with EPA 360.1.

Chemical Oxygen Demand: The COD was measured by HACH Method 8000, the Reactor
Digestion method approved by US EPA for wastewater analysis.

UV Absorbance: The UV absorbance values represent the amount of total organic matter in
samples. The UV absorbance was determined by a LAMBDA 35 UV/Vis Spectrophotometers
(PerkinElmer, MA) set at a wavelength of 254 nm.

Total Coliform: The presence of high coliform density in water samples indicates that pathogenic
bacteria may potentially exist at a high probability. The coliform bacterial density was estimated
by the MPN Index Method (APHA et al., 2005). The measurement followed the Standard Total
62

Coliform Fermentation Technique discussed in Standard Method 9221B (APHA et al., 2005)

Turbidity: Turbidity was determined by using Hach 2100N Laboratory Turbidimeter equipped
with a stable halogen-filled tungsten filament lamp to meet the requirements of EPA Method 180.1.

Total Suspended Solids: The total suspended solids were determined according to procedures
outlined in Standard Method 2540D (APHA et al., 2005).

Alkalinity: Alkalinity was determined by titration method using 0.02 N sulfuric acid, bromosol
green, and phenolphthalein indicators, in accordance with the Standard Method 2320B (APHA et
al., 2005).

3.3 Ozonation Batch Studies
The ozonation studies were intended to evaluate the effect of ozone oxidation of organic
matter on membrane fouling and permeate flux reduction. The completely mixed batch (CMB)
ozonation reactor system shown in Figure 3.3 consisted of a glass reactor with a heat exchanger to
control constant temperature, an ozone diffuser connected to an ozone generator and an air
compressor. A combination of ozone and hydrogen peroxide was used to study the effectiveness
of oxidation on permeate flux. The concentrations of ozone (O3) and hydrogen peroxide (H2O2)
were each 5 mg L
-1
. The ozonation performance was investigated for ozone alone and a
combination ozone and hydrogen peroxide (peroxone). The total reaction time was between 1 to 2
hours. Samples were periodically withdrawn from the reactor, and the concentrations of total
organic compounds (TOC) were determined by the TOC analyzer and the UV-Vis
spectrophotometer. The reactor was maintained at a temperature of 25
o
C.
63

Figure 3.3 - Schematic of experimental setup for ozonation system

3.4 Adsorption Isotherm Studies
Adsorption isotherm studies are employed to evaluate the adsorption phenomena including
three interactions: (1) adsorbate-water interactions, (2) adsorbate-surface interactions, and (3)
water-surface interactions (MWH, 2005). In general, activated carbon particles are widely used as
adsorbents in water and wastewater treatments because they have nonpolar properties, large
surface areas, and the ability to effectively remove organic matter. The adsorption experimental
data are fitted to adsorption equilibrium relationships such as the Langmuir and Freundlich models.
The Langmuir adsorption isotherm describes the equilibrium between adsorbent and water
interaction (chemisorption) resulting in monolayer adsorption phenomena (Langmuir, 1918). The
Langmuir model is expressed as Eqs. (1) and (2) (Chung, 2015).
𝑞 𝑒 = (
𝐾 𝐿 ∙ 𝐶 𝑒 1 + 𝐾 𝐿 ∙ 𝐶 𝑒 ) ∙ 𝑞 𝑚𝑎𝑥
Eq. (1)

Ozone generator

Heat exchanger
controller

Sampling port
Stirrer
Dry air
Ozone reactor
Fine diffuser
Ozone bubble
Ozone vent
64

1
𝑞 𝑒 = (
1
𝐾 𝐿 ∙ 𝑞 𝑚𝑎𝑥
) ∙
1
𝐶 𝑒 +
1
𝑞 𝑚𝑎𝑥
Eq. (2)

The Freundlich adsorption isotherm is an empirical adsorption model for multilayer adsorption
(Foo, 2010; Halsey, 1948). The Freundlich equation in its non-linear and linear forms are described
in Equations (3) and (4), respectively (Chung, 2015).
𝑞 𝑒 = 𝐾 𝐹 ∙ 𝐶 𝑒 1/𝑛
Eq. (3)

log 𝑞 𝑒 =
1
𝑛 log 𝐶 𝑒 + log 𝐾 𝐹
Eq. (4)

An adsorption kinetic test was conducted to determine the equilibrium time. Powdered
activated carbon (PAC) was added into 2 L of secondary clarifier effluent (SCE) to achieve a final
PAC concentration of 40 mg L
-1
. The solution was mixed homogenously using an agitator.
Samples were periodically taken to measure total organic carbon (TOC) concentration. Figure 3.4
shows residual TOC concentration in the reactor as a function of time. As can be observed, the
TOC concentration was drastically declined during the first 2 hours and approached steady state
after 24 hours, indicating that most of the adsorption occurred during the initial stages of the
experiment.
Batch adsorption experiments were conducted to determine adsorption isotherm
parameters. Various amount of PAC ranging from 0 to 25 mg were placed into 50 mL of centrifuge
tubes containing SCE solution, 7.2 mg L
-1
of TOC. The tubes were vigorously shaken using a
horizontal shaker at 25
o
C for 7 days. The residual TOC was measured by a TOC Analyzer
(Shimadzu Corp., Japan) after filtering the samples using 0.45 μm syringe filters. The experimental
data were fitted with the linearized equations (Equations (2) and (4)) and the results are graphically
presented in Figure 3.5. The estimated values of the parameters are presented in the Table 3.2. As
can be observed, the experimental data fitted the Langmuir and Freundlich isotherm models very
65

well. However, the Freundlich isotherm model provided a better fit and was therefore used to
predict TOC removal.

Figure 3.4 - Determination of equilibration time with 100 mg L
-1
of PAC in SCE

Table 3.2 - Adsorption isotherm parameters for Freundlich and Langmuir isotherms
Isotherm Parameter Value Unit
Freundlich
Freundlich isotherm constant (KF) 1.7486 (mg/g)(L/mg)
(1/n)

Freundlich exponent (1/n) 0.2161 unitless
Regression (R
2
) = 0.963
Langmuir
Langmuir isotherm constant (KL) 0.0544 L/mg
Maximum adsorption capacity
(qmax)
6.013 mg/g
Regression (R
2
) = 0.945
0
1
2
3
4
5
6
7
8
9
10
0 24 48 72 96 120 144 168 192 216 240 264
Residual TOC (mg/L)
Time (hr)
66

(a)

(b)

Figure 3.5 - Determination of adsorption patterns by fitting experimental data to (a) Freundlich
and (b) Langmuir isotherms

y = 3.0577x + 0.1663
R² = 0.9451
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 0.01 0.02 0.03 0.04 0.05 0.06
1/qe
1/Ce
y = 0.2161x + 0.2427
R² = 0.963
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 0.5 1 1.5 2 2.5 3
log qe
log Ce
67

3.5 Results and Discussion
Permeate Flux and Trans-membrane Pressure
In all the membrane filtration tests ranging from microfiltration to reverse osmosis
membranes, the relationship between the permeate flux and trans-membrane pressure (TMP) is
important in evaluating the aqueous transport properties (Arkhangelsky et al., 2008; Daniş et al.,
2009). With different feed solutions and operating conditions, the relationship describes the
transport characteristics of membranes such as the intrinsic membrane resistance and membrane
structure (Wijmans et al., 1985).
Flat sheet membrane filtration tests were conducted using DDI water to investigate the
effect of TMP on permeate flux. Figure 3.6 shows the permeate flux as a function of the TMP. It
can be seen from the results that the flux increased almost linearly with the TMP in the range of
20 - 60 psi, similar to the observations of Tansel et al., 2009. It must be noted that DDI was devoid
of organic, inorganic, and biological foulants, and generally, had low range TOC of 0.05 to 0.15
mg L
-1
.

Figure 3.6 - Relationship between permeate flux and trans-membrane pressure using UF flat sheet
membrane using distilled-deionized water
R² = 0.9924
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70
Permeate Flux (L/m
2
/hr)
Trans-membrane Pressure (psi)
68

It is important to note that the secondary clarifier effluent (SCE) contains particulates,
organics, inorganics, minerals, and biomass that can potentially cause membrane fouling, and as a
result adversely affecting membrane performances (Wijmans et al., 1984). Figure 3.7 presents the
permeate flux as a function of operation time at different TMPs for a typical SCE. As can be
observed, the initial permeate flux was proportional to the increase in the applied TMP as noted
by Acero et al. (2010). The permeate flux dropped significantly after the first 30 minutes of
operation due to gel layer formation attributable to organic, inorganic, and biological foulants. The
flux decline patterns for different TMPs were relatively similar throughout the operation time.
Figure 3.8 shows the effect of TMP on the permeate TOC concentration. The TOC concentration
in feed reservoir was 7.2 mg L
-1
and TMPs employed were 30, 45, and 60 psi. As evident in Figure
3.8, the extent of TOC removal was nearly the same at 20% in all cases, indicating that the TMP
had no influence on the overall TOC rejection.

Figure 3.7 - Effect of trans-membrane pressure on permeate flux for SCE using UF flat sheet
membrane

0
100
200
300
400
500
600
0 1 2 3 4 5 6 7 8
Permeate Flux (L/m
2
/hr)
Time (hr)
60 psi
45 psi
30 psi
69

Figure 3.8 - Effect of trans-membrane pressure on TOC concentration of SCE using UF flat sheet
membrane (TOC concentration in feed reservoir = 7.2 mg L
-1
)

Effect of PAC on Permeate Flux and TOC
Membrane filtration tests were conducted to investigate the effect of PAC on permeate flux
decline as well as the extent of TOC removal. Figure 3.9 shows that the addition of PAC at 40 mg
L
-1
lowered initial flux from 250 to 150 L m
-2
hr
-1
subsequently increased the flux by about 20%
owing to fouling mitigation. The effect of PAC on TOC removal is presented in Figure 3.10. As
can be observed, the presence of PAC lowered TOC concentration by nearly 40%.

0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8
TOC Concentration (mg/L)
Time (hr)
60 psi
45 psi
30 psi
Initial feed=7.2mg/L
70

Figure 3.9 - Effect of PAC addition on permeate flux of SCE using UF flat sheet membrane

Figure 3.10 - Effect of PAC addition on TOC concentration of SCE using UF flat sheet membrane
(TOC concentration in feed reservoir = 7.2 mg L
-1
)

Effect of E.coli on Membrane Flux and TOC
Experiments were carried out to study the effect of E.coli concentration on the permeate
flux and TOC removal. Two concentration of E.coli were employed in these tests - 10
6
and 10
8

CFU per 100 mL. Figure 3.11 shows the permeate flux decline patterns at the two E.coli
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
40 mg/L PAC addition
Trans-membrane pressure = 30 psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
40 mg/L PAC addition
Initial feed (both)=7.2mg/L
Trans-membrane pressure = 30 psi
71

concentrations. As evident, the flux decline due to the presence of E.coli was significant within
the first three hours of the experiments (50 – 30% drop). The flux decline during the first three
hours was nearly 80%. The flux decline between the two E.coli concentrations was only about 10
– 15%. It is evident that the flux decline is attributable to the formation of bacterial films on the
membrane surface and some internal pore fouling due to the presence of bioorganic compounds.
Figure 3.12 shows that the presence of E.coli had insignificant effect on TOC removals. This may
be explained by the fact that the E.coli did not make a significant contribution towards the
mineralization of the TOC content.

Figure 3.11 - Effect of E.coli addition on permeate flux of SCE using UF flat sheet membrane

0
50
100
150
200
250
300
0 1 2 3 4 5 6 7 8
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
1E+6 CFU/100mL E.coli
1E+8 CFU/100mL E.coli
Trans-membrane pressure = 30 psi
72

Figure 3.12 - Effect of E.coli on TOC concentration of SCE using UF flat sheet membrane (TOC
concentration in feed reservoir = 9.5 mg L
-1
)

Combined Effect of PAC and E.coli on Permeate Flux and TOC
Studies were performed to assess the combined effect of PAC and E.coli on permeate flux
decline and TOC removal. Figure 3.13 shows the differences in flux decline patterns between the
two systems. As can be seen, there is insignificant difference between the two flux decline patterns
during the entire period of operation, except during the first 30 – 60 minutes. Figure 3.14 presents
the effect of PAC and E.coli addition on the TOC removal. The TOC removal increased by 40%,
similar to the system described earlier (Figure 3.10). It appears that the presence of E.coli had no
significant effect on TOC removal.

0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
1E+6 CFU/100mL E.coli
1E+8 CFU/100mL E.coli
Initial feed of SCE=7.2mg/L
Feed reservoir=9.5mg/L
Trans-membrane pressure = 30 psi
73

Figure 3.13 - Effect of E.coli and PAC addition on permeate flux of SCE using UF flat sheet
membrane

Figure 3.14 - Effect of E.coli and PAC addition on TOC concentration of SCE using UF flat
sheet membrane (TOC concentration in feed reservoir = 9.5 mg L
-1
)

Effect of Ozonation and Combination of Ozone-Hydrogen Peroxide on SCE
An important aspect in membrane operations is the impact of advanced oxidation processes
(AOPs) on membrane fouling, permeate flux, and rejection characteristics. This factor is
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
PAC (40mg/L) and E.coli (1E+8CFU/100mL)
Trans-membrane pressure = 30 psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
PAC and E.coli addition
Initial feed of SCE=7.2mg/L
Feed reservoir=9.5mg/L
Trans-membrane pressure = 30 psi
74

significant because oxidants such as ozone (O3) and hydrogen peroxide (H2O2) can break down
and alter the structural characteristics of organic compounds such an endocrine disrupting
chemicals (EDCs) in water samples. Several researchers have reported the effect of oxidants on
membrane filtration performances (Langlais et al., 1991; Song et al., 2004; Song et al., 2008).
This study investigated the effect of ozonation as well as a combination of ozone and
hydrogen peroxide on the efficiency of membrane filtration. Figure 3.15 shows the permeate flux
pattern for ozonated secondary clarifier effluent. As can be observed, pre-ozonated wastewater
showed less flux decline than in the case of un-ozonated wastewater. After eight hours of operation,
the permeate flux of ozonated wastewater (63 L m
-2
hr
-1
) was nearly three times larger than that of
un-ozonated (174 L m
-2
hr
-1
). Figure 3.15 also presents the permeate flux for the combination of
ozone and hydrozen peroxide, demonstrating that the flux decline is substantially higher for the
combination than for ozone alone. It can be postulated that O3/H2O2 decomposed more organic
compounds than O3 alone, resulting in more internal pore fouling and greater flux decline. Figure
3.16 compares the TOC removal patterns for ozone and ozone-peroxide systems. It appears that
the extent of TOC removal for O3 alone exceeds that of the O3/H2O2 combination by about 10 –
15%. The reason for this difference may be attributed to the ability of O3/H2O2 to produce hydroxyl
radicals (∙OH) that can potentially react with a larger spectrum of organic molecules and
decompose them into smaller molecules that can pass through the membrane.

75

Figure 3.15 - Effect of ozone (5 mg L
-1
) and a combination of ozone (5 mg L
-1
) and hydrogen
peroxide (5 mg L
-1
) on permeate flux for UF flat sheet membrane (Temperature = 25
o
C, and time
duration = 1 hour)

Figure 3.16 - Effect of ozone (5 mg L
-1
) and a combination of ozone (5 mg L
-1
) and hydrogen
peroxide (5 mg L
-1
) on TOC concentration for UF flat sheet membrane (Temperature = 25
o
C, and
time duration = 1 hour)

0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
Preoxidized by ozone
Preoxidized by ozone and hydrogen peroxide
Trans-membrane pressure = 30 psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8
TOC Concentration (mg/L)
Time (hr)
Secondary effluent
Preoxidized by ozone
Preoxidized by ozone and hydrogen peroxide
Initial secondary effluent=7.2mg/L
Ozonated feed solution=8.5mg/L
Trans-membrane pressure = 30 psi
76

Membrane Cleaning Processes
Membrane cleaning is an essential operation for minimizing flux decline and making the
process cost-effective. Depending on the type of membrane fouling, specific cleaning strategies
should be adopted and appropriate cleaning agents be used. In general, fouled membranes are
cleaned by physical, chemical, or biochemical processes with in situ or ex situ configuration
depending on the membrane fouling characteristics (Wang et al., 2014).
In this study, organic and biological fouling were targeted because of the characteristics of
secondary clarifier effluent. Backwashing could remove non-adhesive foulants on the membrane
surfaces or pores. Organic matter or biological foulants adsorbed on membranes could be removed
by employing chemical cleaning reagents such as sodium hydroxide, or an appropriate surfactant
such as Triton X-100, or a suitable bioactive enzyme such as RID-X.
Figure 3.17 (a) shows the permeate flux using 10
-3
M sodium hydroxide (pH 11). As can
be observed, after 8 hours and 16 hours of operation, the flux recoveries were 94% and 88%,
respectively. It appears from the results that 10
-3
M NaOH could effectively remove most of the
foulants and lead to substantive flux recovery. Figure 3.17 (b) demonstrates the flux recovery after
backwashing with DDI water after 6 hours, cleaning with surfactant Triton X-100 after 12 hours,
and cleaning with enzyme RID-X after 18 hours; wherein the permeate flux were 158 L m
-2
hr
-1
,
222 L m
-2
hr
-1
, and 201 L m
-2
hr
-1
, respectively. As can be deduced, application of chemical agents
such as surfactant and enzyme could effectively remove foulants and consequently lead to
significant permeate flux recovery.
Figure 3.18 (a) presents the permeate TOC concentration profile after two cleaning cycles
(8 hours and 16 hours) using 10
-3
M NaOH. The results demonstrate that the TOC levels are not
affected by membrane cleaning. Figure 3.18 (b) shows the TOC concentration changes after
77

membrane backwashing with DDI water (6 hours), cleaning with Triton X-100 (12 hours), and
cleaning with RID-X (18 hours). Based on the above observation, it appears that all three chemicals
(NaOH, Triton X-100, and RID-X) are almost similar in terms of foulant cleaning ability and flux
recovery.

(a)

(b)

Figure 3.17 - Flux recovery and filtration after membrane cleaning processes for SCE with UF flat
sheet membrane (TMP = 30 psi): (a) 1 × 10
-3
M of sodium hydroxide; (b) backwash with DDI
water and cleaning with Triton X-100 and RID-X
0
50
100
150
200
250
300
0 2 4 6 8 10 12 14 16 18 20 22 24
Permeate Flux (L/m
2
/hr)
Time (hr)
0
50
100
150
200
250
300
0 2 4 6 8 10 12 14 16 18 20 22 24
Permeate Flux (L/m
2
/hr)
Time (hr)
10
-3
M NaOH
(Cycle 1)
10
-3
M NaOH
(Cycle 2)
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning with
5 mg/L of
RID-X
78

(a)

(b)

Figure 3.18 - TOC concentration after membrane cleaning processes for UF flat sheet membrane:
(a) 1 × 10
-3
M of sodium hydroxide at 8 and 16 hours; (b) backwash with DDI water at 6 hours,
cleaning with 5 mg L
-1
of Triton X-100 at 12 hours, and 5 mg L
-1
of RID-X at 18 hours
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24
TOC concentration (mg/L)
Time (hr)
SCE wastewater
TOC in feed = 7.9 mg/L
Trans-membrane pressure = 30 psi
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24
TOC concentration (mg/L)
Time (hr)
SCE wastewater
TOC in feed = 6.8 mg/L
Trans-membrane pressure = 30 psi
10
-3
M NaOH
(Cycle 1)
10
-3
M NaOH
(Cycle 2)
DDI
backwash
5 mg/L
Triton X-100
5 mg/L
RID-X
79

Model Predictions and Simulations
Table 2.1 (Chapter 2) lists parameters for flat sheet UF membrane and their values used in
this study. The experimental data and model predictions for permeate flux under different TMPs
are presented in Figure 3.19 (a, b, and c). As can be observed, good agreement between the
experimental data and model predictions is achieved. Figure 3.20 (a, b, and c) presents
experimental data and model predictions for permeate TOC concentrations at different TMPs
demonstrating reasonable agreement between experimental data and model profiles.
The effects of membrane cross-flow rates on permeate flux and TOC concentrations are
depicted in Figures 3.21 and 3.22, respectively. Figure 3.21 (a, b, and c) shows the experimental
data and model predictions for different cross-flow rates of 0.04, 0.4 and 1.0 gpm. As is evident, a
satisfactory agreement between experimental data and model predictions were observed. Figure
3.22 (a, b, and c) presents experimental data and model profiles for permeate TOC concentrations,
and the results further attest the predictive capability of the model.
The experimental data and model simulation/prediction for permeate flux with addition of
PAC is shown in Figure 3.23. The simulation of permeate flux with addition of PAC to the feed is
shown in Figure 3.23. It must be noted that the model incorporated an additional mass transfer
resistance term due to the formation of the PAC layer. The mass transfer resistance was estimated
by running the filtration system using DDI water and 40 mg L
-1
of PAC. As shown in Figure 3.23,
a reasonable agreement was observed between experimental data and model profiles.
Figure 3.24 presents the experimental data and predictive model profiles for permeate TOC
concentrations. The results show small deviations between the experimental data and model
profiles. Nonetheless, the model predictions are able to fit the experimental results within the
practical limits of acceptability.
80

(a) 30 psi

(b) 45 psi

(c) 60 psi
Figure 3.19 - Permeate flux pattern at different trans-membrane pressure for UF flat sheet
membrane filtration (feed TOC concentration = 7.2 mg L
-1
)
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=30psi
0
100
200
300
400
500
600
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=45psi
0
100
200
300
400
500
600
700
800
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=60psi
model profile
model profile
model profile
81

(a) 30 psi

(b) 45 psi

(c) 60 psi
Figure 3.20 - TOC concentration profiles at different trans-membrane pressure for UF flat sheet
membrane filtration (feed TOC concentration = 7.2 mg L
-1
)
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10 TOC Concentration (mg/L)
Time (hr)
Experimental data
Predicted model profile
Feed reservoir=7.2mg/L
Trans-membrane pressure=30psi
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10
TOC concentration (mg/L)
Time (hr)
Experimental data
Predicted model profile
Feed reservoir=7.2mg/L
Trans-membrane pressure=45psi
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10
TOC concentration (mg/L)
Time (hr)
Experimental data
Predicted model profile
Feed reservoir=7.2mg/L
Trans-membrane pressure=60psi
model profile
model profile
model profile
feed TOC
concentration
feed TOC
concentration
feed TOC
concentration
82

(a) 0.04 gpm

(b) 0.4 gpm

(c) 1.0 gpm
Figure 3.21 - Effect of cross-flow rate on permeate flux for UF flat sheet membrane filtration (feed
TOC concentration = 7.2 mg L
-1
)
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=30psi
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=30psi
Cross-flow rate=0.4 gpm
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=30psi
model profile
model profile
model profile
83

(a) 0.04 gpm

(b) 0.4 gpm

(c) 1.0 gpm
Figure 3.22 - Effect of cross-flow rate on TOC concentration for UF flat sheet membrane filtration
(feed TOC concentration = 7.2 mg L
-1
)
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10
TOC concentration (mg/L)
Time (hr)
Experimental data
Predicted model profile
Cross-flow rate=0.04gpm
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10
TOC concentration (mg/L)
Time (hr)
Experimental data
Predicted model profile
Cross-flow rate=0.4gpm
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10
TOC concentration (mg/L)
Time (hr)
Experimental data
Predicted model profile
Cross-flow rate=1.0gpm
model profile
model profile
model profile
feed TOC
concentration
feed TOC
concentration
feed TOC
concentration
84

Figure 3.23 - Effect of PAC addition on permeate flux for UF flat sheet membrane filtration

Figure 3.24 - Effect of PAC addition on TOC concentration for UF flat sheet membrane filtration
(trans-membrane pressure = 30 psi, and feed TOC concentration = 7.2 mg L
-1
)

Model Sensitivity Analysis
Figure 3.25 shows the permeate flux predictions for the flat sheet UF membrane with
variations in parameters including diffusion coefficient (D), and gel layer concentration (Cg).
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8 9 10
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data (without PAC)
Predcited model profile (without PAC)
Experimental data (40 mg/L PAC)
Predicted model profile (40mg/L PAC)
Trans-membrane pressure=30psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10
TOC Concentration (mg/L)
Time (hr)
Experimental data (without PAC)
Predicted model profile (without PAC)
Experimental data (40mg/L PAC)
Predicted model profile (40mg/L PAC)
Feed reservoir=7.2mg/L
Trans-membrane pressure=30psi
model profile
model profile
feed TOC
concentration
85

Diffusion coefficient is a significant parameter because it directly affects mass transfer coefficient
parameter and the permeate flux profiles. As can be observed in Figure 3.25 (a), permeate flux was
slightly influenced by diffusion coefficient. In Figure 3.25 (b), the extent of gel layer formation
affects the permeate flux profiles because increasing the concentration of gel layer would increase
the resistance of membrane flux.

(a)

(b)

Figure 3.25 - Experimental data and model profiles for model sensitivity analysis for flat sheet UF
membrane: (a) Effect of diffusion coefficient; (b) effect of gel layer
0
50
100
150
200
250
300
350
400
450
0 1 2 3 4 5 6 7 8 9 10
Flux (L/m
2
/hr)
Time (hr)
+50% of D
D
-50% of D
Experiment
0
50
100
150
200
250
300
350
400
450
0 1 2 3 4 5 6 7 8 9 10
Flux (L/m
2
/hr)
Time (hr)
+50% of Cg
Cg
-50% of Cg
Experiment
Model Profile
Experimental data
Experimental data
Model Profile
86

3.6 Summary and Conclusions
Flat sheet membranes have been widely used in membrane filtration tests because they
afford rapid testability, have easy implementability and scalability at pilot-scale or full-scale levels
and maintain cost effectiveness.
In this chapter, UF flat sheet membrane tests were conducted to evaluate the permeate flux
patterns, TOC removals, and the effectiveness of PAC and/or microorganisms in the overall
membrane performances for secondary clarifier effluent. The TMP for UF was 30 psi for all the
tests. The initial permeate flux was approximately 250 L m
-2
L
-1
and it gradually increased as the
TMP was increased. The TOC removal was nearly 20% for an influent TOC of 7.2 mg L
-1
. The
addition of PAC increased the TOC removal to approximately 40%. The combination of PAC and
microorganisms was expected to increase the TOC removal due to the synergistic effects of
adsorption and biodegradation. Nonetheless, the synergism towards enhanced TOC removal was
not witnessed owing to the short contact time between the activated carbon particles and the
microorganisms.
According to the results of flat-sheet UF membrane filtration, the produced water could be
used for certain reuse applications such as agricultural irrigation. The permeate from this study
presumably contained sufficient amounts of organic matter and minerals that would be used as
carbon sources or nutrients for the growth of crops or the storage in soils.
The following section discusses the different water reuse applications employing
nanofiltration membrane filtration.

87

3.7 Limited Study of Nanofiltration Membrane for Water Reuse Application
Permeate Flux Pattern and TOC Removal
Filtration tests for flat sheet NF membranes were conducted using the UF permeate as feed.
Figure 3.26 shows permeate flux pattern as a function of time for a TMP of 70 psi. The flux was
initially 58.1 L m
-2
hr
-1
and it gradually dropped to 44.5 L m
-2
hr
-1
after 12 hours, corresponding
to a flux reduction of 23.4%. This flux decline was caused by dissolved organic carbons (DOCs)
because the feed solution was obtained a permeate of a UF process.
It is important to note that the pore size of UF used in this study was 4 - 10 nm and a
MWCO of 10,000 Daltons. Logan et al. (1990) and Shon et al. (2006) reported that most of the
DOC in wastewater consisted of organic matter of lower molecular weight than 1,000 Daltons,
and were sizes smaller than about 0.8 nm. Thus, the DOC in the wastewater cannot be retained by
UF membranes, but can be partly or mostly retained by NF membranes according to the size
exclusion principle. Figure 3.27 shows the TOC concentrations of the feed and the permeate. The
initial TOC concentration in the feed was 5.05 mg L
-1
and it increased slightly to 6.3 mg L
-1
over
12 hours of operation in a batch reactor system. Meanwhile, the permeate TOC that was initially
2.8 mg L
-1
, dropped to 2.2 mg L
-1
after 12 hours. The results showed that the NF90 membrane
could effectively remove organic matter (TOC) to the extent of 57%.
88

Figure 3.26 - Permeate flux pattern for NF flat sheet membranes

Figure 3.27 - TOC concentration for NF flat sheet membranes

Effect of Powdered Activated Carbon Addition on Flux and TOC Removal
Activated carbon is widely used in water and wastewater treatment applications to remove
organic compounds and potential pollutants including endocrine disrupting chemicals (EDCs)
and/or pharmaceuticals and personal care products (PPCPs). These pollutants cannot be generally
removed by conventional treatment processes, but can be effectively removed only by the
0
10
20
30
40
50
60
70
0 1 2 3 4 5 6 7 8 9 10 11 12
Flux (L/m
2
/hr)
Time (hr)
NF permeate using UF permeate as feed
Trans-membrane pressure=70 psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC (mg/L)
Time (hr)
Feed characteristics
NF permeate
Initial feed=5.1mg/L
Trans-membrane pressure=70 psi
89

advanced treatment technologies. An earlier experimental study by Solak et al (2013) had
established that activated carbon could effectively remove some types of EDCs to the extent of
88% to 93%, even at trace concentration levels.
The present study investigated the feasibility of nanofiltration for the removal of potential
contaminants from SCE. It evaluated the TOC removal and membrane flux using NF membranes.
Figure 3.28 shows the permeate flux patterns in the presence and absence of PAC (the
concentration of PAC used was of 15 mg L
-1
). As can be observed, the initial fluxes were similar
to each other, and were approximately 60 L m
-2
hr
-1
. The permeate fluxes in the absence of PAC
were slightly higher than those experienced in the presence of PAC during the first 6 hours of
operation. However, after this period, the permeate fluxes exhibited a different patterns, as they
were relatively higher in the presence of PAC. These changes in permeate flux patterns indicated
that after the initial adsorption of organic carbon, the membrane fouling was reduced and the
overall aqueous permeability increased.
Figure 3.29 shows the effect of PAC on the TOC concentration in permeate. Evidently, the
PAC successfully removed the TOC from the treated wastewater leaving a residual of only 0.2 mg
L
-1
. These results demonstrate the potential of PAC use for the removal of a broad spectrum of
organic contaminants.

90

Figure 3.28 - Effect of 15 mg L
-1
PAC addition on NF permeate flux pattern

Figure 3.29 - Effect of 15 mg L
-1
PAC addition on NF permeate TOC removal

Reuse of wastewater is defined generally as the direct use of treated wastewater for
irrigation, groundwater recharge, toilet flushing, and other purposes, except discharge to surface
water bodies such as rivers and oceans, because it is considered as an abandonment of water rights.
0
10
20
30
40
50
60
70
0 1 2 3 4 5 6 7 8 9 10 11 12
Flux (L/m
2
/hr)
Time (hr)
NF permeate using UF permeate as feed
NF permeate using UF permeate as feed with PAC
Trans-membrane pressure=70psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC (mg/L)
Time (hr)
UF permeate TOC=5.05mg/L
NF permeate using UF permeate as feed with PAC
NF permeate using UF permeate as feed
Trans-membrane pressure=70 psi
91

A general public health concern is that the water reuse applications must meet the water quality
standards for all specified contaminants including TOC and microorganisms. Table 3.3 shows the
water reuse applications for different advanced wastewater treatment configurations. In this regard
there are four classifications that are considered (Figure 3.30): (1) filtration of SCE with UF
membrane alone, (2) filtration of SCE with UF membrane and PAC addition, (3) filtration of UF
permeate through an NF membrane (UF followed by NF), and (4) filtration of UF permeate with
NF membrane and PAC addition (UF followed by NF with PAC). The use of recycled water for
irrigation is described in Section 60304, Titles 17 and 22, California Code of Regulations
(California Department of Public Health, 2014). The water can be used for agricultural irrigation
for the cultivation of food crops and edible root crops, urban irrigation of land such as school yards,
residential landscaping, and leisure irrigation of parks, playgrounds, and unrestricted golf courses.
The use of recycled water for impoundment, and industrial as well as commercial cooling are
stated in Sections 60305 and 60306 of the above regulations. The other applications of water reuse
are listed in Section 60307 of these regulations and they include the following: flushing toilets,
priming drain traps, fire fighting of structures, decorative fountains, commercial laundries,
artificial snow for commercial outdoor use, commercial car washes, industrial boiler fee, soil
compaction, mixing concrete, dust control on roads and streets, and industrial process water. The
combination of nanofiltration membrane and PAC addition can be employed for groundwater
replenishment applications under Section 60320.

92

Table 3.3 - Reuse of wastewater applications on a basis of different advanced wastewater treatment
Configuration
Steady
state flux
(L m
-2
hr
-1
)
Steady
state TOC*
(mg L
-1
)
Steady state
anion
concentrations**
(mg L
-1
)
References Applications***
UF 63 5.4
Cl
-
: 165
NO3
-
-N: 3.0
PO4
3-
-P: 0.3
SO4
2-
: 140
Fig. 3-10
Table 3.1
- Agricultural reuse such
as food crops
commercially processed
- Urban reuse including
landscape irrigation, golf
courses, parks, vehicle
washing, toilet flushing
- Recreational
impoundments
- Construction and
Industrial use
UF+PAC 80 3.0
Cl
-
: 80
NO3
-
-N: 1.5
PO4
3-
-P: 0.1
SO4
2-
: 60
Fig. 3-10
UF-NF 45 2.2
Cl
-
: 15.6
NO3
-
-N: 1.27
PO4
3-
-P: 0
SO4
2-
: 1.12
Fig. 3-26
Table 5.3
UF-NF+PAC 50 0.2
Cl
-
: 6.84
NO3
-
-N: 0.66
PO4
3-
-P: 0
SO4
2-
: 0.44
Fig. 3-28
Table 5.3
Groundwater recharge
and indirect potable
reuse
Note: Microorganisms were not detected in all cases, that is, 0 CFU mL
-1
.
* Initial TOC concentrations were 5.1, 6.7, and 7.2 mg L
-1
.
** Initial concentrations were 170 mg L
-1
of Cl
-
, 3 mg L
-1
of NO3
-
-N, 0.5 mg L
-1
of PO4
3-
-P, and
140 mg L
-1
of SO4
2-

*** California Department of Public Health (2014) and US EPA (2012) – see the following two
references:
1. California Department of Public Health (2014). Title 17 and 22 California Code of Regulations.
Retrieved from http://www.waterboards.ca.gov/drinking_water/certlic/drinkingwater/documents/
lawbook/RW regulations_20140618.pdf.
2. US EPA (2012). Guidelines for Water Reuse. EPA/600/R-12/618. Retrieved from
http://nepis.epa.gov/ Adobe/PDF/P100FS7K.pdf

93

Figure 3.30 - Actual and proposed water reclamation schemes
Primary Clarifier
Secondary
Clarifier
Biological
Treatment
Conventional
wastewater
treatment
Sand Filter Chlorination Dechlorination
Typical
water reclamation
treatment
Sand Filter Microfiltration
Reverse
Osmosis
Sand Filter Microfiltration
Reverse
Osmosis
UV light
Advanced
water reclamation
treatment
in Los Angles
Proposed
water reclamation
schemes
(see Table 3.3)
Ultrafiltration
Advanced
water reclamation
treatment
in Orange County
Ultrafiltration and PAC
Ultrafiltration Nanofiltration
Ultrafiltration Nanofiltration and PAC
Chlorination Dechlorination
Chlorination Dechlorination
94

3.8 References
Acero, J. L., Benitez, F. J., Leal, A. I., Real, F. J., & Teva, F. (2010). Membrane filtration
technologies applied to municipal secondary effluents for potential reuse. Journal of Hazardous
Materials, 177, 390-398.

Andreozzi, R., Caprio, V., Insola, A., & Marotta, R. (1999). Advanced oxidation processes (AOP)
for water purification and recovery. Catalysis Today, 53, 51-59.

APHA, AWWA, & WEF (2005). Standard Methods for the Examination of Water and Wastewater,
21st ed., American Public Health Association, Washington, D.C.

Arkhangelsky, E. & Gitis, V. (2008). Effect of transmembrane pressure on rejection of viruses by
ultrafiltration membranes. Separation and Purification Technology, 62, 619-628.

Ates, N., Yilmaz, L., Kitis, M., & Yetis, U. (2009). Removal of disinfection by-product precursors
by UF and NF membranes in low-SUVA waters. Journal of Membrane Science, 328, 104-112.

California Department of Public Health (2014). Title 17 and 22 California Code of Regulations.
Retrieved from http://www.waterboards.ca.gov/drinking_water/certlic/drinkingwater/documents/
lawbook/RW regulations_20140618.pdf.

Campos, C., Mariñas, B. J., Snoeyink, V. L., Baudin, I., & Lainé, J. M. (2000). PAC-membrane
filtration process. II: Model application. Journal of Environmental Engineering, 126, 104-111.

Chiang, P. C., Chang, E. E., Chang, P. C., & Huang, C. P. (2009). Effects of pre-ozonation on the
removal of THM precursors by coagulation. Science of the Total Environment, 407, 5735-5742.

Chung, H. K., Kim, W. H., Park, J., Cho, J., Jeong, T. Y., & Park, P. K. (2015). Application of
Langmuir and Freundlich isotherms to predict adsorbate removal efficiency or required amount of
adsorbent. Journal of Industrial and Engineering Chemistry, 28, 241-246.

95

Daniş, Ü., & Keskinler, B. (2009). Chromate removal from wastewater using micellar enhanced
crossflow filtration: Effect of transmembrane pressure and crossflow velocity. Desalination,
249(3), 1356-1364.

Dang, H. T., Nnarbaitz, R. M., & Matsuura, T. (2010). Evaluation of apparatus for membrane
cleaning tests. Journal of Environmental Engineering, 136, 1161-1170.

Foo, K. Y. & Hameed, B. H. (2010). Insights into the modeling of adsorption isotherm systems.
Chemical Engineering Journal, 156, 2-10.

Gao, Y. & Deshusses, M. A. (2011). Adsorption of clofibric acid and ketoprofen onto powdered
activated carbon: Effect of natural organic matter. Environmental Technology, 32(15), 1719-1727.

Halsey, G. (1948). Physical adsorption on non-uniform surfaces. Journal of Chemical Physics,
6(10), 931-937.

Jin, G., Englande, A. J., Bradford, H., & Jeng, H. W. (2004). Comparison of E.coli, Enterococci,
and fecal coliform as indicators for brackish water quality assessment. Water Environment
Research, 76(3), 245-255.

Kwon, B., Cho, J., Park, N., & Pellegrino, J. (2006). Organic nanocolloid fouling in UF membranes.
Journal of Membrane Science, 279(1-2), 209-219.

Langlais, B., Reckhow, D. A., & Brink, D. R., AWWA Research Foundation., & Compagnie
générale des eaux (Paris, France). (1991). Ozone in water treatment: Application and engineering:
cooperative research report. Chelsea, Mich: Lewis Publishers.

Langmuire, I. (1918). The adsorption of gases on plane surfaces of glass, mica, and platinum,
Journal of the American Chemical Society, 40, 1361-1402.

96

Lehman, S. G. & Liu, L. (2009). Application of ceramic membranes with pre-ozonation for
treatment of secondary wastewater effluent. Water Research, 43, 2020-2028.

Logan, B. E. & Jiang, Q. (1990). Molecular size distributions of dissolved organic matter, Journal
of Environmental Engineering, 116, 1046-1062.

Maartens, A., Swar , P. & Jacobs, E. P. (1996). Characterisation techniques for organic foulants
adsorbed onto flat-sheet UF membranes used in abattoir effluent. Journal of Membrane Science,
119, 1-8.

Mosqueda-Jimenez, D. B., Narbaitz, R. M. & Matsuura, T. (2004). Membrane fouling test:
Apparatus evaluation, Journal of Environmental Engineering, 130, 90-99.

Motlagh, A. M., Pant, S., & Gruden, C. (2013). The impact of cell metabolic activity on biofilm
formation and flux decline during cross-flow filtration of ultrafiltration membranes. Desalination,
316, 85-90.

MWH, Water Treatment: Principles and Design, 2nd ed., John Wiley and Sons, 2005.

Pezeshk, N., Rana, D., Narbaitz, R.M, & Matsuura, T. (2012). Novel modified PVDF ultrafiltration
flat-sheet membranes. Journal of Membrane Science, 389, 280-286.

Rudie, B. J., Ross, G. S., Harrold, S. J., & Paulson, D. J. (1993). Effects of surface force interations
on an NF/UF membrane. Desalination, 90, 107-118.

San Jose Creek Water Reclamation Plant. (n.d.). Retrieved May 1, 2015, from
http://www.lacsd.org/wastewater/wwfacilities/joint_outfall_system_wrp/san_jose_creek.asp

Shon, H. K., Kim, S. H., Erdei, L. & Vigneswaran, S. (2006). Analytical methods of size
distribution for organic matter in water and wastewater. Korean Journal of Chemical Engineering,
23(4), 581-591.
97

Solak, S., Vakondios, N., Tzatzimaki, I., Diamadopoulos, E., Arda, M. Kabay, N. & Yuksel, M.
(2013). A comparative study of removal of endocrine disrupting compounds (EDCs) from treated
wastewater using highly crosslinked polymeric adsorbents and activated carbon. Journal of
Chemical Technology and Biotechnology, 89(6), 819-824.

Song, W., Ravindran, V., Koel, B. E., & Pirbazari, M (2004). Nanofiltration of natural organic
matter with H2O2/UV pretreatment: fouling mitigation and membrane surface characterization,
Journal of Membrane Science, 241, 143-160.

Song, W., Ravindran, V., & Pirbazari, M. (2008). Process optimization using a kinetic model for
the ultraviolet radiation-hydrogen peroxide decomposition of natural and synthetic organic
compounds in groundwater. Chemical Engineering Science, 63, 3249-3270.

Tansel, B., Sager, J., Garland, J., & Xu, S. (2009). Effect of transmembrane pressure on overall
membrane resistance during cross-flow filtration of solutions with high-ionic content. Journal of
Membrane Science, 328(1-2), 205-210

Ulbricht, M. (2006). Advanced functional polymer membranes. Polymer, 47, 2217-2262.

US EPA (2012). Guidelines for Water Reuse. EPA/600/R-12/618. Retrieved from
http://nepis.epa.gov/ Adobe/PDF/P100FS7K.pdf

Volk, C. J. & LeChevallier, M. W. (2000). Assessing biodegradable organic matter. Journal
(American Water Works Association), 92(5), Disinfection, 64-76.

Wang, Z., Ma, J., Tang, C. Y., Kimura, K., Wang, Q. & Han, X. (2014). Membrane cleaning in
membrane bioreactors: A review. Journal of Membrane Science, 468, 279-307.

Wijmans, J. G., Nakao, S., & Smolders, C. A. (1984). Flux limitations in ultrafiltration: osmotic
pressure model and gel layer model. Journal of Membrane Science, 20, 115-124.

98

Wijmans, J. G., Nakao, S., Van Den Berg, J. W. A., Troelstra, F. R., & Smolders, C. A. (1985).
Hydrodynamic resistance of concentration polarization boundary layers in ultrafiltration. Journal
of Membrane Science, 22, 117-135.

Zhou, L. X. & Wong, J. W. C. (2000). Organic compounds in the environment. Journal of
Environmental Quality, 29(6), 1852-1856.

99

Chapter 4
Investigating the Potential of Hollow Fiber Ultrafiltration
Membranes for Water Reclamation and Reuse
4.1 Introduction
Among various technologies, membrane bioreactor (MBR) processes have shown
excellent potential for water reclamation, water reuse, groundwater recharge, and similar
applications. Superior membranes with better aqueous transport and anti-fouling characteristics
can make the technology more efficient and economical. In fact, they can significantly reduce
energy costs that constitute a substantial fraction of total operation costs. The continuous flow
hybrid MBR system discussed in this study offers several technical advantages over conventional
biological processes in environmental applications: small size or footprint requirements for reactor
systems, long solids retention times, and efficient retention of particulates, colloids, contaminants
and microorganisms. The treated effluent from such hybrid MBR system processes must be low
in suspended solids, biochemical oxygen demand (BOD), chemical oxygen demand (COD), total
organic carbon (TOC) and most pathogens, so as to meet groundwater recharge requirements. The
MBR process could be made more compact, efficient and economical with the advent of modern,
computerized, and programmable process control systems. The application of an adsorbent such
as powder activated carbon (PAC) will defoul the membranes and maintain high permeate fluxes
(Pirbazari et al, 1996; Williams and Pirbazari, 2007; Williams et al., 2012). An added advantage
will be that the adsorbent has the ability to remove most trace-level or residual micro-pollutants
including endocrine disrupting chemicals (EDCs) such as pharmaceutical and personal care
products (PPCPs), pesticides and solvents during water reclamation (Snyder et al., 2007).
100

Water reclamation for reuse and groundwater recharge requires that tertiary treatment
standards for various components be met, that can be briefly summarized as follows: (i) virus
removal or inactivation must exceed 5 logs; (ii) total coliform levels must be below 2.2 coli per
100 mL; (iii) turbidity must be below 2 NTU; (iv) total organic carbon concentrations must not
exceed 0.5 mg/L (of wastewater origin); (v) organic and inorganic contaminants must meet
drinking water maximum contaminant levels (MCLs) (specified by the United States
Environmental Protection Agency and/or the State Water Quality Control Board); (vi) lead and
copper concentrations must conform to actions levels, and (vii) nitrate limits are based on the
treatment technology. These standards could be achieved with high efficiency and favorable
economics by the MBR technology using superior ultrafiltration membranes.
Ultrafiltration (UF) membranes are designed for clarification and disinfection processes in
water and wastewater treatment. These membranes have pore sizes in the range of 0.01 to 0.1
micrometer (µm) and molecular weight cutoff (MWCO) range of 10,000 to 100,000 Daltons (Da).
Since the osmotic backpressure can be neglected, UF membrane systems can be generally operated
at relatively low trans-membrane pressures ranging from 200 to 700 kPa (30 and 100 psi).
Hollow-fiber UF membranes have been used in various experimental scales to evaluate
performances and feasibilities of membrane operations. In this chapter, two different scales of tests
such as micro-scale and mini-pilot-scale were employed. Micro-scale hollow-fiber membrane
filtration experiments provide not only preliminary results for flux pattern and TOC removal but
also estimates of parameters required for mathematical models employed for performance
prediction. The data from micro-scale tests would be used to predict the dynamics of mini-pilot-
scale membrane process systems. The mini-pilot-scale setup is a prelude to a pilot-scale membrane
bioreactor (MBR) system used to investigate process feasibility and upscaling.
101

These MBR processes typically refer to the membrane filtrations used in conjunction with
biological systems in wastewater treatment facilities. Fane and coworkers (1980) employed MBR
technology to wastewater treatment applications in Australia. These researchers combined
activated sludge process with an ultrafiltration unit and were able to obtain an effluent of high
quality by membrane exclusion of suspended solids including silt, clay and microorganisms.
Chang et al. (1993) employed a similar technology using a bioreactor and an ultrafiltration module
for the denitrification of drinking water. Cicek et al. (1998) employed a pilot-scale version in the
reclamation of municipal wastewater for non-potable uses. Nowadays, MBRs also employ
submerged membrane configuration for applications such as municipal wastewater treatment
(Rosenberger et al., 2002; Witzig et al., 2002). The use of submerged membranes has certain
advantages including a significant reduction in power consumption, a factor that has increased the
potential for membranes in wastewater treatment. However, submerged membrane configuration
also has certain disadvantages that offset the advantages of energy saving, and they include severe
membrane fouling, frequent membrane backwashing and cleaning requirements, and exposure to
material damage. The relative merits and demerits of submerged membrane reactors have been
discussed elsewhere (Rosenberger et al., 2002; Witzig et al., 2002). The MBR system can
effectively replace conventional activated sludge treatment systems with microbial growth in
wastewater treatment and water reclamation applications (Ravindran et al., 2009).

Strategies for membrane fouling control
The factors influencing membrane fouling include biomass, colloids, natural organics,
inorganic precipitates or scalants, and extracellular polymers; and these factors are dependent upon
operating conditions. Membrane fouling is attributed to the following causes: (i) macromolecular
102

and colloid sorption, (ii) biofilm growth and attachment, and (iii) inorganic matter
precipitation or scaling (Tsai et al., 2004; Williams and Pirbazari, 2007; Ravindran et al.,
2009). Membrane fouling can be strategically controlled by concentration polarization
suppression, optimization of physical and chemical cleaning protocols, and pre-treatment of feed
(Tu et al., 2005; Williams and Pirbazari, 2007).
Fouling in MBR systems and other integrated membrane processes will be caused by
foulant sorption onto and within membrane pores, and deposition of cake or gel layer on the
membrane surface. Biological fouling is predominantly caused by extracellular polymeric
substances (EPS) that mainly consist of carbohydrates, proteins, humic substances and nucleic
acids, constituting the infra-structure for bacterial floc formation and biofilm adhesion (Williams
and Pirbazari, 2007; Ravindran et al., 2009). Permeate flux decline due to concentration
polarization and membrane fouling can be mitigated by employing PAC adsorbent and fluid
management (Williams and Pirbazari, 2007; Ravindran et al., 2009; Williams et al., 2012). This is
achieved by using structural features such as baffles and vortex generators to promote local
turbulence, eddies, vortices and instabilities inside the membrane shell and the membrane module,
and suppresses boundary layer and concentration polarization effects. The PAC depolarizes
dissolved biological and organic matter and re-entrains colloids and suspended solids from the
viscous sub-layer. The PAC adsorbs most organic and bio-organic foulants in wastewaters and
reclaimed waters, including humic substances, proteins, carbohydrates and fats (Kilduff and
Weber, 1992). The PAC particles also reduce thicknesses of mass-transfer and hydrodynamic
boundary layers, minimize concentration polarization, and control gel deposition on membrane
surfaces and pores (Pirbazari et al., 1996; Tsai et al., 2004; Ravindran et al., 2009).

103

The significance of single hollow fiber membrane filtration experiments
Hollow-fiber membranes are commonly used in membrane bioreactor (MBR) systems
because a single module of these membranes has relatively large surface area than a module of
flat- sheet membranes used in plate-and-frame or spiral-wound configurations. The larger surface
areas shall provide more permeate flux and more biofilm formation for biological treatment.
The MBR process is a relatively complicated system to design for water and wastewater
applications because of many dependent factors such as permeability, biomass loading, organic
loading, hydraulic retention time, biomass retention time, biokinetic parameters, cost effectiveness,
and energy consumption. These factors may influence and depend on each other, and therefore
should be carefully evaluated by simple experimental design in the form of bench-scale or
laboratory-scale systems that will greatly economize the consumption of time, costs, energy and
chemicals.
This chapter includes the results of micro-scale and mini-pilot-scale membrane
experiments. The study of micro-scale UF hollow fiber membranes was designed for the following
purpose:
i: Investigate the permeate flux and TOC removal
ii: Determine the effect of PAC on permeate flux and TOC
iii: Determine the effect of E.coli on permeate flux and TOC
iv: Determine the effect of combination of PAC and E.coli on permeate flux and TOC
The study of mini-pilot-scale UF hollow-fiber membranes intended to achieve the following
purpose:
i: Investigate the effect of trans-membrane pressure on permeate flux
ii: Determine the effect of PAC on permeate flux and TOC
104

iii: Determine the effect of E.coli on permeate flux and TOC
iv: Investigate the effect of combination of PAC and E.coli on permeate flux and TOC
v: Examine the effect of ozonation on permeate flux and TOC
vi: Predict the performance of the MBR system under different scenarios regarding TOC
removal using mathematical modeling.

4.2 Materials and Methods
Materials:
Membrane
Polyethersulfone (PES) UF hollow-fiber membranes provided by Matrix Membranes was
used for the membrane bioreactor (MBR) research. These hollow-fiber membranes were
characterized by a pore size of 4 - 10 nm and a MWCO of 10,000 Daltons. These membranes were
of an asymmetric structure and were moderately hydrophobic. The effective membrane surface
areas of micro-scale and mini-pilot-scale hollow-fibers were 9.42 cm
2
and 3768 cm
2
, respectively
(1 inch
2
= 6.45cm
2
).

Feed Solution
A biologically treated wastewater was used in this research. Secondary clarifier effluent
(SCE) was obtained from the San Jose Creek Water Reclamation Plant (SJCWRP) in Los Angeles
County (Figure 3.1). The SCE was the effluent of a secondary clarifier of the treatment plant prior
to filtration and chlorination. The samples of SCE were filtered by using a 5 µm pore filter paper
to remove the larger particulate matter, but without excluding the bacterial population. The
characteristics of wastewater are presented in Table 3.1. The analysis of the wastewater was
periodically performed in accordance with EPA standard methods (See Chapter 3).
105

Microorganism
Membrane processes can achieve the removal of various bacterial species including
coliforms. Escherichia coli (E.coli) type ATCC 8739, cultured by BioMérieux Inc. (Durham, NC),
is widely used in biological studies because of several advantages: the analysis of E.coli is
relatively easier than pathogens in the laboratory, and the generation time is relatively shorter than
for other bacteria. The E.coli is used as indicator organisms for pathogenic bacteria in the realm of
water and wastewater treatment. The E.coli was cultivated in a bio-laboratory for two days to reach
a steady phase. Appropriate amounts of cultivated E.coli were transferred into the feed reservoir
and completely mixed.

Adsorbent
Activated carbon has been widely used as adsorbent for water and wastewater treatment
applications because of its effectiveness in the removal of a broad spectrum of compound including
natural organic matter. Powdered activated carbon (PAC) was used as adsorbent and in this study,
WPH-M PAC manufactured by Calgon Carbon Corporation (Pittsburgh, PA) was employed.

Oxidants
Ozone is widely used oxidant in water and wastewater applications. Ozone typically reacts
with unsaturated double bonds and hence organic compounds may not be completely mineralized.
The ozone gas used in this study was produced by a bench-scale ozone generator, (LAB2B;
Triogen, UK).

106

Chemical Cleaning Agents
Three types of chemical cleaning agents, namely, a caustic solution, a surfactant solution,
and an enzyme solution, were used in these experiments. These included the following: sodium
hydroxide (ACS reagent grade, BDH; distributed by VWR International, PA) that was used for
caustic cleaning; Triton X-100 (laboratory grade, Dow Chemical, MI) and RID-X (Reckitt
Benckiser, NJ) enzyme.

Methods:
Membrane Filtration
The micro-scale membrane filtration system and the mini-pilot-scale continuous flow
membrane bioreactor (MBR) system are presented in Figure 4.1 and Figure 4.2, respectively. The
purpose of the MBR system (designed with membrane module operation) is to evaluate membrane
performance factors including membrane fouling, membrane cleaning, and productivity. The
system is composed of the hollow-fiber membrane module, a reactor with a heat exchanger, a feed
reservoir, permeate collector, an air blower to supply oxygen and mix the reactor, a coarse diffuser
bar, a pressure gauge, a flow meter, a sampling port, and pumps including that used for
recirculation pump.
Different types of feed waters were used in this study, namely: (1) effluent of secondary
clarifier (SCE), (2) SCE with addition of 40 mg L
-1
of PAC, (3) SCE with addition E.coli (10
8
CFU
per 100 mL), (4) SCE with combination of PAC and E.coli, (5) ozonated SCE. The operating
pressure was examined by using distilled deionized (DDI) water as feed so as to determine the
nominal permeate rate. A recirculation ratio of 20 was used in these operations. Trans-membrane
pressures (TMP) for micro-scale and mini-pilot-scale were maintained at 20 psi and 4 psi,
107

respectively (1 bar = 14.5 psi). Coarse bubbles were constantly supplied during the experiments.
The permeate flow rate was periodically measured and the permeate was sampled for the
determination of its composition including measurements of TOC and UV absorbance.

Figure 4.1 - Schematic of micro-scale hollow fiber membrane filtration unit

Figure 4.2 - Schematic of mini-pilot hollow fiber membrane filtration
108

Membrane Backwash and Chemical Cleaning
The backwash process was conducted only for the mini-pilot-scale membrane filtration
system. The fouled membranes were backwashed with DDI water using an in situ method wherein
the flow direction was reversed and a TMP of 8 psi was applied for 1 hour. Chemical cleaning was
also performed using a similar method using a cleaning agent for 2 hours followed by rinsing with
DDI water for 2 to 4 hours until the residuals of chemical agents were completely removed.

Analytical Methods
Total Organic Carbon: The total organic carbon (TOC) was measured by a non-purgeable
organic carbon (NPOC) method. The samples were acidified with phosphoric acid and purged with
carbon free air to remove inorganic carbon species. The TOC was determined using a Shimadzu
TOC-V CSH analyzer (Shimadzu Corporation, Kyoto, Japan), analyzer in accordance with the
analytical procedure specified in EPA 415.1, and Standard Methods 5310B (APHA et al., 2005).

UV Absorbance: The UV absorbance values indirectly indicate the amount of total organic matter
present in samples. It was determined by using Lambda 35 UV/Vis Spectrophotometers
(PerkinElmer, Waltham, MA) set at a wavelength of 254 nm.

4.3 Biokinetic Studies
Biokinetic studies are generally conducted to obtain a good understanding of biological
processes, and their performance levels to investigate the optimal operational conditions for a
process such as a membrane bioreactor (MBR) system. These studies are usually used to estimate
109

the biokinetic parameters including the specific growth rate ( μm), Monod half saturation coefficient
(Ks), yield coefficient (Y), and decay coefficient (kd).

4.3.1 Biokinetic Reactor Studies
The biokinetic system employed in this study consists of a glass reactor of capacity of two
liters. The glass reactor was securely fixed to a frame by a stainless steel holder, as illustrated in
Figure 4.3. The reactor vessel was sealed gas-tight to prevent the entry of oxygen if anoxic or
anaerobic conditions were required. The feed solution containing the carbon source with nutrients
were injected into the reactor by a peristaltic pump. The reactor contained microorganisms and
secondary clarifier effluent. Temperature was maintained at 25
o
C. The pH and the dissolved
oxygen (DO) concentration were continuously monitored by electrode probes that were linked to
controllers of the biokinetic reactors. The pH controller was connected to pumps that could inject
either HCl or NaOH solutions into the system to maintain the pH at the required set point. The
reactor was equipped with a variable speed agitator to provide the mixing necessary to facilitate
the biochemical reactions. Samples were periodically withdrawn from the reactor for determining
carbon source concentrations and biomass.
110

Figure 4.3 - Schematic of experimental setup for biokinetic studies

4.3.2 Estimation of Biokinetic Parameters
Biokinetic experiments are generally employed to determine biological substrate reduction
rates and to estimate the associated biokinetic parameters including the maximum specific growth
rate ( μmax), Monod half saturation coefficient (Ks), cell yield coefficient (Y), and the specific death
rate (kd). The fundamental mass-balance equations describing the Monod biokinetic dynamics are
formulated in a time-variation format (Monod, 1949).
The mass balance for the biomass (X) and substrate (S) for carbon source limiting
conditions can be expressed as follows:
𝑉 𝑑𝑋 𝑑𝑡 = 𝑄 𝑋 0
− 𝑄𝑋 + 𝑉 𝑟 𝑔
Eq. (1)

𝑉 𝑑𝑆 𝑑𝑡 = 𝑄 𝑆 0
− 𝑄𝑆 + 𝑉 𝑟 𝑠𝑢

Eq. (2)

If reaction rate rg and rsu are limited by the carbon source, the Monod equation can be
applied and the terms rg and rsu are expressed as follows:
111

𝑟 𝑔 =
𝜇 𝑚𝑎𝑥
𝑋𝑆
𝐾 𝑠 + 𝑆 − 𝑘 𝑑 𝑋 Eq. (3)

𝑟 𝑠𝑢
= −
𝜇 𝑚𝑎𝑥
𝑋𝑆
𝑌 ( 𝐾 𝑠 + 𝑆 )
= −
𝑘 ′
𝑋𝑆
𝐾 𝑠 + 𝑆 Eq. (4)

Eq. (1) can be expressed by substitution of rg as
𝑉 𝑑𝑋 𝑑𝑡 = 𝑄 𝑋 0
− 𝑄𝑋 + 𝑉 (
𝜇 𝑚𝑎𝑥
𝑋𝑆
𝐾 𝑠 + 𝑆 − 𝑘 𝑑 𝑋 ) Eq. (5)

We can assume that the concentration of biomass in influent can be neglected, and at steady
state (𝑑𝑋 /𝑑𝑡 = 0), Eq. (5) can be expressed as
1
𝜃 =
𝜇 𝑚𝑎𝑥
𝑆 𝐾 𝑠 + 𝑆 − 𝑘 𝑑 Eq. (6)
where the hydraulic retention time 𝜃 is given by 𝑉 /𝑄 , and is the reciprocal of the dilution rate Dr
of the biokinetic reactor.
The term rsu in Eq. (4) is determined using the following expression:
𝑟 𝑠𝑢
= −
𝑄 𝑉 ( 𝑆 0
− 𝑆 ) = −
𝑆 0
− 𝑆 𝜃
Eq. (7)

The combination of Eqs. (4), (6), and (7) leads to the following relation:
1
𝜃 = −𝑌 𝑟 𝑠𝑢
𝑋 − 𝑘 𝑑 = 𝑌 𝑆 0
− 𝑆 𝑋𝜃
− 𝑘 𝑑
Eq. (8)

Substitution of the term rsu into Eq. (2) leads to the relationship
𝑉 𝑑𝑆 𝑑𝑡 = 𝑄 𝑆 0
− 𝑄𝑆 + 𝑉 ( −
𝑘𝑋𝑆 𝐾 𝑠 + 𝑆 ) Eq. (9)

Under steady-state conditions, 𝑑𝑆 /𝑑𝑡 = 0, Eq. (9) takes the form
( 𝑆 0
− 𝑆 )
𝑋𝜃
=
𝑘𝑆
𝐾 𝑠 + 𝑆 Eq. (10)
112

The Monod kinetic parameters, namely, the half-saturation coefficient Ks, and the maximum
substrate utilization rate k ’ can be estimated from the Lineweaver-Burk plot of [ Xθ/( S0-S)] versus
[1/S], based on Eq. (10). The microbial yield coefficient Y and the decay coefficient kd can be
determined using the [ 1/θ] versus [(S0- S ) /Xθ] plot based on Eq. (8).
The experimental results are shown in Figure 4.4 and Figure 4.5. Figure 4.4 shows the
overall E.coli growth curve in a secondary clarifier effluent at a temperature of 25
o
C. There was a
short lag phase in the beginning, followed by the approach of an exponential phase, wherein the
population of E.coli significantly increased to 10
13
CFU per 100 mL. After the exponential phase,
the population was maintained at steady state because the growth rate was equivalent to the death
rate. Eventually, the bacterial growth was on the last stage of decay phase in a batch system. The
substrate concentration expressed as total organic carbon (TOC) declined due to microbial
degradation and organic mineralization. The biokinetic parameters was estimated as shown in
Figure 4.4 and 4.5, and their values are presented in Table 4.1.

Figure 4.4 - Overall E.coli growth curve in secondary clarifier effluent at 25
o
C

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0 12 24 36 48 60 72 84 96 108 120 132
Number of E.coli (log scale)
Time (hr)
113

Figure 4.5 - TOC reduction by E.coli growth at 25
o
C

Table 4.1 - Biokinetic parameters for secondary clarifier effluent
Parameters Value Unit
Specific growth rate constant ( μ) 0.303 hr
-1

Maximum specific growth rate ( μmax) 0.8 hr
-1

Monod half saturation constant (Ks) 11.13 mg L
-1

Substrate concentration (S) 6.8 mg L
-1

Specific death rate (kd) 0.074 hr
-1

Cell yield (Y) 0.754 mg mg
-1

4.3.3 Biodegradation
The term biodegradation refers to the breakdown of organic contaminants or organic matter
occurring due to microbial activity under aerobic or anaerobic conditions, wherein these organics
constitute the microbial food source or substrate. Complete biodegradation involves oxidation of
the parent compound to form carbon dioxide and water, a process known as minerlization that
0
1
2
3
4
5
6
7
8
0 12 24 36 48 60 72 84 96 108 120 132
TOC concentration (mg/L)
Time (hr)
Initial TOC = 6.8 mg/L
114

provides both carbon and energy for growth and reproduction of cells. Biodegradation of any
organic compound can be thought of as a series of degradation steps or a pathway that ultimately
results in the oxidation of the parent compound. Each degradation step in the pathway is catalyzed
by a specific enzyme produced by the microorganisms. Enzymes are most often found within a
cell but are also produced and released from the cell to help initiate degradation reactions.
Biodegradation will stop at any step if the appropriate enzyme is not present, and hence some
organic matter is degraded partially. Furthermore, another type of incomplete degradation that
occurs is cometabolism, in which a partial oxidation of the substrate occurs but the energy derived
from the oxidation is not used to support microbial growth, and therefore, such enzymes are called
nonspecific enzymes (for example, methane monooxygenase).
A number of parameters influence the survival and activity of microorganisms in any
environment. Their abundance in an environment is determined not only by available the carbon
but also by various physical and chemical factors. These include oxygen availability, nutrient
availability, temperature, pH, salinity, water activity, and toxin. Inhibition of biodegradation can
be caused by limitation imposed by any one of these factors. In fact, redox conditions are very
important in determining the extent and rate of contaminant biodegradation. Oxygen is important
for the degradation of highly reduced hydrocarbons such as the low molecular weight of alkanes,
methane. The organic matter content is also important for the sustenance of microbial population
because it represents the carbon source for microorganisms, and its concentration is lower in the
deeper regions such as the groundwater region. Nitrogen and phosphorus are necessary for
microbial growth and biodegradation, especially, a C:N:P ratio of 100:10:1. Bacteria can adapt to
extreme temperature in order to maintain metabolic activity; however, seasonal temperature
fluctuations in the natural environments have been shown to affect biodegradation rates. The
115

degradation rates of hydrocarbon degradation are influenced by the pH, and the highest rates are
generally observed at neutral pH.

4.3.4 Biofilm
A biofilm is a surface association of microorganisms that is strongly attached through the
production of an extracellular polymer matrix. Biofilms are characterized by the presence of
bacterial extracellular polymers which can create a visible “slimy” layer on a solid surface
(Marshall, 1992). The biofilm kinetics have been developed by understanding the biological
treatment processes (Rittmann, 1987). The exopolymer matrix is also an integral component
influencing the functioning and survival of biofilms in hostile environments. Biofilm development
is initiated by the attachment of bacteria to a solid surface. Microorganisms benefit from a biofilm
community. Biofilms can protect the attached community from environmentally stressful
conditions such as desiccation or changes in pH and temperature. They can also provide some
protection of the cells from predation by protozoa, from antibiotics and from disinfectants. The
filtration of natural organic matter (NOM) from the water by biofilm occurs in water purification
systems, exemplified by trickling filters and fluidized bed reactors.
In this study, biofilm was developed on UF membranes (Figure 4.6 - a) and on activated
particles (Figure 4.6 - b).
116

(a) (b)
Figure 4.6 - Scanning electron microscope (SEM) showing the growth of E.coli (a) on the UF
membrane surface at 50,000 times magnification, and (b) on the activated carbon particle surface
at 10,000 times magnification

4.4 Results and Discussion
4.4.1 Micro-Scale UF Hollow Fiber Membrane
Membrane filtration test for a single UF hollow-fiber was conducted using secondary
clarifier effluent (SCE) to investigate the effect of trans-membrane pressure (TMP) on permeate
flux and TOC removal. Figure 4.7 shows the permeate flux pattern as a function of operating time
under different TMPs of 10, 20, and 30 psi. The initial fluxes were proportional to the TMP and
were 93.3 L m
-2
hr
-1
, 180 L m
-2
hr
-1
, and 266.7 L m
-2
hr
-1
at 10 psi, 20 psi, and 30 psi, respectively.
However, the flux after 6 hours operation was not exactly linear owing to the gel layer compaction.
The fluxes were 44.6 L m
-2
hr
-1
, 127.4 L m
-2
hr
-1
, and 165.6 L m
-2
hr
-1
at 10 psi, 20 psi, and 30 psi,
respectively. Figure 4.8 shows the effect of TMP on the permeate TOC concentration. The TOC
concentration in feed reservoir was 6.7 mg L
-1
. As can be observed, the TOC removal was
approximately 33 - 34% under TMPs of 10, 20, and 30 psi. In this study, a TMP of 20 psi was
117

chosen because the flux deterioration was minimal as compared to other values, while the TOC
removals were similar at all values.

Figure 4.7 - Effect of trans-membrane pressure on permeate flux for SCE using micro-scale UF
hollow fiber membrane

Figure 4.8 - Effect of trans-membrane pressure on TOC concentration of SCE using micro-scale
UF hollow fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
)

0
50
100
150
200
250
300
0 1 2 3 4 5 6 7
Permeate Flux (L/m
2
/hr)
Time (hr)
10 psi
20 psi
30 psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7
TOC (mg/L)
Time (hr)
10 psi
20 psi
30 psi
Feed solution TOC=6.7mg/L
118

Effect of PAC on Permeate Flux and TOC
Membrane filtration tests were conducted using the SCE to investigate the effect of PAC
on permeate flux and TOC removal. Figure 4.9 shows the permeate flux pattern at a TMP of 20
psi. It is important to note that PAC lowered the initial flux for flat sheet membrane discussed in
Chapter 3. The application of PAC did not affect the initial flux because of the characteristics of
the hollow-fiber configuration and the associated fluid dynamics. The initial fluxes in the presence
and absence of PAC were both equal to 180 L m
-2
hr
-1
. The flux decline for PAC addition was
significant within the first two hours due to the potential transport resistance caused by the PAC
and as time progresses the compaction of the PAC layer could increase the resistance. Figure 4.10
presents the extent of TOC removal with addition of PAC. The presence of PAC slightly lowered
the TOC removal by 46% as compared to 34% without PAC.

Figure 4.9 - Effect of PAC addition on permeate flux of SCE using micro-scale UF hollow fiber
membrane

0
20
40
60
80
100
120
140
160
180
200
0 1 2 3 4 5 6 7
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
40 mg/L PAC
Trans-membrane pressure=20psi
119

Figure 4.10 - Effect of PAC addition on TOC concentration of SCE using micro-scale UF hollow
fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
)

Effect of E.coli on Permeate Flux and TOC
Micro-scale UF hollow fiber membrane filtration tests were conducted to investigate the
effect of E.coli addition on permeate flux and TOC removal. The concentration of E.coli was
maintained at 10
8
CFU per 100 mL. The effect of E.coli on permeate flux is shown in Figure 4.11.
Similar to the experiments of PAC, the initial flux were the same as 180 L m
-2
hr
-1
. However, the
flux gradually declined to 114.7 L m
-2
hr
-1
due to the presence of E.coli possibly due to biological
fouling. Figure 4.12 shows the effect of E.coli on TOC removal. The TOC removal was 55% in
the presence of E.coli compared to 34% when E.coli was absent. It appears that the E.coli had
significant effect on TOC removal due to microbial degradation and mineralization or
consumption of organic matter for cell growth.

0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
40 mg/L PAC
Feed solution TOC=6.7mg/L
120

Figure 4.11 - Effect of E.coli addition on permeate flux of SCE using micro-scale UF hollow fiber
membrane

Figure 4.12 - Effect of E.coli on TOC concentration of SCE using micro-scale UF hollow fiber
membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
)

Combined Effect of PAC and E.coli on Permeate Flux and TOC
Micro-scale UF hollow-fiber membrane tests were conducted to evaluate the combined
effect of PAC and E.coli on the permeate flux and extent of TOC removal. Figure 4.13 shows the
0
20
40
60
80
100
120
140
160
180
200
0 1 2 3 4 5 6 7
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
10 CFU/100mL E.coli
Trans-membrane pressure=20psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
10 CFU/100mL E.coli
Feed solution TOC=6.7mg/L
8
8
121

permeate flux decline pattern as a function of operation time. As can be observed, the flux decline
was significant within the first two hours similar to the trends observed for PAC alone (Figure 4.9).
The initial flux was 180 L m
-2
hr
-1
and after 2 hours dropped to 125 L m
-2
hr
-1
. The flux decline
was possibly due to the shearing of microorganisms from PAC, and the formation of bacterial film
on the membrane surface. Figure 4.14 presents the effect of both PAC and E.coli on the TOC
removal. The TOC removal was 70%, and was higher than those experienced with either PAC or
microorganisms. The permeate TOC concentration was as low as 2 mg L
-1
. Evidently, the
combination of E.coli and PAC has a synergistic effect on TOC removal.

Figure 4.13 - Effect of E.coli and PAC addition on permeate flux of SCE using micro-scale UF
hollow fiber membrane
0
20
40
60
80
100
120
140
160
180
200
0 1 2 3 4 5 6 7
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
PAC (40 mg/L) and E.coli (10 CFU/100mL)
Trans-membrane pressure=20psi
8
122

Figure 4.14 - Effect of E.coli and PAC addition on TOC concentration of SCE using micro-scale
UF hollow fiber membrane (TOC concentration in feed reservoir = 6.7 mg L
-1
)

4.4.2 Mini-Pilot-Scale UF Hollow Fiber Membrane
Permeate Flux and Trans-membrane Pressure
Membrane permeate flux measurements were essential to determine the effect of
microorganism concentration on the overall aqueous transport through hollow-fiber membranes.
In this study, the trans-membrane pressure (TMP) was varied from 2 psi to 8 psi. Figure 4.15 shows
the relationship between the permeate flux and the TMP for a DDI feed water at 20
o
C. As can be
seen, the permeate flux is directly proportional to the applied TMP. It is important to note that DDI
water does not contain any foulants such as particulates, microorganisms, organic compounds, as
previously discussed in the flat sheet membrane tests.
The secondary clarifier effluent (SCE) contains particulates, organics, mineral, and
microorganisms that adversely affect membrane performances due to membrane fouling. The
permeate flux pattern as a function of operation time at different TMPs for SCE is shown in Figure
4.16. As can be seen, the initial flux was proportional to the increase in the applied TMP. The
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
PAC (40 mg/L) and E.coli (10 CFU/100mL)
Feed solution TOC=6.7mg/L
8
123

initial flux corresponding to 2 psi, 4 psi, and 6 psi were 9.1 L m
-2
hr
-1
, 18.1 L m
-2
hr
-1
, and 26 L
m
-2
hr
-1
, respectively. As can be observed in Figure 4.16, the permeate flux converged to
approximately 5 - 7 L m
-2
hr
-1
. The permeate flux decline after 12 hours of operation time were
5 (45%), 6.2 (65.7%), and 7.3 L m
-2
hr
-1
(71.9%) for 2, 4, and 6 psi, respectively. The reduction of
fractional permeate flux at higher TMPs can be explained by increase in gel layer compactness
and in internal pore fouling. Figure 4.17 shows the effect of different TMPs on the permeate TOC
concentration. The feed TOC concentration was 6.5 mg L
-1
. As can be observed, the extent of TOC
removal ranged from 4.6% to 9.2% for the three TMPs.
The mini-pilot-scale UF hollow fiber membrane filtration experiment is a continuous flow
system. It is important to determine the hydraulic retention time (HRT) for optimal operation
conditions such as permeate flux rate, constituents in the bioreactor, and recirculation rate. Figure
4.18 shows the effect of HRT on permeate flux at 4 psi of TMP. Two different HRTs were
investigated in this study. As can be seen, the permeate flux patterns were nearly the same under
different HRTs of 14.4 hours and 27.7 hours. The fact that the permeate flux was not affected at
different HRTs could be explained by that the constituents in both the reactor and the feed reservoir
were homogeneously well-mixed, and the shear off rates of surface foulants were nearly the same
in each case. Figure 4.19 shows the effect of HRT on permeate TOC concentrations. Evidently,
the extent of TOC removal was practically the same at 7% for the two HRTs.

124

Figure 4.15 - Relationship between permeate flux and trans-membrane pressure using mini-pilot-
scale UF hollow fiber membrane using distilled deionized water

Figure 4.16 - Effect of trans-membrane pressure on permeate flux for SCE using mini-pilot-scale
UF hollow fiber membrane

y = 17.4x + 1
R² = 0.9995
0
20
40
60
80
100
120
140
160
0 1 2 3 4 5 6 7 8 9
Permeate Flux (L/m
2
/hr)
Trans-membrane Pressure (psi)
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
6 psi
4 psi
2 psi
125

Figure 4.17 - Effect of trans-membrane pressure on TOC concentration of SCE using mini-pilot-
scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
)

Figure 4.18 - Effect of hydraulic retention time on permeate flux for SCE using mini-pilot-scale
UF hollow fiber membrane
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC Concentration (mg/L)
Time (hr)
6 psi
4 psi
2 psi
Feed solution TOC=6.5mg/L
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
HRT=14.4 hrs
HRT=27.7 hrs
Trans-membrane pressure=4psi
126

Figure 4.19 - Effect of hydraulic retention time on TOC concentration of SCE using mini-pilot-
scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
)

Effect of PAC on Permeate Flux and TOC
Mini-pilot-scale UF hollow fiber membrane filtration experiments were conducted to
investigate the effect of PAC on performance factors such as permeate flux and TOC removal.
Figure 4.20 shows the effect of PAC addition on permeate flux at 4 psi of TMP. As can be observed,
the presence of PAC lowered the initial flux from 18.1 L m
-2
hr
-1
to 13.5 (25%) and 11.9 L m
-2

hr
-1
(34%) for 40 and 100 mg L
-1
of PAC, respectively. The reduction in initial flux can be
explained by the resistance of the compacted PAC layer on the membrane surface. Figure 4.21
depicts the effect of PAC addition on TOC removal. As evident, the presence of PAC drastically
decreased TOC concentration in the permeate by 46.2% (3.5 mg L
-1
) for 40 mg L
-1
of PAC, and
53.8% (3.0 mg L
-1
) for 100 mg L
-1
of PAC.
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC Concentration (mg/L)
Time (hr)
HRT=14.4 hrs
HRT=27.7 hrs
Feed solution TOC=6.5mg/L
127

Figure 4.20 - Effect of PAC addition on permeate flux of SCE using mini-pilot-scale UF hollow
fiber membrane

Figure 4.21 - Effect of PAC addition on TOC concentration of SCE using mini-pilot-scale UF
hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
)

Effect of E.coli on Membrane Flux and TOC
The rejection of microorganisms and their impact on the membrane permeate flux due to
microbial and biological fouling is important from the standpoint of membrane processes in water
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
40 mg/L PAC addition
100 mg/L PAC addition
Trans-membrane pressure=4psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
40 mg/L PAC addition
100 mg/L PAC addition
Feed solution TOC=6.5mg/L
128

reclamation applications. The E.coli was chosen as a model microorganism and it appeared logical
to evaluate its effect on membrane performance. As can be seen in Figure 4.22, the result shows
that E.coli population affected the initial flux due to microbial and biological fouling. However,
after 6 hours of operation, the extent of flux decline was mitigated by the presence of E.coli in this
study due to the reduction of organic fouling in the bioreactor. Figure 4.23 shows the effect of
E.coli addition on the permeate TOC concentration. It must be noted that the initial TOC in feed
was 6.5 mg L
-1
. It can be observed that in the presence of E.coli the extent of TOC removal was
35.4% (from 6.5 to 4.2 mg L
-1
). Nonetheless, no E.coli was detected in permeate, as these
microorganisms were completely removed by the UF membranes based on size exclusion.

Figure 4.22 - Effect of E.coli addition on permeate flux of SCE using mini-pilot-scale UF hollow
fiber membrane
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
10 CFU/100mL E.coli
Trans-membrane pressure=4psi
8
129

Figure 4.23 - Effect of E.coli on TOC concentration of SCE using mini-pilot-scale UF hollow fiber
membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
)

Combined Effect of PAC and E.coli on Permeate Flux and TOC
Experiments were conducted to evaluate the combined effect of PAC and E.coli on the
membrane filtration performances. Figure 4.24 shows the effect of combination of PAC and E.coli
on permeate flux pattern. The flux patterns were similar to those observed in Figure 4.22 due to
the presence of E.coli alone. The extent of flux decline was about 30% (12.8 L m
-2
hr
-1
) after 4
hours of operation. The fact that fouling mitigation occurred after 4 hours is due to the combined
effect of biofilm on PAC, adsorption of foulant organic matter, and organic removal. In the case
of E.coli used without PAC (Figure 4.22) the fouling mitigation was experienced only after 6 hours.
Figure 4.25 shows the effect of combined PAC and E.coli on TOC removal. The TOC removal
was 60% highlighting the fact that the combination of PAC and E.coli significantly enhanced
organic removal.

0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
10 CFU/100mL E.coli
Feed solution TOC=6.5mg/L
8
130

Figure 4.24 - Effect of E.coli and PAC addition on permeate flux of SCE using mini-pilot-scale
UF hollow fiber membrane

Figure 4.25 - Effect of E.coli and PAC addition on TOC concentration of SCE using mini-pilot-
scale UF hollow fiber membrane (TOC concentration in feed reservoir = 6.5 mg L
-1
)

Effect of Ozonation on SCE
In order to study the effect of ozonation on membrane filtration, 5 mg L
-1
of ozone was
applied for duration of 6 hours, 15 hours, and 24 hours. These durations corresponded to TOC
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
PAC (40mg/L) and E.coli (10 CFU/100mL)
Trans-membrane pressure=4psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
PAC (40mg/L) and E.coli (10 CFU/100mL)
Feed solution TOC=6.5mg/L
8
8
131

reductions of about 10%, 20%, and 30%, respectively. Figure 4.26 shows the effect of ozoated
SCE on permeate flux. As can be seen, the permeate flux of the ozonated SCE exhibited lower
flux loss than without ozonation due to reduction of organic foulants. However, no significant
differences between the TOC reductions were observed after 2 hours of ozonation due to break
down of most of the organic matter to smaller molecules. Figure 4.27 shows the effect of ozonation
on TOC removal. The initial TOC of the SCE was 6.5 mg L
-1
, and the TOC loads after ozonation
were 5.3, 4.8, and 4.2 mg L
-1
, corresponding to TOC reductions of 10%, 20%, and 30%,
respectively. As can be observed in Figure 4.27, after ozonation the TOC was not significantly
reduced. It can be postulated that UF membrane could not retain ozonated SCE by size exclusion
mechanisms. It is important to note that ozonation processes break down organics to smaller
molecules but cannot mineralize them completely. Furthermore, the smaller organic molecules
may cause internal pore fouling.

Figure 4.26 - Effect of ozonated SCE on permeate flux using mini-pilot-scale UF hollow fiber
membrane

0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
10% TOC reduction
20% TOC reduction
30% TOC reduction
Trans-membrane pressure=4psi
132

Figure 4.27 - Effect of ozonated SCE on TOC concentration using mini-pilot-scale UF hollow
fiber membrane: TOC concentrations for 10% TOC reduction, 20%, and 30% were 5.3, 4.8, and
4.2 mg L
-1
, respectively

Membrane Permeate Flux Recovery
Permeate flux recovery is an essential aspect of membrane processes for sustained
operations in various applications. Membrane cleaning and flux recovery are significant for
membrane longevity and durability besides energy conservation. There are several methods to
recover the membrane filtration performances by removing membrane fouling. Fouled membrane
can be cleaned by physical, chemical, or biochemical processes. In this chapter, physical and
chemical cleaning methods were investigated. The term “Relaxation” refers to leaving the fouled
membrane in a bioreactor as it stands and releasing the trans-membrane pressure. Figure 4.28
shows the effect of pressure releasing, “relaxation,” on the flux recovery of (a) SCE and (b) SCE
with 40 mg L
-1
of PAC addition. The pressure was released after 12 hours of operation time, and
the fouled membrane was left alone for 12 hours in the reactor, and subsequently the membrane
filtration was continued for another 12 hours. As evident in Figure 4.28 (a) and (b), the permeate
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
10% TOC reduction
20% TOC reduction
30% TOC reduction
Initial feed of SCE=6.5mg/L
133

flux recoveries were by 90%, 17 L m
-2
hr
-1
(initial flux of 18.7 L m
-2
hr
-1
), and 93%, 11.1 L m
-2

hr
-1
(initial flux of 11.9 L m
-2
hr
-1
), respectively. The “Relaxation” method was effective in
achieving flux recovery to some extent. Figure 4.29 shows the effect of “relaxation” on TOC
concentration. As can be seen, “relaxation” did not affect TOC removals.
Backwash and chemical cleaning are widely used for the removal of different classes of
foulants. In this study, backwash was conducted by DDI water for 2 hours to remove surface
foulants, and chemical cleaning methods were investigated for removing surface and internal
fouling. The cleaning agents used were the followings: 1 × 10
-3
M of sodium hydroxide (NaOH)
and bioactive agent, RID-X with different concentrations such as 5, 50, and 1000 mg L
-1
. The
cleaning was performed for 2 hours and the chemical residuals were thoroughly washed using DDI
water. Figure 4.30 shows the effect of each cleaning procedure on permeate flux recovery. Overall,
membrane cleaning by sodium hydroxide was the most effective with a flux recovery of 75.9%.
As can be observed, the flux recovery was improved by an increase in RID-X concentrations; the
flux recoveries corresponding to 5, 50, and 1000 mg L
-1
of RID-X were 41.7%, 46%, and 69%,
respectively. Figure 4.31 shows the permeate TOC concentration under different cleaning
scenarios for an initial feed TOC of 6.5 mg L
-1
. As can be observed, the extent of TOC removals
exhibited similar trends ranging from 11.1% to 15.9%.
134

(a)

(b)
Figure 4.28 - Effect of “relaxation” on permeate flux recovery of SCE using mini-pilot-scale UF
hollow fiber membrane: (a) SCE at 4 psi of TMP; (b) SCE with 40 mg L
-1
of PAC addition at 4
psi of TMP
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24
Permeate Flux (L/m
2
/hr)
Time (hr)
Secondary clarifier effluent
Trans-membrane pressure=4psi
0
2
4
6
8
10
12
14
0 2 4 6 8 10 12 14 16 18 20 22 24
Permeate Flux (L/m
2
/hr)
Time (hr)
40 mg/L of PAC addition
Trans-membrane pressure=4psi
Relaxation
(disconnect
pressure supply)
Relaxation
(disconnect
pressure supply)
135

(a)

(b)
Figure 4.29 - Effect of “relaxation” on TOC concentration using mini-pilot-scale UF hollow fiber
membrane: (a) SCE at 4 psi of TMP; (b) SCE with 40 mg L
-1
of PAC addition at 4 psi of TMP
(TOC concentration in feed reservoir = 6.5 mg L
-1
)

0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24
TOC Concentration (mg/L)
Time (hr)
Secondary clarifier effluent
Initial feed=6.5mg/L
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24
TOC Concentration (mg/L)
Time (hr)
40 mg/L of PAC addition
Initial feed=6.5mg/L
Relaxation
(disconnect
pressure supply)
Relaxation
(disconnect
pressure supply)
136

Figure 4.30 - Flux recovery and filtration after membrane cleaning processes for SCE with mini-
pilot-scale UF hollow fiber membrane (TMP = 4 psi): Backwash with DDI water at 6 hours,
cleaning with 5 mg L
-1
of RID-X at 12 hours, cleaning with 50 mg L
-1
of RID-X at 18 hours,
“relaxation” at 24 hours, cleaning with 1 × 10
-3
M of sodium hydroxide at30 hours, cleaning with
1000 mg L
-1
of RID-X at 36 hours

Figure 4.31 - TOC concentration after membrane cleaning processes for mini-pilot-scale UF
hollow fiber membrane: Backwash with DDI water at 6 hours, cleaning with 5 mg L
-1
of RID-X
at 12 hours, cleaning with 50 mg L
-1
of RID-X at 18 hours, “relaxation” at 24 hours, cleaning with
1 × 10
-3
M of sodium hydroxide at 30 hours, cleaning with 1000 mg L
-1
of RID-X at 36 hours
(TOC concentration in feed reservoir = 6.5 mg L
-1
)
0
2
4
6
8
10
12
14
16
18
20
0 6 12 18 24 30 36 42
Permeate flux (L/m
2
/hr)
Time (hr)
0
2
4
6
8
10
12
14
16
18
20
0 6 12 18 24 30 36 42
TOC Concentration (mg/L)
Time (hr)
64.7%
41.7% 46%
64.7%
75.9%
69%
18.7 L/m
2
/hr
After
backwash
After
5 mg/L
RID-X
After
50 mg/L
RID-X
After
1000 mg/L
RID-X
After
“relaxation”
After
1mM
NaOH
After
backwash
After
5 mg/L
RID-X
After
50 mg/L
RID-X
After
1000 mg/L
RID-X
After
“relaxation”
After
1mM
NaOH
137

Model for MBR Systems Predicting TOC Removals
An important objective of the present study was to evaluate the MBR process dynamics
efficiencies under different operating conditions, and to assess the relative contributions of
different mechanisms including PAC sorption and microbial biodegradation to the overall TOC
removal efficiencies. In this regard, the TOC removals (besides permeate fluxes) were studied
under different process operation scenarios. The MBR system operational scenarios included the
following: (a) no PAC adsorbent with only E.coli microorganisms at 10
8
CFU per 100 mL
concentration.; (b) 40 mg L
-1
PAC adsorbent with no microorganisms; (c) 100 mg L
-1
PAC
adsorbent with no microorganisms; and (d) 40 mg L
-1
PAC adsorbent with E.coli microorganisms
at 10
8
CFU per 100 mL concentration. The results of the predictive modeling studies and the
comparisons between the experimental data and theoretical predictions are presented in Figures
4.32 to 35.
It can be easily seen from Figure 4.32 that the TOC removals attributable to microbial
degradation in the presence of E.coli at a concentration of 10
8
CFU per 100 mL was only about
35% under steady-state operating conditions. The model predictions show some deviations at the
beginning of the experimental run (during the first two hours), but the steady state dynamics are
well within 5 to 10% of the model predictions at a later stage. It must be noted that the model uses
the biokinetic parameters evaluated from independent bioreactor studies that take into account the
gradual acclimation of the microorganisms to the organic substrates (represented by the TOC),
while the MBR operation is conducted with microorganisms well acclimated before being
employed in the system. This is one plausible explanation for the model not capturing the sudden
reduction in effluent TOC levels at the beginning of the run. The deviations of theoretical
predictions and the experimental results diminish as the MBR operation is continued. A similar
138

trend is observed in the cases employing only PAC adsorbent although in quantitative terms the
deviations between the experimental results and the theoretical predictions disappears at a faster
rate. In the case of the MBR system using both PAC adsorbent and microorganisms, similar trends
in deviations are observed between the model predictions and experimental results.
A qualitative comparison of the experimental results provide a plethora of insights into the
TOC removal mechanisms in the MBR system. It must be remembered that the TOC in the
biologically treated wastewater used in this study shall represent natural organic matter (NOM) of
a broad spectrum of molecular weights and molecular sizes ranging from 20 to 1,000 Daltons
consisting of several classes of natural organic matter including proteins, fats, carbohydrates,
humic substances, microbial enzymes, and other organic macromolecules of biological origin.
There shall be some fractions that are neither amenable to adsorption by PAC nor microbial
degradation by E.coli, some that are adsorbable by PAC but poorly biodegradable by
microorganisms; some that are not adsorbed by PAC but microbiologically degradable; and some
that are both adsorbable and microbiologically degradable. Under the circumstances, a gradation
in TOC removals can be observed in the MBR system under different operational conditions. The
steady state TOC removal was about 35% in the presence of E.coli alone at 10
8
CFU per 100 mL
(Figure 4.32). In comparison, TOC removals at steady state due to PAC adsorption alone were
higher at two different doses of the adsorbent, namely, 40 and 100 mg L
-1
, at 46% and 54%,
respectively, as can be observed from the results depicted in Figures 4.33 and 4.34. These results
indicate that the TOC constituents are more amenable to PAC adsorption than to microbial
degradation by E.coli, and also that higher PAC doses can remove more poorly biodegradable and
yet better adsorbable fractions. On a related note, distinction must be made here between
biodegradation and complete mineralization of the TOC constituents. Only complete
139

mineralization shall reflect TOC reduction due to microbial action, while degradation of larger
organics into smaller molecules shall maintain the same level of TOC as there is no net conversion
of organic carbon to carbon dioxide. The combination of PAC adsorption and microbial
degradation leads to enhanced TOC removals as shown in Figure 4.35. The steady state TOC
removal is about 60%, and it is significantly higher than that observed even at high PAC dosages
of 100 mg L
-1
. The combination of adsorption and microbial degradation results in the removal of
poorly biodegradable and slowly adsorbable organic fractions as well. These qualitative and
quantitative features of the MBR process are reflected by the experimental data and model
predictions.

Figure 4.32 - Experimental data and theoretical predictions of the TOC concentration profiles in
the mini-pilot MBR reactor in the absence of PAC adsorbent and presence of 10
8
CFU per 100 mL
E.coli

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
Experimental data
Predicted model profile
Initial TOC=6.5mg/L
140

Figure 4.33 - Experimental data and theoretical predictions of the TOC concentration profiles in
the MBR using 40 mg L
-1
PAC adsorbent and no microorganisms

Figure 4.34 - Experimental data and theoretical predictions of the TOC concentration profiles in
the MBR using 100 mg L
-1
PAC adsorbent and no microorganisms

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
Experimental data
Predicted model profile
Initial TOC=6.5mg/L
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
Experimental data
Predicted model profile
Initial TOC=6.5mg/L
141

Figure 4.35 - Experimental data and theoretical predictions of the TOC concentration profiles in
the MBR reactor system using 40mg L
-1
PAC adsorbent and 10
8
CFU per 100 mL

Ecoli

Membrane Bioreactor Process Dynamics and Model Sensitivity
An important objective of the study was the verification and validation of the model using
experimental data from MBR mini-pilot-scale studies. While model verification involves
evaluation of experimental data used for model calibration and checking its assumptions,
validation investigates the accuracy of the model. Precise determination of calibration variables is
therefore essential for obtaining accurate and efficient predictions. Sensitivity studies were
employed to evaluate the relative influence of each model parameter on the MBR process
dynamics. Consequently, an informed decision was made regarding the accuracy and precision
required for data acquisition and parameter estimation.
The membrane bioreactor process dynamics is complex owing to its dependence on a
number of parameters and variables pertaining to micro-transport, macro-transport, and
biodegradation of the organic contaminants. The principal objectives of model sensitivity studies
and several advantages that can be realized in process design or process upscaling can be
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
Experimental data
Predicted model profile
Initial TOC=6.5mg/L
142

succinctly described as follows: (a) identification of parameters significantly impacting process
dynamics; (b) improvement in process efficiency (TOC removal) by effecting suitable changes in
process or operating conditions; (c) recognition of predominant operative mechanisms to provide
insight into process dynamics, and quantification of changes in observed variables in response to
variations in certain parameters and/or dependent variables; (d) facilitating improvement in
methodologies for parameter estimation with reference to uncertainities of observable variables,
often indicating unsuitable process conditions or features (e) development of means for model
expansion, contraction, or refinement, if necessary, after considering various model components;
and (f) evaluation of key parameters for process design and upscaling. Thus, sensitivity analyses
provide relevant information on process conditions that can be altered by modifying certain
parameters to enhance the overall efficiency. Although the studies were conducted for several
model parameters, only the few that profoundly influenced process dynamics are worth discussing.
The utility of this approach was to establish the sensitivity of MBR process efficiency (TOC
removal) to various parameters pertaining to reactor flow conditions (flow rate Q, and the influent TOC
concentration C 0), adsorption equilibrium and mass-transfer parameters (KF, 1/n, Ds and kf), and
biodegradation parameters including the Monod coefficients (  max , K s, Y and k d) as well as the biological
variables (X, L fmax and D f). The results are reported for those parameters identified as significantly
influencing transient and steady-state process dynamics. These investigations employed the
nominal value of each parameter, and an increment of 50% or a decrement of 50% over that value
(0.5 and 1.5 times the normal values). The process dynamics were evaluated in terms of the
effluent-to-influent concentration ratio (normalized effluent concentration) as a function of
operation time for all the model parameters presented in Table 2.2 (Chapter 2).
143

Model simulation results are presented in Figures 4.36 (a) through (b) only for cases where
parameters had significant influence on process dynamics. The influence of flow rate (Q) and
influent contaminant (TOC) concentration on the process dynamics pertain to the influence of the
overall contaminant loading, and the results are presented in Figure 4.36 (a). The normal flow rate
was 0.93 mL s
-1
or 56 mL min
-1
and the higher and lower flow rates were 84 mL min
-1
and 28 mL
min
-1
, respectively. The effect of influent reactor contaminant (TOC) concentrations on process
dynamics is presented in Figure 4.36 (b) corresponding to TOC levels of 3.4, 6.7, and 10.0 mg L
-1
.
Figures 4.36 (a) and (b) demonstrate that the MBR dynamics are significantly influenced by
changes in flow rate and influent TOC concentration.\
Model sensitivity to the adsorption equilibrium and mass transfer parameters are presented
in Figures 4.37 (a) through (d). Qualitatively similar variations were observed in the MBR dynamics for
direct and inverse variations in the Freundlich capacity coefficient K F and the Freundlich intensity
coefficient 1/n. The MBR effluent profiles were significantly affected as can be seen in Figures 4.37 (a) and
(b). The capacity coefficient appears to have a profound effect on the TOC removal. In qualitative terms, it
can be seen that the total mass removal of the TOC increased or decreased in proportion to the adsorbent
capacity changes reflected by the parameter K F. The impact of the intensity coefficient 1/n was equally
significant under the process conditions. The overall dynamics was also sensitive to some extent on the
adsorption mass-transfer parameters D s and k f, as shown in Figures 4.37 (c) and (d), respectively. The shapes
of the reactor effluent TOC concentration profiles were affected by the changes in these parameters. The
initial part of the profile was more influenced by the film transfer coefficient, k f, while the overall steady-state
removals were controlled by the surface diffusion coefficient D s as well. These results clearly indicated that
adsorption on PAC is as important as microbial degradation for TOC removal. The importance of adsorption
capacity and adsorption rates have an indirect influence and a synergistic effect on the overall TOC reduction
144

attributable to microbial activity in the bulk liquid phase and biofilm reduction in TOC. On a related note, it
must be noted that TOC reduction in the reactor due to adsorption mechanisms is mainly due to the uptake
of the organic matter. In comparison, the reduction in TOC to microbial activity is more due to mineralization
of the organic matter as mere decomposition of complex organic molecules will not reflect any change in
the overall TOC.

(a)

(b)

Figure 4.36 - Sensitivity of MBR TOC removal to (a) the reactor flow rate Q and (b) the influent
TOC concentration C0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
+50% Q
Q = 56mL/min
-50% Q
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% C0
C0 = 6.7mg/L
+50% C0
Model profile
Model profile
-50% C 0
C 0= 6.7 mg/L
+50% C 0
145

(a)

(b)

(c)
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% KF
KF=1.7846
+50% KF
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% 1/n
1/n = 0.2161
+50% 1/n
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% Ds
Ds = 2.0E-10 cm2/s
+50% Ds
Model profile
Model profile
Model profile
146

(d)

Figure 4.37 - Sensitivity of MBR TOC removal to (a) the Freundlich capacity coefficient KF ; (b)
the Freundlich intensity coefficient 1/n; (c) the surface diffusion coefficient Ds; and (d) the film
transfer coefficient kf

Model sensitivity to the Monod biokinetic parameters is illustrated in Figure 4.38 (a)
through (c). Generally, the transient-state dynamics exhibited moderate to substantial sensitivity to
changes in these parameters (+50% variations), while the steady state dynamics were less influenced
by these variations. Qualitatively similar variations were observed in the MBR system dynamics for direct
and inverse variations in the Monod maximum substrate utilization rate μ max and Monod half-saturation
coefficient K s . The MBR effluent profiles were significantly affected as can be seen in Figures 4.38 (a) and
(b). The Monod microbial yield coefficient Y also had a profound influence on the MBR performance
dynamics, as shown in Figure 4.38 (c). The process dynamics were insensitive to changes in the microbial
decay coefficient k d, and the results were not worth reporting. These investigations qualitatively and
quantitatively reflected that the Monod coefficients had a profound influence on MBR process dynamics
and treatment efficiency with regard to TOC removal. Therefore, in order to obtain accurate predictions of
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% kf
kf = 4.0E-3 cm/s
+50% kf
Model profile
147

MBR performance, the Monod parameters are recommended to be determined with fair amount of precision
and accuracy from biokinetic studies. These results also had far-reaching implications with respect to
improving process efficiencies. The simulation studies indicated that the yield coefficient could be improved
by using the most favorable bacterial strain to enhance the microbial mineralization of organic matter
towards TOC removal. The microbial degradation of the larger organic molecules would also tend to
increase the adsorption of these molecules by the PAC adsorbent, resulting in greater TOC removal.

(a)

(b)

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% umax
umax = 0.8 per hr
+50% umax
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% Ks
Ks = 11.13 mg/L
+50% Ks
Model profile
Model profile
148

(c)

Figure 4.38 - Sensitivity of MBR TOC removal to (a) the maximum substrate utilization rate µ max; (b) the
half saturation coefficient K s; and (c) the microbial yield coefficient Y

The sensitivity of process dynamics to the biomass concentration in the liquid phase (X),
maximum biofilm thickness (Lfmax), and biofilm diffusion coefficient Df are presented in Figures
4.39 (a) through (c), respectively. These three parameters have certain important significance
relating to biological activity and transport resistance to TOC reduction. Figure 4.39 (a) demonstrates
that the MBR performance has been strongly influenced by changes in biomass concentration (E.coli
concentration), strongly implying that process efficiency can be controlled by using optimal biomass levels.
In qualitative and quantitative terms, the effect of the biomass concentration X on process dynamics was
similar to that of the Monod maximum substrate utilization rate coefficient µ max. Figures 4.39 (b) and (c)
show the MBR effluent concentration profile for changes in the maximum biofilm maximum biofilm
thicknesses (Lfmax) and the biofilm diffusion coefficient (D f), respectively. The influences of the
maximum biofilm thicknesses Lfmax on the process performance is significantly lower than that of
biomass concentration in the liquid phase (X). It can be easily observed that the biofilm thicknesses
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% Y
Y = 0.754 mg/mg
+50% Y
Model profile
149

Lfmax of 10 µm had a lower influence than the other biological parameters. This could be attributed
to the fact that the overall biomass content has a stronger influence than biofilm thicknesses as a
substantial portion of TOC mineralization might occur due to microorganisms present in the bulk
liquid phase. Nevertheless, Lfmax influences the overall MBR dynamics to the extent controlled by
diffusion transport of organic substrates through the biofilm. Increases in the values of these two
parameters will lead to better reactor performance and better TOC reduction. However, in the case
of the maximum biofilm thickness (Lfmax), increases beyond a certain level could potentially
increase transport resistance to nutrients and carbon source leading to lower performance levels.
Furthermore, excess biomass levels in the liquid phase or biofilm growth and detachment from
PAC can increase membrane fouling (biological fouling) and adversely affect the MBR permeate
flux.

(a)

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% X
X = 1.0E8 CFU per 100mL
+50% X
Model profile
150

(b)

(c)
Figure 4.39 - Sensitivity of MBR TOC removal to (a) the microorganism concentration X; (b) the
maximum biofilm thickness Lfmax; and (c) the biofilm diffusion coefficient Df

Model for Predicting Permeate Fluxes in MBR Systems
The experimental data and model predictions for permeate flux under different trans-
membrane pressures are depicted in Figure 4.40. As can be observed, the predicted model profiles
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% Lfmax
Lfmax = 10 um
+50% Lfmax
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13
TOC Concentration (C/C
0
)
Time (hr)
-50% Df
Df = 1.6E-6 cm2/s
+50% Df
Model profile
Model profile
-50% D f
D f= 1.6×10
-6
cm
2
/s
+50% D f
151

are well fitted to the experimental data. Figure 4.41 shows the comparison of the permeate flux
between the experimental data and the predictions. The model incorporated an additional mass
transfer resistance due to the formation of PAC layers. The initial flux was slightly decreased as
the concentration of PAC increased (Figure 4.41 a, b).

(a) 2 psi

(b) 4 psi
0
1
2
3
4
5
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=2psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=4psi
152

(c) 6 psi
Figure 4.40 - Permeate flux pattern at different trans-membrane pressure for UF hollow fiber
membrane filtration (feed TOC concentration = 6.5 mg L
-1
)

(a) 40 m/L of PAC

0
5
10
15
20
25
30
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data
Predicted model profile
Trans-membrane pressure=6psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data (without PAC)
Predicted model profile (without PAC)
Experimental data (40mg/L PAC)
Predicted model profile (40mg/L PAC)
153

(b) 100 mg/L of PAC
Figure 4.41 - Effect of PAC addition on permeate flux for UF hollow fiber membrane filtration
(feed TOC concentration = 6.5 mg L
-1
)

4.5 Summary and Conclusions
Experiments were conducted with two different scales, namely, micro-scale and mini-pilot-
scale systems to evaluate the feasibility of process scale-up. The micro-scale system consisted of
a single strain of UF hollow-fiber membrane, while the mini-pilot-scale system involved a UF
hollow-fiber module. The benefits of single hollow-fiber membranes were to provide a guideline
for mini-pilot-scale and pilot-scale system by determining major factors such as number of
modules, trans-membrane pressure, permeability rate, rejection rate, and kinetic parameters, to
save cost, time, and effort. The mini-pilot-scale system was designed on the basis of results
obtained from a micro-scale system. The mini-pilot-scale system represented an intermediate stage
between a micro-scale system and a pilot-scale or full-scale system.
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Permeate Flux (L/m
2
/hr)
Time (hr)
Experimental data (without PAC)
Predicted model profile (without PAC)
Experimental data (100mg/L PAC)
Predicted model profile (100mg/L PAC)
154

Both micro-scale and mini-pilot-scale system showed similar experimental results
including permeate flux patterns, TOC removals, and the effectiveness of microorganisms and
PAC adsorbent. According to these tests, permeate flux decline could be attributed to compaction
of gel layer or gel layer formation due to organic and biological foulants, and also to the internal
pore fouling caused by adsorption of organic matter. The TOC removal in micro-scale and mini-
pilot-scale systems were 35% and 10%, respectively. The reason for the difference between these
systems was the effect of trans-membrane pressure. For instance, the TMP of micro-scale was 20
psi, and that for the mini-pilot-scale system was 4 psi. This TMP difference would cause different
degrees of compaction of the gel layer. In both systems, the PAC effectively removed the TOC.
The TOC removal increased to 30% for the micro-scale system and 25% for mini-pilot-scale
system. It is indeed well known that the PAC as a micro-porous adsorbent has a high specific
surface area that allows adsorption of most organic constituents. The addition of microorganisms
was intended to evaluate the effectiveness of biological degradation due to biofilm formation and
to biological activity in the suspension phase. The combination of PAC and microorganisms was
more effective for TOC removal due to the synergistic effects of adsorption and microbial
degradation. Eventually, the TOC removal was 70% for the micro-scale system and 60% for the
mini-pilot-scale system. The permeate flux was successfully recovered by the application of
effective backwash and chemical cleaning.
Scaling relations would be required for upscaling the micro-scale system to the mini-pilot-
scale, reducing the time, effort and cost in these operations under different process scenarios. The
model would also provide the optimal operating time including membrane cleaning based on the
required TOC removal. In this study, the model predicted the permeate flux pattern and TOC
removal with addition of PAC and/or microorganisms at the different concentrations.
155

4.6 References
APHA, AWWA, & WEF (2005). Standard Methods for the Examination of Water and Wastewater,
21st ed., American Public Health Association, Washington, D.C.

Chang, J., Manem, J., & Beaubien, A. (1993). Membrane bioprocesses for the denitrification of
drinking water supplies. Journal of Membrane Science, 80(1), 233-239.

Cicek, N., Franco, J. P., Suidan, M. T., & Urbain, V. (1998). Using a membrane bioreactor to
reclaim wastewater. Journal of American Water Works Association, 90(11), 105-113.

Fane, A. G., Fell, C. J. D. & Nor, M. T. (1980). Ultrafiltration/activated sludge system:
Development of a predictive model. In Ultrafiltration Membranes and Applications, Cooper, A.
R. (editor), 631-658.

Kilduff, J. E., & Weber, Jr., W. J. (1992). Transport and separation of organic molecules in
ultrafiltration membranes. Environmental Science and Technology, 26 (3), 569-577.

Marshall, K. C. (1992). Biofilms: An overview of bacterial adhesion, activity, and control at
surfaces. Control of biofilm formation awaits the development of a method to prevent bacterial
adhesion. American Society Microbiology (ASM) News, 58, 202-207.

Monod, J. (1949). The growth of bacterial cultures, Annual Review of Microbiology, 3(1), 371-
394.

Pirbazari, M., Kim, S. H., Badriyha, B. N., & Ravindran, V. (1996). Hybrid membrane filtration
process for leachate treatment. Water Research, 30(11), 2691-2706.

Ravindran, V., Tsai, H-H, Williams, M. D., & Pirbazari, M. (2009). Hybrid membrane bioreactor
technology for small water treatment utilities: Process evaluation and primordial considerations.
Journal of Membrane Science, 344, 39-54.

156

Rittmann, B. E. (1987). Aerobic biological treatment, Environmental Science & Technology, 21(2),
128-136.

Rosenberger, S., Krüger, U., Witzig, R., Manz, W., Szewzyk, U., & Kraume, M. (2002).
Performance of a bioreactor with submerged membranes for aerobic treatment of municipal waste
water. Water Research, 36, 413-420.

Snyder, S. A., Adham, S., Redding, A. M., Cannon, F. S., DeCarolis, J., Oppenheimer, J., Wert,
E., & Yoon, Y. (2007). Role of membranes and activated carbon in the removal of endocrine
disruptors and pharmaceuticals. Desalination, 202, 156-181.

Tsai, H. H., Ravindran, V., Williams, M. D., & Pirbazari, M. (2004). Membrane bioreactor process
for water denitrification. Journal of Environmental Engineering and Science, 3(6), 507-521 (2004).

Tu, S. C., Ravindran, V., & Pirbazari, M. (2005). A pore diffusion model for forecasting the
performance of membrane processes. Journal of Membrane Science, 265(1-2), 29-50.

Williams, M. D., & Pirbazari, M. (2007). Membrane bioreactor process for removing
biodegradable organic matter from water. Water Research, 41(17), 3880-3893.

Williams, M. D., Ravindran, V., & Pirbazari, M. (2012). Modeling and process evaluation of
membrane bioreactor for removing biodegradable organic matter from water. Chemical
Engineering Science, 84, 494-511.

Witzig, R., Manz, W., Rosenberger, S., Krüger, U., Kraume, M. & Szewzyk, U. (2002).
Microbiological aspects of a bioreactor with submerged membranes for aerobic treatment of
municipal wastewater. Water Research, 36, 394-402.

157

Chapter 5
Developing Polymeric Nanofiltration Membrane Impregnated with
Graphene Oxide Nanoparticles
5.1 Introduction
Improved membrane separation promise to yield substantial environmental and economic
benefits by significantly reducing energy consumption, increasing industrial productivity,
decreasing waste generation, and addressing water shortages (Crittenden et al., 2005; Zhang et al.,
2009a). Environmental applications of membrane processes include water purification, wastewater
treatment, and water reclamation and reuse. However, membrane technologies face several
scientific and technological challenges that must be overcome before their widespread use can be
considered; and these include membrane fouling and permeate flux decline, poor rejection or
selectivity, and large energy footprints (Crittenden et al., 2005; Peng and Escobar, 2004 and 2005).
State-of-the-art polymers used in membrane filtration include such as polyethersulfone
(PES) and aromatic polyamides (PA) are relatively hydrophobic and hence susceptible to fouling,
resulting in low aqueous transport. One of the aims of this work is to develop high-performance
membranes for use in various applications including integrated membrane systems such as
membrane bioreactor processes. The work focuses on a synthesis-guided strategy to develop a
class of polymer composites through infusion of nanomaterials such as graphene oxide (GO),
graphene derivatives, carbon nanotubes (CNTs) zeolite nanocrystals, and other additives into
chemically modified polymeric membranes, yielding superior permeate fluxes, fouling resistance
and rejection properties, while retaining favorable mechanical characteristics. The preliminary
work, incorporating GO into aromatic PA through interfacial polymerization of m-
158

phenylenediamine (MPD) and trimesoyl chloride (TMC) has yielded some promising results. The
membrane produced have shown increased water fluxes (as much as three fold). Therefore, use of
nanomaterials into polymeric matrices will lead to superior membranes for different environmental,
industrial and commercial applications.
A promising area in membrane technology is the infusion of nanomaterials into polymeric
matrices to yield nano-composites. This holds promise for achieving significant improvements in
aqueous transport and fouling resistance. Important examples include the application of carbon
nanotubes mediated micro-transport of water (Xie et al., 2005; Choi et al., 2006, 2007; Kim and
Van der Bruggen, 2010). For instance, the infused carbon nano tubes (CNTs) can enhance polymer
hydrophilicities due to the presence of polar functional groups, thereby significantly reducing their
potential for decreasing organic, biological and particulate fouling. Secondly, CNTs possess
antibacterial properties that can greatly reduce biofilm formation and bio-fouling (Xie et al., 2005).
Lastly, their use can lead to stronger membranes, reducing the possibility of mechanical failures
(Xie et al., 2005). In short, infusion of CNTs into polymeric materials has the potential for
significantly decreasing energy and operation and maintenance costs of membrane systems. For
instance, zeolite nanocrystals (ZNCs) have been shown to improve membrane hydrophilicity and
to enhance aqueous non-viscous micro-transport properties (Lind et al., 2009). Hence, a
combination of CNTs and ZNCs may provide a synergistic effect in enhancing water transport.
Application of nanoparticles in the manufacturing process of polymeric membranes has
received much attention during the last few years with reference to their ability to improve and
increase aqueous permeate transport, and produce desired structure and functionalities (Geise et
al., 2010). Nanoparticles-based membranes can be developed by assembling engineered
nanoparticles into porous membranes or blending them with polymeric or inorganic membranes.
159

Nano-sized inorganic-material-blended composite membranes are attractive candidates owing to
their enhanced properties such as perm-selectivity, hydrophilicity, and fouling resistance.
Functionalized nanopores in graphene monolayers have been studied and developed (Sint
et al., 2008; Suk et al., 2010a). One of the emerging nano-materials in environmental engineering
applications is grapheme oxide (GO), in the form of a single layer of graphite oxide (Hu et al.,
2013; Joshi et al., 2014). GO is modified from graphite using strong oxidizing agents that provide
oxygen functional groups on carbon-based graphite materials (Dreyer et al., 2010). The
oxygenated functionalities change the properties of GO to enhance the hydrophilicity in polymeric
membranes, dispersibility in water, permeability in fluid transport, separation in molecular weight,
resistance of bacterial destruction, etc. (Koeing et al., 2012; Richards et al., 2012; Suk et al., 2010a).
This chapter is the result of collaborative research work between the Professor Pirbazari
research group in the Department of Civil and Environmental Engineering and Professor Hogen-
Esch research group at the USC Loker Hydrocarbon Research Institute. The initial phase of the
collaborative work involved laboratory-scale flat-sheet membrane tests in plate-and-frame cells,
using actual wastewaters with the idea of optimizing material formulations. The uniqueness of the
tests is the special design for minimizing membrane fouling and permeate flux decline using
adsorbents such as powdered activated carbon (PAC). Laboratory-scale studies to evaluate
synthesized membranes in terms of permeate flux decline patterns, membrane fouling and organic
rejection. In the present context, the following preliminary studies were used in the overall
assessment of the performances of the fabricated membranes:
i: Investigate the permeate flux and TOC removal for synthesized membranes
ii: Determine the effect of PAC on permeate flux and TOC
iii: Determine the effect of the cleaning agents on permeate flux recovery and TOC

160

5.2 Research Background
The most widely used polymeric membrane materials in conventional applications include
the following: cellulose acetate, nitrocellulose, and cellulose esters (CA, CN, and CE), polytetra-
fluoro-ethylene (PTFE), polyvinylidene fluoride (PVDF), polypropylene (PP), polyacrylonitrile
(PAN), polyamides, polyimides, polyvinyl chloride (PVC), polysulfone (PS), polyether sulfone
(PES), and polyethylene (PE).
Reverse osmosis (RO) and nanofiltration (NF) besides ultrafiltration (UF) are currently the
primary processes for water applications including desalination and water reclamation, and are the
subject of several recent reviews (Greenlee et al., 2009; Li et al., 2008a). The membranes used in
these processes typically consist of an active polymer but ultrathin (≤ 0.2μm) but highly high cross-
linked layer, which is dense, amorphous and has extremely small interstitial voids (≤ 0.5 nm)
(Kong et al., 2010; Maruf et al., 2012). RO membranes are typically composed of either CA or
aromatic PA, while NF membranes are often made of PA and PVDF, and UF membranes are often
composed of PS and PES. The older CA RO membranes have been used extensively in desalination
applications (Loeb and Sourirajan, 1963; Kimura and Sourirajan, 1967). However, CA membranes
are limited to relatively narrow pH ranges (4.5- 7.9), susceptible to biological fouling, and are
readily compacted at high pressures. In comparison, PA membranes are superior and are
compatible with high temperatures and a wider pH range. Furthermore, they are more stable to
biological attack and pressure compaction and tend to maintain excellent water permeability and
high salt rejection due to the very thin but efficiently sieving surface layer. However, the PAs
susceptible to degradation from the hypochlorous acid (ClOH) added to municipal water used to
prevent biological fouling as well as exposure to high or low pH conditions (Hydronautics Press
Release, 2007). The other major limitation are inherent difficulties in making polymers with
161

uniform and controllable pore sizes at the sub-nm (< 1nm) level because of the lack of appropriate
building blocks and/or methods for controlling pore architecture on this small size regime. Pore
size control is crucial in finding the optimum tradeoff between high water permeability and
effective ion rejection.
Thin-film composite (TFC) RO and NF membranes are currently the primary membranes
for water desalination and water reclamation applications. These membranes generally consist of
two layers: a dense layer and a porous sub-layer (Matsuura, 1994). The dense layer is usually an
aromatic PA that does the actual desalination. It is fabricated on a porous PS sub-layer which gives
it mechanical support and minimizes the drop in pressure (Fritzmann et al., 2007). The membranes
typically consist of an active ultrathin (≤ 0.2μm) but high density amorphous polymer layer and
has extremely small interstitial voids (≤0.5 nm). Nanostructured materials including nanomaterials
such as zeolites and CNTs have gained attention since they act as high flux molecular sieving
membranes for water desalination (Li et al., 2004, 2008a). However, they have some limitations
including residual ion permeability around zeolite crystals so that perfect salt rejection cannot be
achieved. Similarly, CNT pores diameters are too large to act as molecular sieves for excluding
some of the smaller ions, especially the monovalent ions. Graphene also deserves attention due to
its unique properties. It consists of a single layer of hexagonally arranged sp
2
-hybrized carbon
atoms; it is inexpensive and has excellent mechanical properties as observed by several researchers
(Holt et al., 2006; Fornasiero et al., 2008; Booth et al., 2008; Li et al., 2008a, 2008b, 2009;
Rajbartoreh et al., 2011). These properties suggest the potential of graphene to create thin high
flux membranes. However, graphene is impermeable to molecules as small as helium in its pristine
state, and therefore there is need to produce controllable sub-nano pores to facilitate water passage
through membrane as discussed by Nair et al. (2012).
162

Recent simulations and experimental studies by several investigators suggested that sub-
nanometer pores can be controllably generated by the methods such as oxidation (Zhang et al.,
2003), electron beam irradiation (Fischbein et al., 2004; Hashimoto et al., 2004), ion bombardment
(Krasheninnikov et al., 2001, 2002; Inui et al., 2010; Lucchese et al., 2010), or by deposing (Wei
et al., 2009). Some theoretical studies have been made on water transport and ion rejection of
graphene sheets (Sint et al., 2008; Jiang et al., 2009; Bai et al., 2010) but none on the organic
matter rejection. The transport of ions and water through ~0.5 nm graphene pores and CNTs has
been explored using molecular dynamics (MD) simulations (Sint et al., 2008; Suk and Aluru,
2010b). However, graphene is highly hydrophobic and should be expected to result in major
membrane fouling (Wang et al., 2009). On the other hand, graphene oxide (GO), a partially
oxidized and hydrophilic form of graphene, has also been studied as active layer and was found to
increase water flux. However, GO is partially hydrophilic and can be leached from the membrane
over long periods of time (Hu and Mi, 2013; Perreault et al., 2013). Furthermore, studies have
shown that GO has pore diameters on the order of 1 nm which limits its use as an RO membrane
but is consistent with acting as an effective nanofiltration layer. It was used in membrane coating
for enhancing antifouling properties, but had a limited effect on water flux (Choi et al., 2013;
Perreault et al., 2013).
Although most polymeric membrane materials have good mechanical, thermal and
chemical properties, many of these are generally more hydrophobic than desirable (Asatekin et al.,
2006, 2007; Zhu et al., 2009). Hence their long-term use in water-based separations is severely
limited by their susceptibility to membrane fouling and permeate flux decline. These membrane
surfaces can be made more hydrophilic by several techniques including the following: polymer
grafting (Asatekin et al., 2006, 2007; Zhu et al., 2009), polymer blending (Zanini et al., 2007;
163

Rahimpour et al., 2009; Su et al., 2008, 2009; Zhang et al., 2009b), ion beam radiation
(Chennamsetty and Escobar, 2008), plasma treatment (Chen and Belfort, 1999; Kim et al., 2002),
free-radical polymerization (Li et al., 2008b and 2009; Dong et al., 2009), chemical oxidation
(Yoon et al., 2009), ultraviolet radiation grafting (Pieracci et al., 2002; Wei et al., 2006), grafting
with functionalized polymer and chelating agent (Hausman et al., 2009), redox graft
polymerization (Van der Bruggen, 2009). It is generally known that the polysulfone (PS)
membranes are widely used in biological, pharmaceutical, sterilization and environmental
applications due to their mechanical toughness, greater hydrophilic character, chemical inertness
(pH and oxidation tolerance), low cost and high durability (Kull et al., 2005). Membranes
fabricated on the basis of polymer structural modifications, for instance by one or more of the
techniques discussed above, have the potential to offer higher permeate fluxes, lower trans-
membrane pressures, and greater fouling resistance. In turn, this would increase the efficiency of
long-term applications and hence their technological and economical value for various application
including water reclamation and reuse, wherein the rejection of organic matter is a very important
aspect. Polymer blending and formation in conjunction with nanomaterial infusion appears to be
a promising technique for membrane improvement, and is a significant motivation for this work.

Preparation of Membranes
Modified polymers:
Changes in polymer polarity induced by systematical changes in polymer composition was
carried out by copolymerization and/or chemical functionalization. The advantages of
copolymerization include known and reproducible chemical compositions. It appeared that the
most promising from a performance perspective were aromatic polyamides. Several pathways
164

seemed feasible including the following: (a) chain topology i.e. increases in chain branching by
addition of branching units in the diamine. For instance small variable fractions of 1,3,5-
triaminobenzene may be added in order to increase segments densities and this will be explored.
(b) A second and promising approach is chemical modification of the polyamide (PA) through
chemical modification of the polyamide backbone through changes in the copolymerization of
trimesoyl chloride with 2,4-diamino-N,N-dimethylaniline, 1, along with 1,3, phenylenediamine
(MPD) as illustrated in Figure 5.1. Further modification of chemical structure would involve
transformation of 1 into copolymers 3 and 4 where 3 has a cationic character while the neutral
copolymer 4 is a zwitterionic tertiary ammonium sulfonate copolymer (Figure 5.1). Figure 5.2
shows the example of partially sulfonated polyamides for membrane applications.

Figure 5.1 - Synthesis of polyamide copolymers for membrane fabrication

Figure 5.2 - Partially sulfonated polyamides for membrane applications
165

Polyamide-graphene oxide nano-composites:
The addition of nanomaterials to the polymer matrices will affect surface and bulk
properties, and will be subject to considerable control through the membrane synthesis protocols
that are designed to be modular. The steps of chemical functionalization in this study illustrated in
Figure 5.3. The figure shows the incorporation of hydrophilic nanomaterials to an aromatic PA
matrix through interfacial polymerization of MPD and TMC precursors with the MPD containing
varying amounts of nanomaterials. These nanocomposites functionalized carbon based
nanomaterilas such as CNTs and GO, and in this case was functionalized with GO. The low friction
flow of water has been proposed to occur through two-dimensional capillaries formed by closely
spaced graphene sheets (Nair et al., 2012). GO is known to have strong chemical and mechanical
stability (Suk et al., 2010a). Furthermore, GO is found to have antibacterial properties that would
decrease microbial attack on membranes (Hu et al., 2010; Liu et al., 2011). The GO particles shall
be added to various phenylene diamines in variable quantities so that the scalability as
nanofiltration or reverse osmosis membrane material and the effects on the aqueous flow properties
were to be evaluated. In this manner, water flux and antifouling properties could be optimized
without sacrificing membrane selectivity.

Figure 5.3 - Synthesis of graphene oxide-modified polyamides (adopted from Yurdacan, 2015)
1/1

FIGURE 1

FIGURE 2

FIGURE 3

PDA TMC PA
166

Functionalization of graphene oxide:
An additional surface functionalization with nanomaterials could be also done to make the
surface less prone to biological fouling. This will include a carbodiimide mediated
functionalization of carboxylic groups into an amide that carries a tertiary amine (Figure 5.4). In
turn, this will be alkylated to give a tertiary ammonium halide (structures not shown) and the
introduction of a zwitterionic ammonium sulfonate through reaction of the amine with a sulfone
(Figure 5.4).

Figure 5.4 - Chemical modification of graphene oxide for infusion into polymeric matrices

5.3 Materials and Methods
Materials
Membranes: The synthesized membranes used in this study were prepared by Dr. Yurdacan of
Loker Hydrocarbon Research Institute (Yurdacan, 2015) under the direction of Professor Hogen-
Esch. Membrane A was prepared by the following procedures. A polyethersulfonate membrane
(YMPTSP3001) with a molecular weight cutoff (MWCO) of 5,000 Daltons was used as a base
membrane for synthesis. The membrane was placed into deionized water bath for 12 hours. It was
subsequently removed from water bath and dipped into a mixture of 0.5 wt% of graphene oxide
(GO) and 2% w/v of m-phenylenediamine (MPD) solutions for 3 minutes. The excess solution was
spread by a rubber roller and then 0.1% w/v of trimesoyl chloride (TMC) in hexane solvent was
167

poured onto the membrane and stayed for 1 minute. The membrane was removed from the TMC
solution and rinsed by hexane to remove residual reagents. The synthesized membrane was dried
at room temperature (20
o
C) for 10 minutes. The membrane was stored in DDI water prior to
filtration tests.
Membrane B was prepared by the same procedure of Membrane A but dried at 60
o
C for 10
minutes. Membrane C was prepared by the same procedure of Membrane A except the
concentration of GO. Membrane C finally contained 1 wt% of GO. Membrane D were prepared
by the following procedures (Li, 2014). In the first step, the solution of 5 wt% of GO in N,N-
Dimethylformamide (DMF) was sonicated for 45 minutes. The 30 wt% solution of poly (tetrabutyl
ammonium styrene sulfonate-co-styrene-co-4-chloromethyl styrene), or P(BASS-S-CMS) was
dissolved in DMF. The solution of 80% of polyvinylidene fluoride (PVDF) was dissolved in DMF.
In the second step, the GO solution was poured into the solution of P(BASS-S-CMS) and
vigorously mixed for 30 minutes in order to prevent from “stacking up.” In the third step, the
solution of PVDF was added into the mixture of GO and P(BASS-S-CMS) and stirred for 15
minutes at room temperature. In the fourth step, the mixture of solution was then transferred to
petri dish and place into preheated oven at 165
o
C. Finally, membranes were quenched in water at
25
o
C after annealing for 2 hours and stored in DDI water prior to filtration tests.
Membrane E was prepared by the same procedure of Membrane D except the concentration
of P(BASS-S-CMS), applied to 35 wt%. Table 5.1 is shown the differentiation of each membrane
according to the composition of GO and P(BASS-S-CMS).
A polyethersulfone (PES) microfiltration (MF) flat sheet membrane purchased from GE
Osmonics, Inc. (Minnetonka, MN) was used as a support material for the synthesized membranes.
The membrane pore size was 5 μm that might not be affected on TOC removal but support the
168

active membranes (Figure 5.1). The PES membrane was known as a moderate hydrophobic
polymeric material and broad chemical tolerance.
A polyamide (PA) nanofiltration (NF) flat sheet membrane, DOW Filmtec NF90
manufactured by Dow Chemical Co. (Midland, MI) was used to evaluate the feasibility of the
performance of synthesized membranes comparing the NF membrane experiments. The MWCO
of NF90 membrane ranged from 200 to 400 Daltons, and its pore size was between 0.5 nm and 0.6
nm. The typical flux specification provided by the manufacturer was 78 to 102 L m
-2
hr
-1
at 130
psi.
The effective area of membrane in this study was 24 cm
2
(3.72 inch
2
).

Table 5.1 - Composition of synthesized membranes and commercial nanofiltration membrane

Membranes
Polymer of base
membrane
GO
(wt%)
P(BASS-S-CMS)
(wt%)
Annealing
temperature (
o
C)
Membrane A PES 0.5 - Room (~20)
Membrane B PES 0.5 - 60
Membrane C PES 1.0 - Room (~20)
Membrane D PVDF 5 30 165
Membrane E PVDF 5 35 165
NF90 PES - - -

169

Figure 5.5 - Schematic of the configuration of synthesized membrane filtration

Feed Solution: Two different sources of secondary clarifier effluents (SCE) were used in this
study. One was obtained from the San Jose Creek Water Reclamation Plant (SJCWRP) in Los
Angeles County. The other SCE was obtained from the Hyperion Treatment Plant (HTP) in Los
Angeles County. The sampling locations and characteristics of SCE are described in Figure 3.1
and Table 3.1. The samples were passed through a 0.45 μm pore filter to exclude colloidal particles
including bacteria. The TOC of the SCE obtained from SJCWRP and HTP were 7.5 mg L
-1
and
14.5 mg L
-1
, respectively.

Graphene Oxide: Graphene oxide (GO) solution was purchased from Graphene Laboratories, Inc.
(Graphene Supermarket, Calverton, NY). The size of the GO nanoparticles was in the range of 100
nm to 500 nm.

Adsorbent: Powdered activated carbon (PAC) was used as adsorbent to evaluate permeate flux
and TOC removal. It was manufactured by Calgon Carbon Corporation (Pittsburgh, PA).

170

Chemical Cleaning Agents: In order to evaluate the extent of flux recovery and TOC removal,
three types of chemicals were used to clean the membranes. These cleaning agents were sodium
hydroxide (BDH; VWR International, PA), Triton X-100 (Dow Chemical, MI), and RID-X
(Reckitt Benckiser, NJ), representing caustic, surfactant, and enzymatic chemicals, respectively.

Methods
Membrane Filtration: The membrane performances were investigated by tests using a cross-flow
membrane filtration system made of stainless steel plate-and-flame membrane cell (Figure 3.2).
The synthesized membranes were rinsed by DDI water to stabilize the active layer on the surface
and to remove the effect of adsorption by GO (Ai et al., 2011). The synthesized active layer was
sandwiched between a feed spacer and a microfiltration membrane (Figure 5.5). The applied trans-
membrane pressure was 60 psi (4.14 bar) and the cross-flow rate was 1.5 L min
-1
.

Membrane Cleaning Procedure: During the membrane cleaning operation, the fouled membrane
was first backwashed with DDI water, applying 60 psi of TMP for 1 hour. Subsequently, chemical
cleaning was conducted at an applied TMP of 60 psi for a duration of 2 hours; and this operation
was followed by repeated rinsing with DDI water for several times so as to completely eliminate
the traces and residuals of cleaning agents.

Analytical Methods
A Shimadzu TOC-V CSH analyzer (Shimadzu Corp., Kyoto, Japan) was employed to
measure the total organic carbon (TOC), in accordance with standard methods 5310B (APHA et
al., 2005).
171

5.4 Results and Discussion
Results for Preliminary Experimental Works
The initial work on this study involved the development of polymer synthesis protocols
with appropriate reaction schemes, free-radical processes, syntheses conditions such as reaction
times, curing procedures, and quantitatively controlled incorporation of GO into the polymers.
Superior membranes were manufactured by adjusting these conditions. The membranes used in
the series of preliminary tests were prepared by interfacial polymerization by sequential addition
of MPD and TMC on a commercial PES ultrafiltration membrane base. This is the one of the best
commercially available UF membranes for water reclamation and related applications. The
monomers used in the preparation of PA membrane were MPD and TMC. Another set of
membrane s were cast using these monomers MPD and TMC, but with the addition of camphor
sulfonic acid (CSA) and triethanol amine (TEA) to make the membranes material more solvophilic
in nature, and observe their hydrophilicity, aqueous transport and rejection characteristics.
The membranes designated as #1 and #2 were all synthesized by interfacial polymerization
(for ~1 minute) using MPD and TMC, and cured at 60
o
C for 10minutes, except for the presence
of GO for membrane #2. Membrane #3 was synthesized by a similar procedure using MPD and
TMC followed by CSA and TEA. Membranes designated as #1* and #2* were replicates of
membranes #1 and #2, and were tested with 40 mg L
-1
of PAC added to the feed. The purpose of
these tests was to assess the performances of Membranes #1 and #2 in the presence and absence
of PAC with regard to permeate flux patterns and TOC rejection.
The results presented in Table 5.2 summarize the membrane performances based on
permeate flux and TOC rejection. Membranes #2 and #2* showed significantly improved water
flux (100% and 300% at 2 and 3 hours, respectively), exhibiting superior characteristics in
comparison with the best commercially available ultrafiltration membrane of this type used as a
172

bench-mark standard. The results were highly promising in the pursuit of advanced membranes.
Membranes synthesized with GO as the only surface layer are prone to long-term leaching of GO.
The PAC applications in the feed also demonstrated that the presence of GO in the polymer matrix
yielded better permeate flux and TOC rejection with PAC addition as high as 55.9% shown in
Table 5.2 (Membrane #1* and #2*).
The results presented in Table 5.2 can therefore be briefly summarized as follows: The
presence of GO in the polymer matrix improved not only the permeate flux (Membranes #1 and
#2) but also did not compromise with TOC rejection (slightly higher TOC rejection of 32.6%
versus 30.9%). Qualitatively similar results were observed when PAC was added to the feed to
probe the role of GO, if any, regarding membrane fouling. Thus, the permeate fluxes and TOC
rejections were higher for membrane #2* as compared to #1* (presence of GO, see Table 5.2). The
use of CSA and TEA during the polymerization process (Membrane #3) yielded a flux of 37 L m
-2

hr
-1
at 2 and 3 hours, but yielded a lower TOC rejection of 18.6%. In future work, it is of paramount
interest to observe the effect of GO content in the membrane aqueous transport and organic
rejection as part of optimizing the overall membrane performance.

Table 5.2 - Performance comparison of various membranes
Membrane #1 #2
b
#1* #2*
b
#3
a
UF control
Time (hr) Permeate flux (L/m
2
/hr)
0 265 275 200 210 137.5 100
0.5 150 140 125 155 125 70
1 50 50 75 90 69 40
2 18 40 45 65 37 40
3 10 40 45 65 37 30
TOC rejection (%) 30.9 32.6 44.4 55.9 18.6 3.6
Notes: (a) Membrane #3 was made much earlier than the others listed, and does not reflect typical
performance such as membrane #1; (b) Membrane is infused with GO.
*PAC was used in the feed at 40 mg L
-1

173

Effect of GO Synthesized Membranes on Membrane Filtration
Membrane filtration tests were conducted to investigate the permeate flux patterns and the
TOC removals of the synthesized membrane. Figures 5.6 (a) and (b) present that the comparison
of permeate fluxes among the different membranes. As can be observed, the flux decline patterns
were relatively similar for all the membranes in the experimental runs. It appears that the curing
temperature affected the initial permeate flux (Membrane A at 20
o
C and Membrane B at 60
o
C)
because the temperature might change the potential properties of nanoparticles such as stability
and agglomeration. Higher curing temperature (Membrane B) might have caused some changes in
the general characteristics, and hence slightly lowering the initial flux than in the case of
Membrane B.
According to the Figure 5.6 (a), GO enhanced the permeability because the initial flux of
Membrane C containing 1 wt% GO had relatively higher than the initial flux of Membrane A and
B consisting of 0.5 wt% GO. In other words, the incorporation of higher concentrations of GO in
nanocomposite membranes might lead to superior anti-fouling or fouling-resistant properties.
In Figure 5.6 (b), both Membranes D and E show almost similar permeate flux patterns.
Membrane E manifests slightly higher initial flux, 400 L m
-2
hr
-1
, while 325 L m
-2
hr
-1
for
Membrane D due to the less membrane resistance. According to the measurements performed, the
thicknesses of Membrane D and E were 4 mil ± 0.1 mil (1 mil = 0.0254 mm) and 2 mil ± 0.2 mil,
respectively. The permeate flux decreased proportionally decreased as the thickness of the
membrane was increased as observed by some earlier investigator (Villaluenga et al., 2005).
Figures 5.7 (a) and (b) present the effect of membrane synthesis on TOC concentration.
The TOC concentrations in feed reservoir in Figures 5.7 (a) and (b) were 7.5 mg L
-1
and 14.5 mg
L
-1
, respectively. The applied TMP was 60 psi throughout the experimental runs. As can been
174

observed in Figure 5.7, the TOC removal of synthesized membranes was between 58% and 65%.
In Figure 5.7 (a), Membranes B and C exhibited the better TOC rejection (50%) than Membrane
A (40%). The additional GO was effectively functional on TOC removal due to charge effects as
suggested by Hu et al. (2013). In Figure 5.7 (b), overall TOC removal was between 45% and 48%
for Membranes D and E.

(a)

(b)
Figure 5.6 - Comparison of permeate flux for different membranes: (a) Membranes A, B, and C;
(b) Membranes D and E
0
50
100
150
200
250
300
350
400
450
500
0 0.5 1 1.5 2 2.5 3
Permeate Flux (L/m
2
/hr)
Time (hr)
Membrane A
Membrane B
Membrane C
Trans-membrane pressure = 60 psi
0
50
100
150
200
250
300
350
400
450
0 1 2 3 4 5 6
Flux (L/m
2
/hr)
Time (hr)
Membrane D
Membrane E
Trans-membrane pressure=60psi
175

(a)

(b)

Figure 5.7 - Comparison of TOC removal for different membranes: (a) Membranes A, B, and C;
(b) Membranes D and E

Effect of PAC on Membrane Filtration
Membrane A and C were tested to investigate the effect of PAC on permeability and the
TOC removal. The PAC was added in the feed reservoir at a concentration of 40 mg L
-1
. As evident,
0
2
4
6
8
10
12
14
16
18
20
0 0.5 1 1.5 2 2.5 3
TOC (mg/L)
Time (hr)
Membrane A
Membrane B
Membrane C
Initial feed = 7.5 mg/L
Trans-membrane pressure = 60 psi
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6
TOC (mg/L)
Time (hr)
Membrane D
Membrane E
Initial TOC in feed=14.5mg/L
176

the permeate flux decline was significant in the presence of PAC (Figure 5.8). The flux was
lowered from approximately 400 or 450 L m
-2
hr
-1
without PAC addition to 130 L m
-2
hr
-1
to 120
L m
-2
hr
-1
with the addition of PAC for Membranes A and C, respectively. The flux after 3 hours
of operation converged to about 50 L m
-2
hr
-1
for Membrane A with the use of PAC.
Figure 5.9 shows the effect of PAC on the TOC removal. The TOC concentration was
lowered by approximately 85%, and this was attributable to the fact that the PAC could effectively
remove most of the organic matter by adsorption mechanisms.

Figure 5.8 - Effect of PAC on permeate flux for Membranes A and C

Figure 5.9 - Effect of PAC on TOC concentration for Membranes A and C
0
50
100
150
200
250
300
350
400
450
500
0 0.5 1 1.5 2 2.5 3
Permeate Flux (L/m
2
/hr)
Time (hr)
Membrane A
Membrane C
Membrane A + PAC
Membrane C + PAC
Trans-membrane pressure = 60 psi
PAC addition = 40 mg/L
0
2
4
6
8
10
12
14
16
18
20
0 0.5 1 1.5 2 2.5 3
TOC (mg/L)
Time (hr)
Membrane A
Membrane C
Membrane A + PAC
Membrane C + PAC
Initial feed = 7.5 mg/L
Trans-membrane pressure = 60 psi
PAC addition = 40 mg/L
177

Membrane Cleaning Processes
One of the main aspects for consideration was the cleaning of fouled membranes and its
impact on permeate flux recovery and process sustainability. There are several cleaning methods
such as physical, chemical, physico-chemical, and biological/biochemical cleaning (Song et al.,
2004; Wang et al., 2014). In this study, physical, chemical, and biochemical cleaning methods
were tested for flux recovery. Backwashing and physical cleaning could remove weakly-bond
deposits on membrane surfaces. Irreversible foulants such as organic matter or bio-solids could be
possibly removed only by the application of chemical cleaning agents such as Triton X-100, a
surfactant; or a bioactive agent, RID-X.
In this study, Membranes A and C were used to investigate the effectiveness of cleaning
methods because they had different GO compositions, namely 0.5% and 1%, respectively. Figure
5.10 compares the permeate recovery after backwashing as well as chemical and biochemical
cleaning for each membrane. Figures 5.10 (a) and (b) present the results of flux recovery for
Membranes A and C, respectively. The initial flux each membrane was 400 L m
-2
hr
-1
and 450 L
m
-2
hr
-1
. Figure 5.10 (a) shows the flux recovery after backwash with DDI water at 3 hours,
cleaning with surfactant at 6 hours, and with enzyme at 9 hours, to be 170 L m
-2
hr
-1
(42.5%
recovery), 275 L m
-2
hr
-1
(68.8%), and 325 L m
-2
hr
-1
(81.3%), respectively. The permeate
recovery for Membrane C is demonstrated in Figure 5.10 (b). As can be observed, flux recoveries
after backwash, surfactant cleaning, and enzyme cleaning were 225 L m
-2
hr
-1
(50%), 350 L m
-2

hr
-1
(77.8%), and 325 L m
-2
hr
-1
(88.9%), respectively. It is note that sodium hydroxide cleaning
caused membrane damage because the initial flux was increased from 400 L m
-2
hr
-1
to 850 L m
-2

hr
-1
. Figures 5.10 (c) and (d) depict the results of flux recovery for Membrane D and E, respectively.
As can be observed in Figure 5.10 (c), flux recoveries after backwash with DDI water, surfactant
178

cleaning, and enzyme cleaning were 37.5%, 53.8% and 53.8%, respectively; and the corresponding
flux recoveries for Membrane E presented in Figure 5.10 (d) were 37.5%, 43.8%, and 43.8%. It is
interesting to note that the recoveries were the same when for surfactant cleaning or enzyme
cleaning in Figures 5.10 (c) and (d).
Figure 5.11 shows the Membranes D and E which are PSSA blends annealed at 165
o
C, and
they were both “transparent films.” Yurdacan (2015) reported that the structures of these
membranes were visually more homogeneous and stronger than the membranes annealed at 70
o
C.
Both membranes exhibited anti-fouling properties as there was no accumulation of foulants and
formation of a gel layer on the membrane surfaces as indicated by the photographs shown in Figure
5.11.
Figures 5.12 (a) and (b) present the trends of TOC concentrations in permeates after
different cleaning processes for Membranes A, C, D, and E. In Figure 5.12 (a), TOC contents in
SCE deposited on the fouled membranes were similarly cleaned by backwash, surfactant, and
enzyme; nevertherless the effect of each method could not be clearly visualized as each flux
recovery was evidently different (Figure 5.10). Membranes A and C might have the similar
physical and chemical resistances to fouling, and this aspect is reflected by the results presented
in Figure 5.12. The study showed trends in terms of foulant removal by each of the cleaning
methods. In overall assessment, all the membranes exhibited strong resistances to fouling, and
amenability to foulant removal by various types of cleaning agents such as surfactant Triton X-
100 and enzyme RID-X.
179

(a) Membrane A

(b) Membrane C

0
50
100
150
200
250
300
350
400
450
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
0
50
100
150
200
250
300
350
400
450
500
0 1 2 3 4 5 6 7 8 9 10 11 12
Permeate Flux (L/m
2
/hr)
Time (hr)
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning
with 5 mg/L
of RID-X
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning with
5 mg/L of
RID-X
180

(c) Membrane D

(d) Membrane E
Figure 5.10 - Flux recovery and filtration after membrane cleaning processes for Membranes A,
C, D, and E

0
50
100
150
200
250
300
350
0 2 4 6 8 10 12 14 16 18 20 22 24
Flux (L/m
2
/hr)
Time (hr)
0
50
100
150
200
250
300
350
400
450
0 2 4 6 8 10 12 14 16 18 20 22 24
Flux (L/m
2
/hr)
Time (hr)
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning with
5 mg/L of
RID-X
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning with
5 mg/L of
RID-X
181

(a) Membrane D (b) Membrane E
Figure 5.11 - Sulfonated membranes annealed at 165
o
C: (a) Membrane D, 30% PSSA; (b)
Membrane E, 35% PSSA

(a)

0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC (mg/L)
Time (hr)
Membrane A
Membrane C
Initial feed = 7.5 mg/L
Trans-membrane pressure = 60 psi
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning with
5 mg/L of
RID-X
182

(b)

Figure 5.12 - TOC concentration after membrane cleaning processes for Membranes A, C, D, and
E

5.5 Nanofiltration Membranes for Water Reclamation
Nanofiltration is widely used in wastewater treatment and water reclamation applications
followed by ultrafiltration and reverse osmosis. Microfiltration is often used as a pretreatment to
protect the nanofiltration or reverse osmosis membranes from organic and biological fouling.
Nonetheless, nanofiltration membranes and reverse osmosis membranes are highly prone to
inorganic fouling or inorganic scaling due to deposition of inorganic precipitates including calcium
carbonate, magnesium sulfate, and several others. In the context of inorganic fouling or scaling,
the membrane must be capable of repulsion of cations such as calcium and magnesium ions as
well as anions such as sulfates and carbonates so that the scaling due to precipitation of the
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20 22 24
TOC (mg/L)
Time (hr)
Membrane D
Membrane E
Initial TOC in feed=14.5mg/L
Trans-membrane pressure=60 psi
Backwash with
DDI
Cleaning with
5 mg/L of
Triton X-100
Cleaning with
5 mg/L of
RID-X
183

membrane is substantially reduced. We are looking at membrane polymer and nanomaterial
formulations to reduce inorganic scaling by taking advantage of surface charge mechanisms.
An important consideration in the formulation of nanofiltration membranes (or for that
matter reverse osmosis membranes) using polymer blends with the infusion or impregnation of
nanomaterials is the reduction of all types of inorganic fouling. In wastewater treatment and water
reclamation applications, the organic fouling potential and rejection characteristics of
nanofiltration membranes used for wastewater treatment are based on the molecular size and
weight ranges, charge effects and the hydrophilicity as well as the hydrophobicity of the following
organic constituents that constitute dissolved organic carbon (DOC) as discussed extensively by
Imai et al (2002). The six main dissolved organic matter constituents are the following: (1) Aquatic
humic substances (AHS), (2) Hydrophobic bases (HOB), (3) Hydrophobic neutrals (HON), (4)
Hydrophilic acids (HIA), (5) Hydrophilic bases (HIB), and (6) Hydrophilic neutrals (HIN). The
following aspects must be carefully considered from the standpoint of organic fouling of
membranes, particularly nanofiltration membranes that are best suited to remove them:
 The AHS and HIA constitute 55% of dissolved organic carbon (DOC)
 The average fractional ranges of HIA is 32-75 % of DOC
 The average range of AHS IS 2-28 % OF DOC
 The average range of HN is 0-12 % of DOC
 The AHS are negatively charged species
 The exo-polymeric substances (EPS) from microorganisms are hydrophobic and cover a
molecular range of 380-850 Daltons.

184

5.6 Application of Polymers for Membrane Fabrication
The most widely used polymeric materials for fabricating membranes used in conventional
applications are the following, as previously discussed: cellulose acetate, nitrocellulose, and
cellulose esters (CA, CN, and CE), polytetra-fluoro-ethylene (PTFE), polyvinylidene fluoride
(PVDF), polypropylene (PP), polyacrylonitrile (PAN), polyamides, polyimides, polyvinyl
chloride (PVC), polysulfone (PS), polyether sulfone(PES) polyether sulfone (PES), polyethylene
and polypropylene (PE and PP).
Several studies have been conducted to increase the hydrophilic characteristics of polymers,
particularly in the case of PS and PES membranes (Richards et al., 2012). There techniques have
been employed in this regard: (1) blending polymers like PS with hydrophilic nanoparticles such
as the following oxides, SiO2, ZnO2, and TiO2; (2) grafting the polymer such as PS or PES with
more hydrophilic polymers, monomers, or functional groups; and (3) coating the polymer with
more hydrophilic polymers. Blending of polymers affords the advantage of producing membranes
with excellent separation performance, favorable aqueous transport characteristics, good chemical
resistance as well as thermal resistance and pH tolerance, and general adaptability to harsh
environments in wastewater treatment and water reclamation situations (Richards et al., 2012).
An important aspect in the membrane development scheme is the consideration new
polymer formulations. For example, the application of polyvinylidene difluoride (PVDF) and
polystyrene with and without graphene or graphene oxide may be considered for membrane
development. Such polymer and copolymer blends have hydrophobic and hydrophilic functional
groups, with the latter on the outer surfaces of the polymer films. The purpose of using these types
of specific polymer blends is to attain the favorable micro-structural, mechanical, transportive, and
fouling resistant properties of the polymer components. The use of polymer blends such as those
185

of PVDF and polystyrene shall increase the mechanical strength, robustness and durability of the
fabricated membranes. The variation in blending and composition shall also afford flexibility and
tunability of the membrane pore sizes and charge effects to enhance transport properties of water
molecules and at the same time reduce fouling potential by promoting repulsion of foulants ---
organic, inorganic, and microbiological. The reduction in fouling refers to both surface fouling
and internal pore fouling of the formulated membranes. The polymer formulation shall also
enhance the mechanical strength and the overall durability and longevity, chemical tolerance, pH
tolerance and cleanability (using various chemical agents), besides micro-structural integrity of
the membrane.

5.7 Application of Nanomaterials in Membrane Matrices
In recent years, various nanoparticles have been incorporated in membranes such as those
of metals or metal oxides or non-metal oxides including TiO2, SiO2, Al2O3, ZnO2, ZrO2, Ag and
Fe, besides zeolites as well as single and multiwalled carbon nano-tubes (CNTs) for various
applications including wastewater treatment. These membranes can remediate two types of fouling,
namely, organic fouling due to natural organic matter and synthetic organic matter, and biological
fouling due to microorganisms and their exudates including exo-polymeric substances (EPS), and
these aspects are extensively reviewed by Richards et al (2012). These nano-particles also
contribute to improvements in aqueous transport characteristics. Li et al. (2006) showed that the
water flux through the PES-TiO2 membrane was significantly enhanced by the inclusion of the
titanium dioxide nanoparticles, but the flux was dependent on the nanoparticle concentrations.
They also observed that owing to their high diffusivity, nanoparticles exist only for a short time
and are susceptible to rapid agglomeration. These nanoparticles can also control biological fouling
186

due to their bactericidal and biocidal effects, and silver nanoparticles are most commonly used for
this purpose (Zodrow et al., 2009; Richards et al, 2012). Maximous et al. (2009, 2010) had
investigated polyether membranes containing nanoparticles aluminum and zirconium oxides
(Al2O3 and ZrO2), and observed increased hydrophilic characteristics towards enhanced aqueous
transport and fouling resistance than plain polymeric membranes. Bae and Tak (2005) observed
that titanium dioxide (TiO2) not only adsorbed on to the membrane surface but to the membrane
pores as well, so that their concentrations beyond a certain level caused reduced membrane
permeability and increased filtration resistance, a factor that highlighted the importance of
nanoparticle concentration in membrane synthesis. Nevertheless, the general observation of these
studies was that metal or metal oxide nanoparticles had higher affinity for water molecules than
plain polymeric membranes and so their impregnation in polymer matrices had positive
contributions towards water flux and fouling control. The use of carbon-based nanomaterials such
as CNTs, graphene (G) and graphene oxide (GO) can have a more positive impact on polymeric
membrane performances in terms of aqueous transport and fouling control. The use of GO in the
impregnation of polymeric matrices will be expected have a larger impact on account of some of
its inherent characteristics discussed here. The function of nanomaterials such as G or GO to the
polymer formulations in various concentrations is to facilitate ant-friction micro-flow water
transport, promote biocidal characteristics (regarding resistance to biological fouling and
destruction of pathogenic organisms), resistance to organic fouling and inorganic fouling or scaling,
and chemical stability, besides irreversible oxidation of or reduction of certain contaminants in
water.
The application of nanomaterials other than graphene oxide and graphene are also
considered to cover a spectrum of multifaceted properties. These properties include high
187

hydrophilicity, super-paramagnetic properties, antibacterial properties, flexible chemical
functionalities and strong hydration capability. These nano-materials shall include metals, metal
oxides, semi-conductor oxides, carbon nanotubes, polymers, and others. However the general
concept is that graphene oxide and graphene related nanoparticles may be far superior to these
nanomaterials in most membrane separation applications.
Engineered graphene and graphene oxide (G and GO) have demonstrated significant
potential for ultrathin, ultrafast, and yet precise sieving membrane for separation of gas molecules
and aqueous ions (Jiang et al., 2015). Observations of intrinsic anti-microbial material properties
further highlight the use of graphene based carbon materials for water treatment membranes with
anti-microbial and fouling resistant properties such as the destruction of Escherichia coli (Perrault
et al., 2013; Jiang et al., 2015). Flat GO membranes (using GO without any polymer matrix) have
shown about 4-10 times the water flux observed in commercial nanofiltration membranes (Perrault
et al., 2013). Crumpled GO as used in the present study is structurally three-dimensional as
compared to GO sheets, while remaining intrinsically porous in the polymer matrix. The crumpled
GO structures have physical defects in the form of vacancies and holes with high ridges and low
valleys, readily forming nanoscale channels and pathways for potential rapid water transport and
permeation. The crumpled GO cross-linked with polysulfone, polyamide or any typical polymeric
material used in membrane fabrication imparts hydrophilic characteristics owing to the abundance
of hydrophilic functional groups exemplified by hydroxyl and carboxyl groups (–OH and –
COOH). The assembled GO-polymer membranes have an effective pore-size of less than 10 nm
in the nanofiltration range, making it suitable for removing macro-molecular organics, colloidal
and biological constituents, and several ions though a combination of mechanism – size, exclusion,
depth filtration, and surface charge effects (Jiang et al., 2015). These membranes will be suitable
188

for several applications including water treatment, wastewater treatment, water reclamation and
industrial separations. More importantly, the modification of hydrophilic groups such as the
sulfonic groups will impart the fine tunability of membrane properties such as pore-size, molecular
weight cut-off, hydration capability, aqueous transport rates and fouling resistance.

5.8 Membrane Fouling Control and Cleaning Strategies
In the context of membrane synthesis and development, factors influencing membrane
fouling in integrated membrane processes are important. These factors include biomass, colloids,
natural organics, inorganic precipitates or scalants, and extracellular polymers; and are dependent
upon process operating conditions. Membrane fouling in such systems is attributed to the
following causes: (i) macromolecular and colloid sorption, (ii) biofilm growth and attachment; and
(iii) inorganic precipitation or scaling (Tsai et al., 2004; Williams and Pirbazari, 2007; Ravindran
et al., 2009). Fouling is generally caused by absorption of hydrophobic compounds onto and within
membrane pores, and deposition of cake or gel layer on the membrane surface. Biological fouling
is predominantly caused by extracellular polymeric substances (EPS) that mainly consist of
carbohydrates, proteins, humic substances and nucleic acids, constituting the infra-structure for
bacterial floc and biofilm formation (Williams and Pirbazari, 2007; Ravindran et al., 2009).
Permeate flux decline due to concentration polarization and membrane fouling can be mitigated
by employing powder activated carbon (PAC) adsorbent and fluid management (Williams and
Pirbazari, 2007; Ravindran et al., 2009; Williams et al., 2012). The PAC depolarizes dissolved
biological and organic matter and re-entrains colloids and suspended solids from the viscous sub-
layer. It adsorbs most organic and bio-organic foulants in wastewaters and reclaimed waters,
including humic substances, proteins, carbohydrates and fats (Kilduff and Weber, 1992,
189

Environmental Science and Technology, 26 (3), 569-577, 1992). The PAC also reduces the
thicknesses of mass-transfer and hydrodynamic boundary layers, lowers concentration polarization,
and controls gel deposition on membrane surfaces or pores (Pirbazari et al., 1996; Tsai et al., 2004;
Ravindran et al., 2009).
Membrane fouling can be strategically controlled by concentration polarization
suppression, optimization of physical and chemical cleaning protocols, and pre-treatment of feed
(Tu et al., 2005; Williams and Pirbazari, 2007). Membrane cleaning strategies are important for
fouling control, flux recovery, and rejection maintenance. The efficiencies of various cleaning
agents such as alkalies, acids, surfactants, redox chemicals, and enzymes may be evaluated for
removing hydrophobic compounds and substances that are major contributors to organic, bio-
organic and in organic fouling. The chemical tolerance of the membranes to these cleaning agents
may also be determined.
Graphene oxide (GO), a nano sized material synthesized by the controlled oxidation of
graphite, is an inexpensive and commercially available material. Typically, nano sized structures
such as GO will have size ranges of less than a micron. In other refinements, nano sized structures
may have size ranges of less than 500 nm, less than 100 nm, or on the order of several nanometers.
In still other refinements, the size ranges may be greater than 3 angstroms, or greater than 5
angstroms. GO includes partially oxidized graphene bearing carboxyl, hydroxyl, and epoxide
functional groups that render it water-soluble. GO is also soluble in several organic solvents and
is known to have a strong chemical and mechanical stability (Dikin et al., 2007; Suk et al., 2010a).
It must be noted that GO also has antibacterial properties (Hu et al., 2010; Liu et al., 2011). GO,
when used as an active layer or when mixed within a polysulfone (PS) membrane, has been shown
to be only suitable as a nanofiltration and microfiltration membrane due to lack of blocking
190

monovalent ions (Hu and Mi, 2013; Ganesh et al., 2013), likely since graphene sheets are separated
by approximately 1 nm (Nair et al., 2012).

5.9 Development of Novel Polymeric Nanofiltration Membranes
The work discussed in this chapter is intended to develop polymer-nanomaterial composite
membranes for water reclamation and water reuse applications. The applications of these
membranes might at a later stage be extrapolated to wastewater treatment, water purification and
other industrial uses. In the initial stages, the purpose was to develop and synthesize flat-sheet
nanofiltration membranes because they are the most widely used class in these applications. The
ongoing research in progress and future work involves the development of suitable polymers and
polymer blends infused with nanomaterials exemplified by graphene oxide and graphene
derivatives.
The fundamental idea is to enhance the aqueous transport, fouling resistance, rejection
characteristics, chemical cleanability, and mechanical durability of the membranes to make them
far superior to existing ones in performance levels. The polymer and nanomaterial combinations
are intended to have fine levels of tunability of membrane characteristics including pore sizes,
pore-size distributions, charge effects, rejection characteristics and fouling resistance besides
chemical tolerance and tunability. These variations can be achieved by altering the functional
group properties of the polymeric matrices and the nanomaterial particle surface. The concepts
applicable to nanofiltration membranes may later be extrapolated to other classes of membranes
including those pertaining to microfiltration, ultrafiltration and reverse osmosis processes.
A significant aspect that is ignored in the synthesis and development of nano-composite
polymeric membranes is the design of membrane filtration tests, and cleaning protocols. Several
191

cycles of cleaning with different combinations of cleaning agents have to be performed elaborately
for evaluating the membrane permeate flux and rejection characteristics, performance recovery,
and sustainability. The acquisition of experimental data from these elaborate testing methods to
examine the membrane aqueous permeability and permeate fluxes, rejection characteristics for
target contaminant or species, fouling potentials and anti-fouling resistances, chemical tolerance
and cleaning potentials for performance recovery, and maintenance of chemical and mechanical
structural integrity of the composite material. It must be noted that these membranes must be
amenable to long-standing, durable and economic applications. These tests constitute an integral
part of membrane synthesis and development, involving the right formulations and combinations
of polymers or polymer blends with nanomaterials (graphene and graphene derivatives like
graphene oxide). The optimal compositions of membrane materials for the specific applications
discussed above will directly hinge on these elaborately designed performance evaluation tests.

5.9.1 Results and Discussion
Permeate Flux Pattern and TOC Removal
Filtration tests for flat sheet NF membranes were conducted using the UF permeate as feed.
Figure 5.13 shows permeate flux pattern as a function of time for a TMP of 70 psi. The flux was
initially 58.1 L m
-2
hr
-1
and it gradually dropped to 44.5 L m
-2
hr
-1
after 12 hours, corresponding
to a flux reduction of 23.4%. The flux decline was caused by dissolved organic carbons (DOCs)
because the feed solution was obtained a permeate of a UF process.
It is important to note that the nominal MWCO is 10,000 Daltons corresponding to pore
size range of 4 – 10 nm. Logan et al. (1990) and Shon et al. (2006) reported that most of the DOC
in biologically treated secondary clarifier effluent consisted of organic matter of lower molecular
192

weight than 1,000 Daltons with sizes smaller than about 0.8 nm. Thus, the DOC in the wastewater
cannot be retained by UF membranes, but can be partly retained by NF membranes according to
the size exclusion principle. Figure 5.14 shows the TOC concentrations of the feed and the
permeate. The initial TOC concentration in the feed was 5.05 mg L
-1
and it increased slightly to
6.3 mg L
-1
over 12 hours of operation in a batch reactor system. Meanwhile, the permeate TOC
that was initially 2.8 mg L
-1
, subsequently dropped to 2.2 mg L
-1
after 12 hours. The results showed
that the NF90 membrane could effectively remove organic matter (TOC) to the extent of 57%.

Figure 5.13 - Permeate flux pattern for NF flat sheet membranes (same as Figure 3.26)

0
10
20
30
40
50
60
70
0 1 2 3 4 5 6 7 8 9 10 11 12
Flux (L/m
2
/hr)
Time (hr)
NF permeate using UF permeate as feed
Trans-membrane pressure=70 psi
193

Figure 5.14 - TOC concentration for NF flat sheet membranes (same as Figure 3.27)

Effect of Powdered Activated Carbon Addition on Flux and TOC Removal
Activated carbon is widely used in water and wastewater treatment applications to remove
organic compounds and potential pollutants including endocrine disrupting chemicals (EDCs)
and/or pharmaceuticals and personal care products (PPCPs). These pollutants cannot be generally
removed by conventional treatment processes, but can be effectively removed only by the
advanced treatment technologies. An earlier experimental study by Solak et al (2013) had
established that activated carbon could effectively remove some types of EDCs to the extent of
88% to 93%, even at trace concentration levels.
The present study investigated the feasibility of nanofiltration for the removal of potential
contaminants from SCE. It evaluated the TOC removal and membrane flux using NF membranes.
Figure 5.15 shows the permeate flux patterns in the presence and absence of PAC (the
concentration of PAC used was of 15 mg L
-1
). As can be observed, the initial fluxes were similar
to each other, and were approximately 60 L m
-2
hr
-1
. The permeate fluxes in the absence of PAC
were slightly higher than those experienced in the presence of PAC during the first 6 hours of
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC (mg/L)
Time (hr)
Feed characteristics
NF permeate
Initial feed=5.1mg/L
Trans-membrane pressure=70 psi
194

operation. However, after this period, the permeate fluxes exhibited a different patterns, as they
were relatively higher in the presence of PAC. These changes in permeate flux patterns indicated
that after the initial adsorption of organic carbon, the membrane fouling was reduced and the
overall aqueous permeability increased.
Figure 5.16 shows the effect of PAC on the TOC concentration in permeate. Evidently, the
PAC successfully removed the TOC from the treated wastewater leaving a residual of only 0.2 mg
L
-1
. These results demonstrate the excellent potential of PAC use for the removal of a broad
spectrum of organic contaminants.

Figure 5.15 - Effect of 15 mg L
-1
PAC addition on NF permeate flux pattern (same as Figure 3.28)

0
10
20
30
40
50
60
70
0 1 2 3 4 5 6 7 8 9 10 11 12
Flux (L/m
2
/hr)
Time (hr)
NF permeate using UF permeate as feed
NF permeate using UF permeate as feed with PAC
Trans-membrane pressure=70psi
195

Figure 5.16 - Effect of 15 mg L
-1
PAC addition on NF permeate TOC removal (same as Figure
3.29)

Table 5.3 summarizes the water quality of the feed and permeate for the NF90 membrane.
In the filtration tests as stated earlier, the permeate of UF membrane was used to evaluate the
removal of TOC and the anionic species including chloride, nitrate, phosphate, and sulfate. As can
be observed, the anions were effectively removed by the NF90 membrane to the extent of 88% to
100%, except nitrate, whose removal was only 45%. The TOC removal was 56.4% using the NF90
membrane alone. As evident, the presence of PAC improved the removals of TOC and anions. The
TOC removal was increased substantially to 96%, and the residual TOC in the permeate was low
at 0.2 mg L
-1
. The phosphate and sulfate ions were almost completely removed by the combination
of NF membranes and PAC at 100% and 99.6%, while the chloride and nitrate removals were
substantially improved to 94.8% and 71.4%, respectively.

0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12
TOC (mg/L)
Time (hr)
UF permeate TOC=5.05mg/L
NF permeate using UF permeate as feed with PAC
NF permeate using UF permeate as feed
Trans-membrane pressure=70 psi
196

Table 5.3 - Feed and permeate water quality for NF90 membrane filtration experiments
Items SCE
NF feed
(UF permeate)
NF membrane NF membrane and PAC
Steady
state
permeate
Steady state
removal (%)
Steady
state
permeate
Steady state
removal (%)
TOC
(mg/L)
6.7 5.05 2.2 56.4 0.2 96.0
Chloride,
Cl
-

(mg/L)
170 132 15.6 88.2 6.84 94.8
Nitrogen,
NO3-N
(mg/L)
3 2.31 1.27 45.0 0.66 71.4
Phosphorus,
PO4-P
(mg/L)
0.5 0.19 ND* 100 ND* 100
Sulfate,
SO4
2-

(mg/L)
140 108 1.12 98.9 0.44 99.6
References Table 3.1 Fig. 5.13 Fig. 5.15

* ND: not detected

Flux Recovery after Cleaning Processes
A caustic cleaning procedure employing was conducted using an in situ method at an
applied TMP of 70 psi. As it evident, sodium hydroxide (10
-3
M) would effectively remove
organics and microorganisms. After caustic cleaning, the membrane was thoroughly rinsed by DDI
water to remove residuals of the chemical cleaning agent. Figure 5.17 shows the flux recovery
after the application of sodium hydroxide. As can be observed, the permeate flux recovery was
almost 100% clearly indicating the successful removal of membrane fouling.
197

Figure 5.17 - Flux recovery and filtration after membrane cleaning processes for UF permeate
water with NF flat sheet membrane: 1 × 10
-3
M of sodium hydroxide at 12 hours

5.10 Summary and Conclusions
Graphene oxide (GO) can be employed in polymeric matrices to develop superior
membranes for various environmental applications ranging from water and wastewater treatment
to water reclamation and reuse. The present study showed that the incorporation of GO
nanoparticles yielded composite membranes with improved membrane permeability and superior
anti-fouling properties.
The synthesized membranes exhibited superior aqueous permeability and anti-fouling
characteristics. The synthesized membranes tolerated the chemical cleaning agents. The cleaning
agents, however, did not completely recover the membrane permeability, approximately 50%
recovery. However, it was difficult to observe the surface fouling by the naked eyes because the
surface of novel membranes was pretty smooth and transparent. The synthesized membranes
annealed at high temperature, 165
o
C had the potential properties of anti-fouling due to the in
hydrophilic characteristics resulting in mitigation of gel layer formation and surface fouling.
0
10
20
30
40
50
60
70
0 2 4 6 8 10 12 14 16 18 20 22 24
Flux (L/m
2
/hr)
Time (hr)
Feed from UF permeate
Trans-membrane pressure=70 psi
Cleaning with
10
-3
M NaOH
198

Limited nanofiltration membranes were investigated regarding the feasibility of water
reclamation and reuse applications. This study contained significant results in terms of organic
compounds because estimated major foulant was dissolved organic compounds as the feed was
obtained from the permeate of UF membrane filtration tests. According to the combination of NF
and PAC processes, they might effectively remove potential trace organic compounds such as
EDCs and PPCPs.
Caustic chemicals such as sodium hydroxide were the most effective cleaning agents in
terms of flux recovery because they could remove organic and biological foulants and biofilms
deposited on the membrane surfaces.

5.11 Future Work
One of the important objectives of the present work is to develop novel membranes
involving the impregnation of graphene oxide and graphene-related nanoparticles into polymer
matrices specially formulated, so that the synthesized membranes have superior characteristics
regarding aqueous permeability and transport, rejection properties, fouling resistance, chemical
tolerance, and cleanability. These novel membranes are intended to manifest greater material
stability and durability for sustained applications without compromising their separation
capabilities. In the present context, the NF90 nanofiltration membranes (manufactured by Filmtec,
Dow Chemical Company) are widely recognized as one of the best (or the best) commercially
available membranes for water reclamation and reuse application among various other uses. The
ongoing and future research activities are directed towards the synthesis and development of new
membrane that are distinctly superior to the best commercially available membranes, particularly
in the area of water reclamation and reuse.
199

5.12 References
Ai, L., Zhang, C., & Chen, Z. (2011). Removal of methylene blue from aqueous solution by a
solvothermal-synthesized grapheme/magnetite composite. Journal of Hazardous Materials,
192(3), 1515-1524.

APHA, AWWA, & WEF (2005). Standard Methods for the Examination of Water and Wastewater,
21st ed., American Public Health Association, Washington, D.C.

Asatekin, A., Menniti, A., Kang, S., Elimelech, M., & Mayes, A. M. (2006). Antifouling
nanofiltration membranes for membrane bioreactors from self-assembling graft polymers. Journal
of Membrane Science, 285(1-2), 81-89.

Asatekin, A., Kang, S., Elimelech, M., & Mayes, A. M. (2007). Anti-fouling ultrafiltration
membranes containing polyacrylonitrile-graft-poly(ethylene oxide) comb copolymer additives.
Journal of Membrane Science, 298(1-2), 136-146.

Bae, T. & Tak, T. (2005). Effect of TiO2 nanoparticles on fouling mitigation of ultrafiltration
membranes for activated sludge filtration. Journal of Membrane Science, 249(1-2), 1-8.

Bai, J., Zhong, X., Jiang, S., Huang, Y., & Duan, X. (2010). Graphene nanomesh. Nature
Nanotechnology, 5(3), 190-194.

Booth, T. J., Blake, P., Nair, R. R., Jiang, D., Hill, E. W., Bangert, U., Bleloch A, & Geim, A. K.
(2008). Macroscopic graphene membranes and their extraordinary stiffness. Nano letters, 8(8),
2442-2446.

Chen, H., & Belfort, G. (1999). Surface modification of [poly(ether sulfone) ultrafiltration
membranes by low-temperature plasma-induced graft polymerization. Journal of Applied Polymer
Sciences, 72(13), 1699-1711.

200

Chennamsetty, R. & Escobar, I. C. (2008). Evolution of a polysulfone nanofiltration membrane
following ion beam radiation. Langmuir, 24(10), 5569-5579.

Choi, J. H., Jegal, J., & Kim, W. N. (2006). Fabrication and characterization of multi-walled carbon
nanotubes/polymer blend membranes. Journal of Membrane Science, 284(1-2), 406-415.

Choi, J. H., Jegal, J., & Kim, W. N. (2007). Modification of performances of various membranes
using MWNTs (multiwalled nanotubes) as a modifier. Macromolecular Symposia, Advanced
Polymers for Emerging Technologies, 249 –250, 610–617.

Choi, W., Choi, J., Bang, J., & Lee, J. H. (2013). Layer-by-layer assembly of graphene oxide
nanosheets on polyamide membranes for durable reverse-osmosis applications. ACS Applied
Materials and Interfaces, 5(23), 12510-12519.

Crittenden, J. C., Trussell, R. R., Hand, D. W., Howe, K. J., & Tchobanoglous, G. (2005). Water
Treatment Principles on Design, Second Edition (revised), John Wiley, Hoboken, New Jersey.

Dikin, D. A., Stankovich, S., Zimney, E., Piner, R. D., Domett, G. H. B., Evmenko, G., Nguyen,
S. T., & Ruoff, R. S. (2007). Preparation and characterization of graphene oxdide paper. Nature,
448(7152), 457-460.

Dong, H. B., Xu, Y. Y., Yi, Z., & Shi, J. L. (2009). Modification of polysulfone membranes via
surface-initiated atom transfer radical polymerization. Applied Surface Science, 255(21), 8860-
8866.

Dreyer, D. R., Park, S., Bielawski, C. W., & Ruoff, R. S. (2010). The chemistry of graphene oxide.
Chemical Society Reviews, 39(1), 228-240.

Fischbein, M. D. & Drndic, M. (2008). Electron beam nanosculpting of suspended graphene sheets.
Applied Physics Letters, 93(11), 113107-113107.

201

Fornasiero, F., Park, H. G., Holt, J. K., Stadermann, M., Grigoropoulos, C. P., Noy, A., & Bakajin,
O. (2008). Ion exclusion by sub-2-nm carbon nanotube pores. Proceedings of the National
Academy of Sciences, 105(45), 17250-17255.

Fritzmann, C., Löwenberg, J., Wintgens, T., & Melin, T. (2007). State-of-the-art of reverse
osmosis desalination. Desalination, 216(1), 1-76.

Ganesh, B. M., Isloor, A. M., & Ismail, A. F. (2013). Enhanced hydrophilicirty and salt rejection
study of graphene oxide-polysulfone mixed matrix membrane. Desalination, 313, 199-207.

Geise, G. M., Lee, H-S., Miller, D. J., Freeman, B. D., McGrath, J. E., & Paul, D. R. (2010). Water
purification by membranes: The role of polymer science. Journal of Polymer Science Part B:
Polymer Physics, 48(15), 1685-1718.

Greenlee, L. F., Lawler, D. F., Freeman, B. D., Marrot, B., & Moulin, P. (2009). Reverse osmosis
desalination: Water sources, technology and today’s challenges. Water Research, 43, 2317-2348.

Hashimoto, A., Suenaga, K., Gloter, A., Urita, K., & Iijima, S. (2004). Direct evidence for atomic
defects in graphene layers. Nature, 430(7002), 870-873.

Hausman, R., Gullinkala, T., & Escobar, I. C. (2009). Development of low-biofouling
polypropylene feed spacers for reverse osmosis. Journal of Applied Polymer Science, 114(5),
3068-3073.

Holt, J. K., Park, H. G., Wang, Y., Stradermann, M., Artyukhin, A. B., Grigoropoulos, C. P., Noy,
A., & Bakajin, O. (2006). Fast mass transport through sub–2-nanometer carbon nanotubes, Science,
312, 1034-1037.

Hu, M. & Mi, B. (2013). Enabling graphene oxide nanosheets as water separation membranes.
Environmental Science & Technology, 47(8), 3715-3723.

202

Hu, W., Peng, C., Luo, W., Lv, M., Li, X., Li, D., Huang Q., & Fan, C. (2010). Graphene-based
antibacterial paper. ACS Nano, 4(7), 4317-4323.

Hydronautics Press Release (2007). Integrated membrane solutions at work in Southern Spain;
(available from: http://www.membranes.com/press/Escombreras), released on January 20, 2007.

Imai, A., Fukushima, T., Matsushinge, K., Kim, Y. H., & Choi, K. (2002). Characterization of
dissolved organic matter in effluents from wastewater treatment plants. Water Research, 36(4),
859-870.

Inui, N., Mochiji, K., Moritani, K., & Nakashima, N. (2010). Molecular dynamics simulations of
nanopore processing in a graphene sheet by using gas cluster ion beam. Applied Physics A, 98(4),
787-794.

Jiang, D. E., Cooper, V. R., & Dai, S. (2009). Porous graphene as the ultimate membrane for gas
separation. Nano letters, 9(12), 4019-4024.

Jiang, Y., Wang, W. N., Liu, D., Nie, Y., Li, W., Wu, J., Zhang, F., Biswas, P., & Fortner, J. D.
(2015). Engineered crumpled graphene oxide nanocomposite membrane for advanced water
treatment processes. Environmental Science and Technology, 49(11), 6846-6854.

Joshi, R. K., Carbone, P., Wang, F. C., Kravets, V. G., Su, Y., Grigorieva, I. V., Wu, H. A., Geim,
A. K., & Nair, R. R. (2014). Precise and ultrafast molecular sieving through grapheme oxide
membranes. Science, 343(6172), 752-754.

Kilduff, J. E. & Weber, Jr., W. J. (1992). Transport and separation of organic molecules in
ultrafiltration membranes. Environmental Science and Technology, 26(3), 569-577.

Kim, J., & Van der Bruggen, B. (2010). The use of nanoparticles in polymeric and ceramic
membrane structures: Review of manufacturing procedures and performance improvement for
water treatment. Environmental Pollution, 158(7), 2335-2349.
203

Kim, K. S., Lee, K. H., Cho, K., & Park, C. E. (2002). Surface modification of polysulfone
ultrafiltration membranes by oxygen plasma treatment, Journal of Membrane Science, 199(1-2),
135-145.

Kimura, S. & Sourirajan, S. (1967). Analysis of data in reverse osmosis with porous cellulose
acetate membranes used. AIChE Journal, 13(3), 497-503.

Koeing, S. P., Wang, L., Pellegrino, J., & Bunch, J. S. (2012). Selective molecular sieving through
porous grapheme. Nature Nanotechnology, 7(11), 728-732.

Kong, C., Shintani, T., & Tsuru, T. (2010). Pre-seeding-assisted synthesis of a high performance
polyamide-zeolite nano-composite membrane for water purification. New Journal of Chemistry,
34, 2101–2104.

Krasheninnikov, A. V., Nordlund, K., & Keinonen, J. (2002). Production of defects in supported
carbon nanotubes under ion irradiation. Physical Reviews B, 65(16), 165423.

Krasheninnikov, A. V., Nordlund, K., Sirviö, M., Salonen, E., & Keinonen, J. (2001). Formation
of ion-irradiation-induced atomic-scale defects on walls of carbon nanotubes. Physical Review B,
63(24), 245405.

Kull, K. R., Steen, M. L., & Fisher, E. R. (2005). Surface modification with nitrogen-containing
plasmas to produce hydrophilic, low-fouling membranes. Journal of Membrane Science, 246, 203-
215.

Li, H., Cao, Y., Qin, J., Jie, X., Wang, T., Lin, J., & Yuan, Q. (2006). Development and
characterization of anti-fouling cellulose hollow fiber UF membrane for oil-water separation.
Journal of Membrane Science, 279(1-2), 328-335.

204

Li, H., Gao, Y., Pan, L., Zhang, Y., Chen, Y., & Sun, Z. (2008a). Electrosorptive desalination by
carbon nanotubes and nanofibres electrodes and ion-exchange membranes. Water Research, 42,
4923–4928.

Li, L., Dong, J., Nenoff, T. M., & Lee, R. (2004). Desalination by reverse osmosis using MFI
zeolite membranes. Journal of Membrane Science, 243(1), 401-404.

Li, L., Yan, G., & Wu, J. (2009). Modification of polysulfone membranes via surface-initiated
atom transfer radical polymerization and their antifouling properties. Journal of Applied Polymer
Science, 111(4), 1942-1946.

Li, L., Yan, G., Wu, Y., Yu, X., & Guo, Q. (2008b). Surface-initiated atom transfer radical
polymerization from polyimide films and their anti-fouling properties. Journal of Macromolecular
Science, Part A: Pure and Applied Chemistry, 45(10), 828-832.

Li, M. (2014). Development of polystyrene sulfonic acid-polyvinylidene fluoride (PSSA-PVDF)
belnds for direct methanol fuel cells (DMFCS) (Doctoral dissertation). University of Southern
California.

Lind, M. L., Ghosh, A. K., Jawor, A., Huang, X., Hou, W., Yang, Ya., & Hoek, E. M. V. (2009).
Influence of zeolite crystal size on zeolite-polyamide thin film nanocomposite membranes.
Langmuir, 25(17), 10139-10145.

Liu, S., Zeng, T. H., Hofmann, M., Burcombe, E., Wei, J., Jiang, R., Kong, J., & Chen, Y. (2011).
Antibacterial activity of graphite, graphite oxide, grapheme oxide, and reduced grapheme oxide:
Membrane and oxidative stress. ACS Nano, 5(9), 6971-6980.

Loeb, S. & Sourirajan, S. (1963). Seawater demineralization by means of an osmotic membrane.
Journal of American Chemical Society, 117–132.

205

Logan, B. E. & Jiang, Q. (1990). Molecular size distributions of dissolved organic matter, Journal
of Environmental Engineering, 116, 1046-1062.

Lucchese, M. M., Stavale, F., Ferreira, E. H., Vilani, C., Moutinho, M. V. O., Capaz, R. B., Achete
C.A., & Jorio, A. (2010). Quantifying ion-induced defects and Raman relaxation length in
graphene. Carbon, 48(5), 1592-1597.

Maruf, S. H., Ahn, D. U., Pellegrino, J., Killgore, J. P., Greenberg A. R., & Ding, Y. (2012).
Correlation between barrier layer Tg (glass temperature) and a thin-film composite polyamide
membrane performance: effect of chlorine treatment. Journal of Membrane Science, 405 –406,
167–175.

Matsuura, T. (1994). Synthetic membranes and membranes separation processes. CRC press,
Boca Raton, Florida.

Maximous, N., Nakhle, G., Wan, W., & Wong, K. (2009). Preparation, characterization and
performance of Al2O3/PES membrane for wastewater filtration. Journal of Membrane Science,
341(1-2), 67-75.

Maximous, N., Nakhle, G., Wan, W., & Wong, K. (2010). Performance of a novel characterization
ZrO2/PES membrane for wastewater filtration. Journal of Membrane Science, 352(1-2), 222-230.

Nair, R. R., Wu, H. A., Jayaram, P. N., Grigorieva, I. V., & Geim, A. K. (2012). Unimpeded
permeation of water through helium-leak–tight graphene-based membranes. Science, 335(6067),
442-444.

Peng, W., Escobar, I. C., & White, D. B. (2004). Effects of water chemistries and properties of
membrane on the performance and fouling: A0model development. Journal of Membrane Science,
238(1-2), 33-46.

206

Peng, W. & Escobar, I. C. (2005). Evaluation of factors influencing membrane performances.
Environmental Progress, 24(4), 382-399.

Perreault, F., Tousley, M. E., & Elimelech, M. (2013). Thin-film composite polyamide membranes
functionalized with biocidal graphene oxide nanosheets. Environmental Science and Technology
Letters, 1(1), 71-76.

Pieracci, J., Crivelle, J. V., & Belfort, G. (2002). Increasing membrane permeability of UV-
modified poly (ether sulfone) ultrafiltration membranes. Journal of Membrane Science, 202(1-2),
1-16.

Pirbazari, M., Kim, S. H., Badriyha, B. N., & Ravindran, V. (1996). Hybrid membrane filtration
process for leachate treatment. Water Research, 30(11), 2691-2706.

Rahimpour, A., Madaeni, S. S., Zereshki, S., & Mansourpanah, Y. (2009). Preparation and
characterization of modified nano-porous PVDF membrane with high antifouling property using
UV photo-grafting. Applied Surface Science, 255(16), 7455-7461.

Rajbartoreh, A. R., Wang, B., Shen, X., & Wang, G. (2011). Advanced mechanical properties of
graphene paper. Journal of Applied Physics, 109(1), 014306-014306.

Ravindran, V., Tsai, H. H., Williams, M. D., Pirbazari, M. (2009). Hybrid membrane bioreactor
technology for small water treatment utilities: process evaluation and primordial considerations.
Journal of Membrane Science. 344(1-2), 39-54.

Richards, H. L., Baker, P. G. L., & Iwuoha, E. (2012). Metal nanoparticle modified polysulfone
membranes for use in wastewater treatment: A critical review. Journal of Surface Engineered
Materials and Advanced Technology, 2, 183-193.

207

Shon, H. K., Kim, S. H., Erdei, L. & Vigneswaran, S. (2006). Analytical methods of size
distribution for organic matter in water and wastewater. Korean Journal of Chemical Engineering,
23(4), 581-591.

Sint, K., Wang, B., & Král, P. (2008). Selective ion passage through fuctionalized graphene
nanopores. Journal of the American Chemical Society, 130(49), 16448-16449.

Solak, S., Vakondios, N., Tzatzimaki, I., Diamadopoulos, E., Arda, M. Kabay, N. & Yuksel, M.
(2013). A comparative study of removal of endocrine disrupting compounds (EDCs) from treated
wastewater using highly crosslinked polymeric adsorbents and activated carbon. Journal of
Chemical Technology and Biotechnology, 89(6), 819-824.

Song, W., Ravindran, V., Koel, B. E., & Pirbazari, M (2004). Nanofiltration of natural organic
matter with H2O2/UV pretreatment: fouling mitigation and membrane surface characterization,
Journal of Membrane Science, 241, 143-160.

Su, Y. L., Cheng, W., Li, C., & Jiang, Z. (2009). Preparation of antifouling ultrafiltration
membranes with poly(ethylene glycol)-graft-polyacrylonitrile copolymers. Journal of Membrane
Science, 329(1-2), 246-252.

Su, Y., Li, C., Zhao, W., Shi, Q., Wang, H., Jiang, Z., & Zhu, S. (2008). Modification of
polyethersulfone ultrafiltration membranes with phosphorylcholine copolymer can remarkably
improve the antifouling and permeation properties. Journal of Membrane Science, 322(1), 171-
177.

Suk, M. E. & Aluru, N. R. (2010b). Water transport through ultrathin grapheme. The Journal of
Physical Chemistry Letters, 1, 1590-1594.

Suk, J. W., Piner, R. D., An, J., & Ruoff, R. S. (2010a). Mechanical properties of monolayer
graphene oxide. ACS Nano, 4(11), 6557-6564.

208

Tsai, H. H., Ravindran, V., Williams, M. D., & Pirbazari, M. (2004). Membrane bioreactor process
for water denitrification. Journal of Environmental Engineering and Science, 3(6), 507-521.

Tu, S. C., Ravindran, V., & Pirbazari, M. (2005). A pore diffusion model for forecasting the
performance of membrane processes. Journal of Membrane Science, 265(1-2), 29-50.

Van der Bruggen, B. (2009). Comparison of redox initiated graft polymerization and sulfonation
for hydrophilisation of polyethersulfone nanofiltration membranes. European Polymer Journal,
45(7), 1873-1882.

Villaluenga, J. P. G., Khayet, M., Godino, P., Seoane, B. & Mengual, J. I. (2005). Analysis of the
membrane thickness effect on the pervaporation separation of methanol/methyl tertiary butyl ether
mixtures. Separation and Purification Technology, 47, 80-87.

Wang, S., Zhang, Y., Abidi, N., & Cabrales, L. (2009). Wettability and surface free energy of
graphene films. Langmuir, 25(18), 11078-11081.

Wang, Z., Ma, J., Tang, C. Y., Kimura, K., Wang, Q. & Han, X. (2014). Membrane cleaning in
membrane bioreactors: A review. Journal of Membrane Science, 468, 279-307.

Wei, D., Liu, Y., Wang, Y., Zhang, H., Huang, L., & Yu, G. (2009). Synthesis of N-doped
graphene by chemical vapor deposition and its electrical properties. Nano letters, 9(5), 1752-1758.

Wei, X., Wang, R., Li, Z., & Fane, A. G. (2006). Development of a novel electrophoresis-UV
grafting technique to modify PES UF membranes used in NOM removal. Journal of Membrane
Science, 273(1-2), 47-57.

Williams, M. D. & Pirbazari, M. (2007). Membrane bioreactor process for removing
biodegradable organic matter from water. Water Research, 41(17), 3880-3893.

209

Williams, M. D., Ravindran, V., & Pirbazari, M. (2012). Modeling and process evaluation of
membrane bioreactor for removing biodegradable organic matter from water. Chemical
Engineering Science, 84, 494-511.

Xie, X. L., Maia, Y. W., & Zhou, X. P. (2005). Dispersion and alignment of carbon nanotubes in
polymer matrix: a review. Materials Science and Engineering, R 49, 89–112.

Yoon, K., Hsiao, B. S., & Chu, B. (2009). Formation of functional polyethersulfone electrospun
membrane for water purification by mixed solvent and oxidation process. Polymer, 50, 2893-2899.

Yurdacan, H. M. (2015). Investigation of mechanical, thermal and rheological properties of
flourocarbaon functionalized polystyrene; and, Development of increased flow graphene based
polymer nanocomposite membranes (Doctoral dissertation). Retrieved from University of
Southern California Digital Library.

Zanini, S., Muller, M., Riccardi, C., & Orlandi, M. (2007). Polyethylene glycol grafting on
polypropylene membranes for anti-fouling properties. Plasma Chemistry and Plasma Processes,
27(4), 446-457.

Zhang, J., Zou, H., Qing, Q., Yang, Y., Li, Q., Liu, Z., Guo X., & Du, Z. (2003). Effect of chemical
oxidation on the structure of single-walled carbon nanotubes. Journal of Physical Chemistry B,
107(16), 3712-3718.

Zhang, M. G., Nguyen, Q. T., & Ping, Z. (2009b). Hydrophilic modification of poly (vinylidene
fluoride) microporous membranes. Journal of Membrane Science, 327(1-2), 78-86.

Zhang, X., Chen, Y., Konsowa, A. H., Zhu, X., & Crittenden, J. C. (2009a). Evaluation of an
innovative polyvinyl chloride (PVC) ultrafiltration membrane for wastewater treatment.
Separation and Purification Technology, 70(1), 71-78.

210

Zhu, L. P., Dong, H. B., Yi, Z., & Zhu, B. K. (2009). Amphiphilic PPESK-graft-P(PEGMA)
copolymer for surface modification of PPESK membranes. Materials Chemistry and Physics,
115(1), 223-228.

Zodrow, K., Brunet, L., Mahendra, S., Li, D., Zhang, A., Li, Q., & Alvarez, P. J. (2009).
Polysulfone ultrafiltration membranes impregnated with silver nanoparticles. Water Research,
43(3), 715-723.

211

Chapter 6
Summary, Conclusions, and Recommendation
6.1 Summary and Conclusions
6.1.1 General
 Membrane separations promise to yield substantial environmental and economic benefits
leading to enhanced global competitiveness by significantly reducing energy consumption,
increasing industrial productivity, lowering waste generation, and addressing global water
shortage problems.
 Membrane technologies are widely used in water and wastewater treatment and water
reclamation and reuse. Their application towards water reclamation and reuse is
specifically important owing to scarcity of water resources and water shortages.
 Membrane technologies have to overcome several challenges such as permeate flux decline
and membrane fouling to maintain performance efficiencies regarding permeability,
rejection characteristics and durability.
 The research investigated the permeate flux patterns and the TOC removals for each
membrane. Furthermore, this study aimed to evaluate the effects of activated carbon and
microorganisms on the permeate flux patterns and the TOC removals.

6.1.2 Flat Sheet UF Membranes and Modeling
 The study showed that membrane fouling caused rapid permeate flux decline, and that the
flux deterioration was due to different classes of foulants that led to surface fouling and
internal pore fouling.
212

 The TOC removal using the UF membrane was 10-20% because the secondary clarifier
effluent contained relatively small molecular weight dissolved organic matter (less than
10,000 Daltons molecular weight cutoff). However, the PAC application increased the
TOC removal to 40-60%.
 The cleaning agents, namely a caustic chemical, surfactant, and biological enzyme were
tested to evaluate the flux recovery due to membrane cleaning. Sodium hydroxide was
more effective than Trion X-100 (surfactant) and RID-X (enzyme) as the former removed
both surface fouling and internal pore fouling.
 The wastewater was preoxidized by ozone, hydrogen peroxide, and peroxone (a
combination of ozone and peroxide) to evaluate their effects on permeate flux and organic
(TOC) removal. The pre-oxidation enhanced permeate fluxes but did not reduce the TOC
removal because the oxidants degraded larger organic molecules into smaller ones that
passed through the membrane pores.
 Mathematical models were used to predict the permeate flux and TOC removal. The
permeate flux model is based on the resistances of the membrane material, gel layer and
concentration polarization layer as well as internal pore blocking. Both models predicted
the experimental results with a reasonable degree of accuracy.
 Model sensitivity studies regarding permeate fluxes were conducted for several parameters
including the diffusion coefficient and gel layer concentration. The gel layer concentration
significantly affected the permeate fluxes owing to increase in the gel resistance. The effect
of the diffusion coefficient was insignificant.

213

6.1.3 Hollow Fiber UF Membrane Bioreactor Experiments and Modeling
 Feasibility of scale-up was evaluated by comparing the performance of the micro-scale and
mini-pilot-scale hollow fiber membrane modules. Micro-scale experiments provided a
guideline for upscaling with factors such as permeability, rejection rate, and kinetic
parameter. Permeate fluxes were relatively proportional to membrane surface areas and
trans-membrane pressures in both cases. However, the TOC removal was affected by a
trans-membrane pressure; it was higher as the TMP increased due to the increase
compaction of the gel layer.
 The membrane bioreactor experiments showed that the combination of PAC and
microorganisms system manifested significantly higher TOC removals than the use of
either PAC or microorganisms. The synergistic effect of adsorption and microbial
degradation (in biofilms and suspension) contributed to significantly enhanced TOC
removals.
 The mathematical model for the MBR system accurately predicted the performances
regarding TOC removal for all scenarios, namely, applications of PAC alone,
microorganisms alone, and combinations of both. The model also reflected the synergistic
effect of PAC adsorption and microbial degradation.

6.1.4 Synthesizing New Membranes
 The graphene oxide impregnated polymeric membranes were synthesized and evaluated,
and their performances were compared with polymeric membranes prepared without
graphene oxide.
214

 The introduction of graphene oxide into the polymer matrix enhanced the overall aqueous
permeability, yielding higher permeate fluxes and lower flux decline due to increased
fouling resistance when tested with secondary clarifier effluent from a wastewater
treatment plant.
 The study showed that the application of PAC adsorbent drastically increased the TOC
removal. The membrane synthesized with graphene oxide impregnation exhibited higher
permeate fluxes. With the application of PAC, the TOC removal was 96%, while in the
absence of PAC it was only 50-60%.
 The study revealed that even the best commercially available nanofiltration membranes
such as the NF90 membranes could not meet the removal requirements for most water
reclamation applications. Novel polymeric membranes can be synthesized with the
impregnation of graphene oxide nanoparticles into polymeric matrices that are distinctly
superior to existing commercial membranes regarding aqueous transport and rejection
properties. These synthesized membranes will exhibit greater aqueous transport and
permeability, superior anti-fouling resistance and lower fouling potential, better chemical
tolerance and cleanability, and greater mechanical strength and durability for sustained use
in various applications.

6.2 Recommendation
 Conduct dynamic pilot-scale tests and optimize the system with respect to influencing
parameters.
 Investigate the removal efficiency of trace organic such as endocrine disrupting chemicals
or pharmaceutical and personal care products using the combination of PAC and
215

membranes.
 Synthesize novel nanofiltration membranes impregnated with GO nanomaterials and
evaluate permeability and rejection ability compared to commercially available
membranes.

216

Bibliography
Acero, J. L., Benitez, F. J., Leal, A. I., Real, F. J., & Teva, F. (2010). Membrane filtration
technologies applied to municipal secondary effluents for potential reuse. Journal of Hazardous
Materials, 177, 390-398.

Ai, L., Zhang, C., & Chen, Z. (2011). Removal of methylene blue from aqueous solution by a
solvothermal-synthesized grapheme/magnetite composite. Journal of Hazardous Materials,
192(3), 1515-1524.

Andreozzi, R., Caprio, V., Insola, A., & Marotta, R. (1999). Advanced oxidation processes (AOP)
for water purification and recovery. Catalysis Today, 53, 51-59.

Andrews, G. F. & Tien, C. (1981). Bacterial film growth in adsorbent surfaces. American Institute
of Chemical Engineers Journal, 27(3), 396-403.

Ang, W. S., Lee, S., & Menachem Elimelech (2006). Chemical and physical aspects of cleaning
of organic-fouled reverse osmosis membranes, Journal of Membrane Science, 272, 198-210.

APHA, AWWA, & WEF (2005). Standard Methods for the Examination of Water and Wastewater,
21st ed., American Public Health Association, Washington, D.C.

Arkhangelsky, E. & Gitis, V. (2008). Effect of transmembrane pressure on rejection of viruses by
ultrafiltration membranes. Separation and Purification Technology, 62, 619-628.

Asatekin, A., Menniti, A., Kang, S., Elimelech, M., & Mayes, A. M. (2006). Antifouling
nanofiltration membranes for membrane bioreactors from self-assembling graft polymers. Journal
of Membrane Science, 285(1-2), 81-89.

Asatekin, A., Kang, S., Elimelech, M., & Mayes, A. M. (2007). Anti-fouling ultrafiltration
membranes containing polyacrylonitrile-graft-poly(ethylene oxide) comb copolymer additives.
Journal of Membrane Science, 298(1-2), 136-146.
217

Ates, N., Yilmaz, L., Kitis, M., & Yetis, U. (2009). Removal of disinfection by-product precursors
by UF and NF membranes in low-SUVA waters. Journal of Membrane Science, 328, 104-112.

Badriyha, B. N. Ravindran, V., Den, W., & Pirbazari, M. (2003). Bioadsorber efficiency, design,
and performance forecasting for alachlor removal. Water Research, 37(17), 4051-4072.

Bai, J., Zhong, X., Jiang, S., Huang, Y., & Duan, X. (2010). Graphene nanomesh. Nature
Nanotechnology, 5(3), 190-194.

Booth, T. J., Blake, P., Nair, R. R., Jiang, D., Hill, E. W., Bangert, U., Bleloch A, & Geim, A. K.
(2008). Macroscopic graphene membranes and their extraordinary stiffness. Nano letters, 8(8),
2442-2446.

Bouwer, E. J. & McCarty, P. L. (1982). Removal of trace chlorinated organic compounds by
activated carbon and fixed-film bacteria. Environmental Science and Technology, 16(12), 836-843.

Brian, P. L. T., Hales, H. B., & Sherwood, T. K. (1969). Transport of heat and mass between
liquids and spherical particles in an agitated tank. American Institute of Chemical Engineers
Journal, 15(5), 727-733.

California Department of Public Health (2014). Title 22 California Code of Regulations,
Regulations Related to Recycled Water.

Campos, C., Mariñas, B. J., Snoeyink, V. L., Baudin, I., & Lainé, J. M. (2000). PAC-membrane
filtration process. II: Model application. Journal of Environmental Engineering, 126, 104-111.

Chang, H. T. & Rittmann, B. E. (1987). Mathematical modeling of biofilm on activated carbon.
Environmental Science and Technology, 21(3), 273-280.

Chang, J., Manem, J., & Beaubien, A. (1993). Membrane bioprocesses for the denitrification of
drinking water supplies. Journal of Membrane Science, 80(1-3), 233-239.
218

Chen, H., & Belfort, G. (1999). Surface modification of [poly(ether sulfone) ultrafiltration
membranes by low-temperature plasma-induced graft polymerization. Journal of Applied Polymer
Sciences, 72(13), 1699-1711.

Chennamsetty, R. & Escobar, I. C. (2008). Evolution of a polysulfone nanofiltration membrane
following ion beam radiation. Langmuir, 24(10), 5569-5579.

Chiang, P. C., Chang, E. E., Chang, P. C., & Huang, C. P. (2009). Effects of pre-ozonation on the
removal of THM precursors by coagulation. Science of the Total Environment, 407, 5735-5742.

Choi, J. H., Jegal, J., & Kim, W. N. (2006). Fabrication and characterization of multi-walled carbon
nanotubes/polymer blend membranes. Journal of Membrane Science, 284(1-2), 406-415.

Choi, J. H., Jegal, J., & Kim, W. N. (2007). Modification of performances of various membranes
using MWNTs as modifier. Macromolecular Symposia, Advanced Polymers for Emerging
Technologies, 249-250, 610-617 (Special topic issue).

Choi, W., Choi, J., Bang, J., & Lee, J. H. (2013). Layer-by-layer assembly of graphene oxide
nanosheets on polyamide membranes for durable reverse-osmosis applications. ACS Applied
Materials and Interfaces, 5(23), 12510-12519.

Chung, H. K., Kim, W. H., Park, J., Cho, J., Jeong, T. Y., & Park, P. K. (2015). Application of
Langmuir and Freundlich isotherms to predict adsorbate removal efficiency or required amount of
adsorbent. Journal of Industrial and Engineering Chemistry, 28, 241-246.

Cicek, N., Franco, J. P., Suidan, M. T., & Urbain, V. (1998). Using a membrane bioreactor to
reclaim wastewater. Journal of American Water Works Association, 90(11), 105-113.

Convention on Biological Diversity (2013). Water and Biodiversity - Natural Solutions for Water
Security. Montreal.
219

Crittenden, J. C., Trussell, R. R., Hand, D. W., Howe, K. J., & Tchobanoglous, G. (2005). Water
Treatment Principles on Design, Second Edition (revised), John Wiley, Hoboken, New Jersey.

Crittende, J. C. & Weber, Jr., W. J. (1978a). Predictive model for design of fixed-bed adsorbers:
Parameter estimation and model development. Journal of the Environmental Engineering Division,
ASCE, 104(2), 185-197.

Crittende, J. C. & Weber, Jr., W. J. (1978b). Predictive model for design of fixed-bed adsorbers:
Single-component model verification. Journal of the Environmental Engineering Division, ASCE,
104(3), 433-443.

Cui, Z. F. & Muralidhara, H. S., Membrane technology: A practical guide to membrane technology
and applications in food and bioprocessing, Elsevier Ltd., 2010.

Daniş, Ü., & Keskinler, B. (2009). Chromate removal from wastewater using micellar enhanced
crossflow filtration: Effect of transmembrane pressure and crossflow velocity. Desalination,
249(3), 1356-1364.

Dang, H. T., Nnarbaitz, R. M., & Matsuura, T. (2010). Evaluation of apparatus for membrane
cleaning tests. Journal of Environmental Engineering, 136, 1161-1170.

Den, W. & Pirbazari, M. (2002). Modeling and design of vapor-phase biofiltration for chlorinated
volarile organic compounds. American Institute of Chemical Engineers Journal, 48(9), 2084-2103.

Dong, H. B., Xu, Y. Y., Yi, Z., & Shi, J. L. (2009). Modification of polysulfone membranes via
surface-initiated atom transfer radical polymerization. Applied Surface Science, 255(21), 8860-
8866.

Dreyer, D. R., Park, S., Bielawski, C. W., & Ruoff, R. S. (2010). The chemistry of graphene oxide.
Chemical Society Reviews, 39(1), 228-240.

220

Drioli, E. & Fontananova, E. (2004). Membrane technology and sustainable growth, Chemical
Engineering Research and Design, 82(A12): 1557-1562.

Ersever, I., Ravindran, V., & Pirbazari, M. (2007). Biological denitrification of reverse osmosis
brine concentrates: I. Batch reactor and chemostat studies. Journal of Environmental Engineering
and Science, 6(5), 503-518.

Fane, A. G., Fell, C. J. D., & Nor, M. T. (1980). Ultrafiltration/activated sludge system:
development of a predictive model. In Ultrafiltration Membranes and Applications, Cooper, A.R.
(editor), pp. 631-658, Plenum Press, New York.

Fischbein, M. D. & Drndic, M. (2008). Electron beam nanosculpting of suspended graphene sheets.
Applied Physics Letters, 93(11), 113107-113107.

Foo, K. Y. & Hameed, B. H. (2010). Insights into the modeling of adsorption isotherm systems.
Chemical Engineering Journal, 156, 2-10.

Fornasiero, F., Park, H. G., Holt, J. K., Stadermann, M., Grigoropoulos, C. P., Noy, A., & Bakajin,
O. (2008). Ion exclusion by sub-2-nm carbon nanotube pores. Proceedings of the National
Academy of Sciences, 105(45), 17250-17255.

Fritzmann, C., Löwenberg, J., Wintgens, T., & Melin, T. (2007). State-of-the-art of reverse
osmosis desalination. Desalination, 216(1), 1-76.

Gao, Y. & Deshusses, M. A. (2011). Adsorption of clofibric acid and ketoprofen onto powdered
activated carbon: Effect of natural organic matter. Environmental Technology, 32(15), 1719-1727.

Geise, G. M., Lee, H. S., Miller, D. J., Freeman, B. D., McGrath, J. E., & Paul, D. R. (2010). Water
purification by membranes: the role of polymer science, Journal of Polymer Science: Part B:
Polymer Physics, 48, 1685-1718.

221

Greenlee, L. F., Lawler, D. F., Freeman, B. D., Marrot, B., & Moulin, P. (2009). Reverse osmosis
desalination: Water sources, technology and today’s challenges. Water Research, 43, 2317-2348.

Halsey, G. (1948). Physical adsorption on non-uniform surfaces. Journal of Chemical Physics,
6(10), 931-937.

Hashimoto, A., Suenaga, K., Gloter, A., Urita, K., & Iijima, S. (2004). Direct evidence for atomic
defects in graphene layers. Nature, 430(7002), 870-873.

Hausman, R., Gullinkala, T., & Escobar, I. C. (2009). Development of low-biofouling
polypropylene feed spacers for reverse osmosis. Journal of Applied Polymer Science, 114(5),
3068-3073.

Ho, W. S. W. & Sirkar, K. K. Membrane Handbook, Van Nostrand Reinhold, 1992.

Holt, J. K., Park, H. G., Wang, Y., Stradermann, M., Artyukhin, A. B., Grigoropoulos, C. P., Noy,
A., & Bakajin, O. (2006). Fast mass transport through sub–2-nanometer carbon nanotubes.
Science, 312, 1034-1037.

Hu, M. & Mi, B. (2013). Enabling graphene oxide nanosheets as water separation membranes.
Environmental Science & Technology, 47(8), 3715-3723.

Hydronautics Press Release (2007). Integrated membrane solutions at work in Southern Spain;
(available from: http://www.membranes.com/press/Escombreras), released on January 20, 2007.

Inui, N., Mochiji, K., Moritani, K., & Nakashima, N. (2010). Molecular dynamics simulations of
nanopore processing in a graphene sheet by using gas cluster ion beam. Applied Physics A, 98(4),
787-794.

Jiang, D. E., Cooper, V. R., & Dai, S. (2009). Porous graphene as the ultimate membrane for gas
separation. Nano letters, 9(12), 4019-4024.
222

Jin, G., Englande, A. J., Bradford, H., & Jeng, H. W. (2004). Comparison of E.coli, Enterococci,
and fecal coliform as indicators for brackish water quality assessment. Water Environment
Research, 76(3), 245-255.

Joshi, R. K., Carbone, P., Wang, F. C., Kravets, V. G., Su, Y., Grigorieva, I. V., Wu, H. A., Geim,
A. K., & Nair, R. R. (2014). Precise and ultrafast molecular sieving through grapheme oxide
membranes. Science, 343(6172), 752-754.

Kamiko, N. & Ohgaki, S. (1993). Multiplication characteristics of FRNA phage and its utility as
an indicator for pathogenic viruses, Water Science and Technology, 27(3-4), 133-136.

Kenndy, M., Kim, S., Mutenyo, I., Broens, L., & Schippers, J. (1998). Intermittent crossflushing
of hollow fiber ultrafiltration systems, Desalination, 118(1-3), 175-187.

Kenny, J. F., Barber, N. L., Hutson, S. S., Linsey, k. S., Lovelace, J. K., & Maupin, M. A. (2009).
Estimated Use of Water in the United States in 2005, U.S. Geological Survey Circular: 1344.

Kilduff, J. E., & Weber, Jr., W. J. (1992). Transport and separation of organic molecules in
ultrafiltration membranes. Environmental Science and Technology, 26(3), 569-577.

Kim, J., & Van der Bruggen, B. (2010). The use of nanoparticles in polymeric and ceramic
membrane structures: Review of manufacturing procedures and performance improvement for
water treatment. Environmental Pollution, 158(7), 2335-2349.

Kim, K. S., Lee, K. H., Cho, K., & Park, C. E. (2002). Surface modification of polysulfone
ultrafiltration membranes by oxygen plasma treatment, Journal of Membrane Science, 199(1-2),
135-145.

Kim, S. H. & Pirbazari, M. (1989). Bioactive adsorber model for industrial wastewater treatment.
Journal of Environmental Engineering, 115(6), 1235-1256.

223

Kimura, S. & Sourirajan, S. (1967). Analysis of data in reverse osmosis with porous cellulose
acetate membranes used. AIChE Journal, 13(3), 497-503.

Koeing, S. P., Wang, L., Pellegrino, J., & Bunch, J. S. (2012). Selective molecular sieving through
porous grapheme. Nature Nanotechnology, 7(11), 728-732.

Kong, C., Shintani, T., & Tsuru, T. (2010). Pre-seeding-assisted synthesis of a high performance
polyamide-zeolite nano-composite membrane for water purification. New Journal of Chemistry,
34, 2101–2104.

Konieczny, K. (1998). Disinfection of surface and ground waters with polymeric ultrafiltration
membranes, Desalination, 119, 251-258.

Kwon, B., Cho, J., Park, N., & Pellegrino, J. (2006). Organic nanocolloid fouling in UF membranes.
Journal of Membrane Science, 279(1-2), 209-219.

Krasheninnikov, A. V., Nordlund, K., & Keinonen, J. (2002). Production of defects in supported
carbon nanotubes under ion irradiation. Physical Reviews B, 65(16), 165423.

Krasheninnikov, A. V., Nordlund, K., Sirviö, M., Salonen, E., & Keinonen, J. (2001). Formation
of ion-irradiation-induced atomic-scale defects on walls of carbon nanotubes. Physical Review B,
63(24), 245405.

Kull, K. R., Steen, M. L., & Fisher, E. R. (2005). Surface modification with nitrogen-containing
plasmas to produce hydrophilic, low-fouling membranes. Journal of Membrane Science, 246, 203-
215.

La, Y. H., Sooriyakumaran, R., McCloskey, B. D., Allen, R. D., Freeman, B. D., & Al-Rasheed,
R. (2012). Enhancing water permeability of fouling-resistant POSS-PEGM hydrogels using
‘addition-extraction’ of sacrificial additives. Journal of Membrane Science, 401-402, 306-312.

224

Langlais, B., Reckhow, D. A., & Brink, D. R., AWWA Research Foundation., & Compagnie
générale des eaux (Paris, France). (1991). Ozone in water treatment: Application and engineering:
cooperative research report. Chelsea, Mich: Lewis Publishers.

Langmuire, I. (1918). The adsorption of gases on plane surfaces of glass, mica, and platinum,
Journal of the American Chemical Society, 40, 1361-1402.

Lehman, S. G. & Liu, L. (2009). Application of ceramic membranes with pre-ozonation for
treatment of secondary wastewater effluent. Water Research, 43, 2020-2028.

Li, H., Cao, Y., Qin, J., Jie, X., Wang, T., Lin, J., & Yuan, Q. (2006). Development and
characterization of anti-fouling cellulose hollow fiber UF membrane for oil-water separation.
Journal of Membrane Science, 279(1-2), 328-335.

Li, H., Gao, Y., Pan, L., Zhang, Y., Chen, Y., & Sun, Z. (2008a). Electrosorptive desalination by
carbon nanotubes and nanofibres electrodes and ion-exchange membranes. Water Research, 42,
4923–4928.

Li, L., Dong, J., Nenoff, T. M., & Lee, R. (2004). Desalination by reverse osmosis using MFI
zeolite membranes. Journal of Membrane Science, 243(1), 401-404.

Li, L., Yan, G., & Wu, J. (2009). Modification of polysulfone membranes via surface-initiated
atom transfer radical polymerization and their antifouling properties. Journal of Applied Polymer
Science, 111(4), 1942-1946.

Li, L., Yan, G., Wu, Y., Yu, X., & Guo, Q. (2008b). Surface-initiated atom transfer radical
polymerization from polyimide films and their anti-fouling properties. Journal of Macromolecular
Science, Part A: Pure and Applied Chemistry, 45(10), 828-832.

225

Li, M. (2014). Development of polystyrene sulfonic acid-polyvinylidene fluoride (PSSA-PVDF)
belnds for direct methanol fuel cells (DMFCS) (Doctoral dissertation). University of Southern
California.

Li, N. N., Fane, A. G., Winston Ho, W. S., & Matsuura, T., Advanced membrane technology and
applications, Wiley, 2008.

Lind, M. L., Ghosh, A. K., Jawor, A., Huang, X., Hou, W., Yang, Y., & Hoek, E. M. V. (2009).
Influence of crystal size on zeolites-polyamide thin film nanocomposite membranes. Langmuir,
25(17), 10139-10145.

Lindau, J. & Jönsson, A. S. (1994). Cleaning of ultrafiltraion membranes after treatment of oily
wastewater, Journal of Membrane Science, 87(1-2), 71-78.

Liu, S., Zeng, T. H., Hofmann, M., Burcombe, E., Wei, J., Jiang, R., Kong, J., & Chen, Y. (2011).
Antibacterial activity of graphite, graphite oxide, grapheme oxide, and reduced grapheme oxide:
Membrane and oxidative stress. ACS Nano, 5(9), 6971-6980.

Loeb, S. & Sourirajan, S. (1963). Seawater demineralization by means of an osmotic membrane.
Journal of American Chemical Society, 117–132.

Logan, B. E. & Jiang, Q. (1990). Molecular size distributions of dissolved organic matter, Journal
of Environmental Engineering, 116, 1046-1062.

Lucchese, M. M., Stavale, F., Ferreira, E. H., Vilani, C., Moutinho, M. V. O., Capaz, R. B., Achete
C.A., & Jorio, A. (2010). Quantifying ion-induced defects and Raman relaxation length in
graphene. Carbon, 48(5), 1592-1597.

Maartens, A., Swar , P. & Jacobs, E. P. (1996). Characterisation techniques for organic foulants
adsorbed onto flat-sheet UF membranes used in abattoir effluent. Journal of Membrane Science,
119, 1-8.
226

Marshall, K. C. (1992). Biofilms: An overview of bacterial adhesion, activity, and control at
surfaces. Control of biofilm formation awaits the development of a method to prevent bacterial
adhesion. American Society Microbiology (ASM) News, 58, 202-207.

Maruf, S. H., Ahn, D. U., Pellegrino, J., Killgore, J. P., Greenberg A. R., & Ding, Y. (2012).
Correlation between barrier layer Tg (glass temperature) and a thin-film composite polyamide
membrane performance: effect of chlorine treatment. Journal of Membrane Science, 405 –406,
167–175.

Matsuura, T. (1994). Synthetic membranes and membranes separation processes. CRC press,
Boca Raton, Florida.

Monod, J. (1949). The growth of bacterial cultures, Annual Review of Microbiology, 3(1), 371-
394.

Mosqueda-Jimenez, D. B., Narbaitz, R. M. & Matsuura, T. (2004). Membrane fouling test:
Apparatus evaluation, Journal of Environmental Engineering, 130, 90-99.

Motlagh, A. M., Pant, S., & Gruden, C. (2013). The impact of cell metabolic activity on biofilm
formation and flux decline during cross-flow filtration of ultrafiltration membranes. Desalination,
316, 85-90.

MWH, Water Treatment: Principles and Design, 2nd ed., John Wiley and Sons, 2005.

Nair, R. R., Wu, H. A., Jayaram, P. N., Grigorieva, I. V., & Geim, A. K. (2012). Unimpeded
permeation of water through helium-leak–tight graphene-based membranes. Science, 335(6067),
442-444.

Ohashi, H., Sugawara, T., Kikuchi, K., & Henmi, T. (1979). Mass transfer between particles and
liquid in solid-liquid two-phase upflow in vertical tubes. Journal of Chemical Engineering of
Japan, 12(3), 190-195.
227

Peng, W. & Escobar, I. C. (2005). Evaluation of factors influencing membrane performances.
Environmental Progress, 24(4), 382-399.

Peng, W., Escobar, I. C., & White, D. B. (2004). Effects of water chemistries and properties of
membrane on the performance and fouling: A0model development. Journal of Membrane Science,
238(1-2), 33-46.

Perreault, F., Tousley, M. E., & Elimelech, M. (2013). Thin-film composite polyamide membranes
functionalized with biocidal graphene oxide nanosheets. Environmental Science and Technology
Letters, 1(1), 71-76.

Pezeshk, N., Rana, D., Narbaitz, R.M, & Matsuura, T. (2012). Novel modified PVDF ultrafiltration
flat-sheet membranes. Journal of Membrane Science, 389, 280-286.

Pieracci, J., Crivelle, J. V., & Belfort, G. (2002). Increasing membrane permeability of UV-
modified poly (ether sulfone) ultrafiltration membranes. Journal of Membrane Science, 202(1-2),
1-16.

Pirbazari, M., Badriyha, B. N., & Miltner, R. J. (1991). GAC adsorber design for removal of
chlorinated pesticides. Journal of Environmental Engineering, 117(1), 80-100.

Pirbazari, M., Badriyha, B. N., & Ravindran, V. (1992). MF-PAC for treating waters contaminated
with natural and synthetic organics. Journal American Water Works Association, 84(12), 95-103.

Pirbazari, M., Ravindran, V., Badriyha, B. N., & Kim, S. H. (1996). Hybrid membrane filtration
process for leachate treatment. Water Research, 30(11). 2691-2706.

Pirbazari, M. & Weber, Jr., W. J. (1984). Removal of dieldrin from water by activated carbon.
Journal of Environmental Engineering, 110(3), 656-669.

228

Rahimpour, A., Madaeni, S. S., Zereshki, S., & Mansourpanah, Y. (2009). Preparation and
characterization of modified nano-porous PVDF membrane with high antifouling property using
UV photo-grafting. Applied Surface Science, 255(16), 7455-7461.

Rajbartoreh, A. R., Wang, B., Shen, X., & Wang, G. (2011). Advanced mechanical properties of
graphene paper. Journal of Applied Physics, 109(1), 014306-014306.

Ravindran, V., Badriyha, B. N., Pirbazari, M., & Kim, S. H. (1996). Modeling of bioactive carbon
adsorbers: A hybrid weighted residual-finite difference numerical technique. Applied Mathematics
and Computation, 76(2-3), 99-131.

Ravindran, V., Kim, S. H., Badriyha, B. N. & Pirbazari, M. (1997). Predictive modeling for
bioactive fluidized bed and stationary bed reactors: Application to dairy wastewater.
Environmental Technology, 18(11), 861-881.

Ravindran, V., Tsai, H. H., Williams, M. D., & Pirbazari, M. (2009). Hybrid membrane bioreactor
technology for small water treatment utilities: process evaluation and primordial considerations,
Journal of Membrane Science, 344, 1-2, 39-54.

Richards, H. L., Baker, P. G. L., & Iwuoha, E. (2012). Metal nanoparticle modified polysulfone
membranes for use in wastewater treatment: A critical review. Journal of Surface Engineered
Materials and Advanced Technology, 2, 183-193.

Rittmann, B. E. (1987). Aerobic biological treatment, Environmental Science & Technology, 21(2),
128-136.

Rosenberger, S., Krüger, U., Witzig, R., Manz, W., Szewzyk, U., & Kraume, M. (2002).
Performance of a bioreactor with submerged membranes for aerobic treatment of municipal waste
water. Water Research, 36, 413-420.

229

Rudie, B. J., Ross, G. S., Harrold, S. J., & Paulson, D. J. (1993). Effects of surface force interations
on an NF/UF membrane. Desalination, 90, 107-118.

San Jose Creek Water Reclamation Plant. (n.d.). Retrieved May 1, 2015, from
http://www.lacsd.org/wastewater/wwfacilities/joint_outfall_system_wrp/san_jose_creek.asp

Shon, H. K., Kim, S. H., Erdei, L. & Vigneswaran, S. (2006). Analytical methods of size
distribution for organic matter in water and wastewater. Korean Journal of Chemical Engineering,
23(4), 581-591.

Sint, K., Wang, B., & Král, P. (2008). Selective ion passage through fuctionalized graphene
nanopores. Journal of the American Chemical Society, 130(49), 16448-16449.

Snyder, S. A., Adham, S., Redding, A. M., Cannon, F. S., DeCarolis, J., Oppenheimer, J., Wert,
E., & Yoon, Y. (2007). Role of membranes and activated carbon in the removal of endocrine
disruptors and pharmaceuticals. Desalination, 202, 156-181.

Solak, S., Vakondios, N., Tzatzimaki, I., Diamadopoulos, E., Arda, M. Kabay, N. & Yuksel, M.
(2013). A comparative study of removal of endocrine disrupting compounds (EDCs) from treated
wastewater using highly crosslinked polymeric adsorbents and activated carbon. Journal of
Chemical Technology and Biotechnology, 89(6), 819-824.

Song, W., Ravindran, V., Koel, B. E., & Pirbazari, M (2004). Nanofiltration of natural organic
matter with H2O2/UV pretreatment: fouling mitigation and membrane surface characterization,
Journal of Membrane Science, 241, 143-160.

Song, W., Ravindran, V., & Pirbazari, M. (2008). Process optimization using a kinetic model for
the ultraviolet radiation-hydrogen peroxide decomposition of natural and synthetic organic
compounds in groundwater. Chemical Engineering Science, 63, 3249-3270.

230

Speitel, G. E. J., Dovantzis, K., & DiGiano, F. A. (1987). Mathematical modeling of
bioregeneration in GAC columns. Journal of Environmental Engineering, 113(3), 32-48.

Su, Y. L., Cheng, W., Li, C., & Jiang, Z. (2009). Preparation of antifouling ultrafiltration
membranes with poly(ethylene glycol)-graft-polyacrylonitrile copolymers. Journal of Membrane
Science, 329(1-2), 246-252.

Su, Y., Li, C., Zhao, W., Shi, Q., Wang, H., Jiang, Z., & Zhu, S. (2008). Modification of
polyethersulfone ultrafiltration membranes with phosphorylcholine copolymer can remarkably
improve the antifouling and permeation properties. Journal of Membrane Science, 322(1), 171-
177.

Suk, M. E. & Aluru, N. R. (2010b). Water transport through ultrathin grapheme. The Journal of
Physical Chemistry Letters, 1, 1590-1594.

Suk, J. W., Piner, R. D., An, J., & Ruoff, R. S. (2010a). Mechanical properties of monolayer
graphene oxide. ACS Nano, 4(11), 6557-6564.

Snyder, S. A., Adham, S. Redding, A. M., Cannon, F. S., DeCarolis, J., Oppenheimer, J., Wert, E.,
& Yoon, Y. (2007). Role of membranes and activated carbon in the removal of endocrine
disruptors and pharmaceuticals. Desalination, 2010, 156-181.

Tansel, B., Sager, J., Garland, J., & Xu, S. (2009). Effect of transmembrane pressure on overall
membrane resistance during cross-flow filtration of solutions with high-ionic content. Journal of
Membrane Science, 328(1-2), 205-210

The Royal Academy of Engineering, Global Water Security – an engineering perspective, London,
United Kingdom, April 2010.

231

Tsai, H. H., Ravindran, V., Williams, M. D., & Pirbazari, M. (2004). Forecasting the performance
of membrane bioreactor process for groundwater denitrification. Journal of Environmental
Engineering and Science, 3(6), 507-521.

Tsai, H. H., Ravindran, V. & Pirbazari, M. (2005). Model for predicting the performance of
membrane bioadsorber reactor process in water treatment applications. Chemical Engineering
Science, 60(20), 5620-5636.

Tu, S. C., Ravindran, V., & Pirbazari, M. (2005). A pore diffusion model for forecasting the
performance of membrane processes. Journal of Membrane Science, 265(1-2), 29-50.

Ulbricht, M. (2006). Advanced functional polymer membranes. Polymer, 47, 2217-2262.

UN-Water (2013). Water security and the global water agenda, A UN-Water Analytical Brief.
United Nations University.

US EPA (2012). Guidelines for Water Reuse. EPA/600/R-12/618. Retrieved from
http://nepis.epa.gov/ Adobe/PDF/P100FS7K.pdf

Van der Bruggen, B. (2009). Comparison of redox initiated graft polymerization and sulfonation
for hydrophilisation of polyethersulfone nanofiltration membranes. European Polymer Journal,
45(7), 1873-1882.

Villaluenga, J. P. G., Khayet, M., Godino, P., Seoane, B. & Mengual, J. I. (2005). Analysis of the
membrane thickness effect on the pervaporation separation of methanol/methyl tertiary butyl ether
mixtures. Separation and Purification Technology, 47, 80-87.

Volk, C. J. & LeChevallier, M. W. (2000). Assessing biodegradable organic matter. Journal
(American Water Works Association), 92(5), Disinfection, 64-76.

232

Wang, S., Zhang, Y., Abidi, N., & Cabrales, L. (2009). Wettability and surface free energy of
graphene films. Langmuir, 25(18), 11078-11081.

Wang, Z., Ma, J., Tang, C. Y., Kimura, K., Wang, Q. & Han, X. (2014). Membrane cleaning in
membrane bioreactors: A review. Journal of Membrane Science, 468, 279-307.

Water treatment: Micro- and ultrafiltration. (2016, March 8), Retrieved from
https://ocw.tudelft.nl/wp-content/uploads/Micro-and-ultrafiltration-1.pdf

Wei, D., Liu, Y., Wang, Y., Zhang, H., Huang, L., & Yu, G. (2009). Synthesis of N-doped
graphene by chemical vapor deposition and its electrical properties. Nano letters, 9(5), 1752-1758.

Wei, X., Wang, R., Li, Z., & Fane, A. G. (2006). Development of a novel electrophoresis-UV
grafting technique to modify PES UF membranes used in NOM removal. Journal of Membrane
Science, 273(1-2), 47-57.

Wijmans, J. G., Nakao, S., & Smolders, C. A. (1984). Flux limitations in ultrafiltration: osmotic
pressure model and gel layer model. Journal of Membrane Science, 20, 115-124.

Wijmans, J. G., Nakao, S., Van Den Berg, J. W. A., Troelstra, F. R., & Smolders, C. A. (1985).
Hydrodynamic resistance of concentration polarization boundary layers in ultrafiltration. Journal
of Membrane Science, 22, 117-135.

Wilke, C. R. & Chang, P. (1955). Correlation of diffusion coefficients in dilute solutions. American
Institute of Chemical Engineers Journal, 1(2), 264-270.

Williams, M. D. & Pirbazari, M. (2007). Membrane bioreactor process for removing
biodegradable organic matter from water. Water Research, 41(17), 3880-3893.

233

Williams, M. D., Ravindran, V., & Pirbazari, M. (2012). Modeling and process evaluation of
membrane bioreactor for removing biodegradable organic matter from water. Chemical
Engineering Science, 84(20), 494-511.

Witzig, R., Manz, W., Rosenberger, S., Krüger, U., Kraume, M. & Szewzyk, U. (2002).
Microbiological aspects of a bioreactor with submerged membranes for aerobic treatment of
municipal wastewater. Water Research, 36, 394-402

Xie, X. L., Maia, Y. W., & Zhou, X. P. (2005). Dispersion and alignment of carbon nanotubes in
polymer matrix: a review. Materials Science and Engineering, R 49, 89–112.

Ying, W. & Weber, Jr., W. J. (1979). Bio-physicochemical adsorption model systems for
wastewater treatment. Water Pollution Control Federation Journal, 51(11), 2661-2677.

Yoon, K., Hsiao, B. S., & Chu, B. (2009). Formation of functional polyethersulfone electrospun
membrane for water purification by mixed solvent and oxidation process. Polymer, 50, 2893-2899.

Yu, D. (2006). Inorganic mesoporous membranes for water purification applications: synthesis,
testing and modeling, (Doctoral Dissertation), The Ohio State University.

Yurdacan, H. M. (2015). Investigation of mechanical, thermal and rheological properties of
flourocarbaon functionalized polystyrene; and, Development of increased flow graphene based
polymer nanocomposite membranes (Doctoral dissertation). Retrieved from University of
Southern California Digital Library.

Zanini, S., Muller, M., Riccardi, C., & Orlandi, M. (2007). Polyethylene glycol grafting on
polypropylene membranes for anti-fouling properties. Plasma Chemistry and Plasma Processes,
27(4), 446-457.

234

Zhang, J., Zou, H., Qing, Q., Yang, Y., Li, Q., Liu, Z., Guo X., & Du, Z. (2003). Effect of chemical
oxidation on the structure of single-walled carbon nanotubes. Journal of Physical Chemistry B,
107(16), 3712-3718.

Zhang, M. G., Nguyen, Q. T., & Ping, Z. (2009b). Hydrophilic modification of poly (vinylidene
fluoride) microporous membranes. Journal of Membrane Science, 327(1-2), 78-86.

Zhang, X., Chen, Y., Konsowa, A. H., Zhu, X., & Crittenden, J. C. (2009a). Evaluation of an
innovative polyvinyl chloride (PVC) ultrafiltration membrane for wastewater treatment.
Separation and Purification Technology, 70(1), 71-78.

Zhou, L. X. & Wong, J. W. C. (2000). Organic compounds in the environment. Journal of
Environmental Quality, 29(6), 1852-1856.

Zhu, L. P., Dong, H. B., Yi, Z., & Zhu, B. K. (2009). Amphiphilic PPESK-graft-P(PEGMA)
copolymer for surface modification of PPESK membranes. Materials Chemistry and Physics,
115(1), 223-228.

Abstract (if available)

Abstract Membrane separations promise to yield substantial environmental and economic benefits leading to enhanced global competitiveness by significantly reducing energy consumption, increasing industrial productivity, lowering waste generation, and addressing global water shortage problems. Membrane technologies face several scientific and technological challenges that must be overcome before witnessing widespread use in environmental, industrial and commercial applications. Environmental applications include wastewater treatment, water reclamation and reuse, water treatment, water purification, and water desalination. The specific challenges include membrane fouling and permeate flux decline, poor rejection or selectivity, and large energy footprint. The research presented here was intended to address most of these critical issues. ❧ An important aspect of this study was the examination of water reclamation processes using combination of ultrafiltration membranes and activated carbon adsorption, and to develop novel, high-performance nanofiltration membranes superior to existing commercial membranes. Firstly, flat-sheet ultrafiltration membranes were tested to evaluate the fundamental performance criteria including permeate fluxes and total organic carbon (TOC) removals in water reclamation applications using real wastewaters. Secondly, ultrafiltration membranes were employed in continuous flow hybrid membrane bioreactor (MBR) systems for treating wastewaters after secondary treatment. Thirdly, the novel membranes were synthesized using impregnation of graphene oxide (GO) nanoparticles in polymeric matrices. These membranes were intended to be superior to existing commercial membranes for water reclamation and reuse as well as other uses regarding various criteria: aqueous transport and permeability properties, anti-fouling potential and fouling resistance, rejection and separation characteristics, cleanability and flux recovery, chemical tolerance, mechanical strength, and overall durability. These membranes were tested in batch systems to evaluate their feasibility for the above applications. The finished water quality was intended to meet the necessary treatment standards for water reclamation and reuse applications regarding chemical and biological purity. Different types of cleaning agents such as caustic solution (NaOH), surfactant (Triton X-100), and biological enzyme (RID-X) were evaluated for foulants removal and permeate flux recovery. Lastly, modeling approaches were employed to predict permeate fluxes and TOC removals for MBR systems in various configurations. ❧ The study further included laboratory-scale flat-sheet plate-and-frame membrane filtration tests for investigating the permeate flux patterns and the TOC removals with secondary clarifier effluent obtained from Los Angeles County. As the TOC removals were not satisfactory with the ultrafiltration membrane itself, additional processes were required such as powder activated carbon (PAC) adsorption, microbial degradation (microorganisms including E.coli), and oxidation (ozone, and peroxone). In all these case, the permeate fluxes were also evaluated. Bench-scale studies were conducted to determine parameters used for prediction of permeate flux and TOC concentration using mathematical models. ❧ A micro-scale hollow fiber ultrafiltration membrane bench setup was designed and tested to evaluate membrane performances regarding permeate fluxes and TOC removals. The micro-scale tests were intended to provide a guideline for the design of hollow fiber membrane modules used in a mini-pilot-scale system. The mini-pilot-scale system represented a continuous flow hybrid MBR process using a hollow-fiber ultrafiltration membrane module, and it was used to assess the feasibility of water reclamation applications using permeate fluxes and TOC removals as criteria. The uniqueness of the MBR unit was the special design for controlling membrane fouling and permeate flux decline, combining powdered activated carbon (PAC) sorption and fluid management techniques. The membrane module was operated in “outside-in” dead-end fluid-dynamic regime, and equipped with structural features to promote local vortex and turbulence for fouling control. The mini pilot-scale MBR studies evaluated permeate flux decline patterns, membrane fouling, and organic rejection as TOC and UV254 (ultraviolet absorbance at 254 nm wavelength). The feed and effluent streams were also analyzed to a limited extent for biochemical oxygen demand (BOD), chemical oxygen demand (COD), biomass, and other relevant water quality parameters. ❧ A transport model using the resistances of various layers constituting concentration polarization and gel layer besides adsorbent and biofilm layer was employed for predicting the permeate fluxes and TOC removals for flat-sheet and hollow-fiber membrane configurations. The model considered membrane surface fouling, internal pore fouling, and membrane rejection in its formulation as well. The necessary model parameters were obtained from bench-scale membrane filtration tests. ❧ This investigation involved a feasibility study of the laboratory-scale MBR process for the purification of potable water sources. Reliable predictions of process performance were obtained regarding organic removal efficiency reflected by the effluent concentration profiles as functions of time. Such predictions were made on the basis of easily determined laboratory experiments, pilot-plant scale studies would be minimized and considerable savings in cost and time can be achieved. This objective was achieved by developing and employing a mathematical modeling approach. Simultaneously, a modeling protocol was observed for the process design and upscaling using dimensional analysis and similitude (although it was not the focus of this research). ❧ The model for performance prediction of organic removals (TOC removals) in the mini-pilot-scale MBR system incorporated the following phenomenological aspects including adsorption mass transfer resistance, adsorption equilibrium, biological reaction due to suspended microorganisms in bulk solution, and biological reaction within biofilms. In this regard adsorption equilibrium studies was performed to evaluate the adsorption equilibria for dissolved organic matter (DOM) using total organic carbon (TOC) as a surrogate parameter. Adsorption rate studies were conducted in batch reactors for determining the adsorption kinetics and the associated mass-transfer parameters. Batch biokinetic studies were undertaken for the estimation of biological parameters pertaining to TOC removal using an indigenous population of microorganisms. Laboratory scale MBR experiments were performed to evaluate organic removals (TOC removals) and membrane permeate fluxes as functions of operating time for a variety of process conditions. The experimental determination of TOC removal efficiencies in MBR systems provided the necessary feedback for MBR model verification and refinement. This adsorption and biodegradation model provided excellent predictions of TOC removals in MBR systems under a variety of operating scenarios including the following: (i) using PAC adsorbent alone (PAC at 40 mg L⁻¹)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Wastewater reclamation and potable reuse with novel processes: membrane performance and system integration

PDF

Fate of antibiotic resistance in anaerobic membrane bioreactors

PDF

Investigation of mechanical, thermal and rheological properties of fluorocarbon functionalized polystyrene; and, Development of increased flow graphene based polymer nanocomposite membranes

PDF

Advancing energy recovery from food waste using anaerobic biotechnologies: performance and microbial ecology

PDF

Electrokinetic transport of Cr(VI) and integration with zero-valent iron nanoparticle and microbial fuel cell technologies for aquifer remediation

PDF

Water desalination: real-time membrane characterization for performance prediction and system analysis for energetic enhancement

PDF

Catalytic oxidation and precursor identification of disinfection byproducts in recycled water

PDF

A hybrid adsorbent-membrane reactor (HAMR) system for hydrogen production

PDF

The use of carbon molecule sieve and Pd membranes for conventional and reactive applications

PDF

An integrated 'one-box' process for hydrogen production

PDF

Integrated technologies, blending schemes, and reuse practices to address contaminant and energy challenges in water reclamation

PDF

Superoxide radical and UV irradiation in ultrasound assisted oxidative desulfurization (UAOD): a potential alternative for green fuels

PDF

Thermally driven water treatment with membrane distillation: membrane performance, waste heat integration, and cooling analysis

PDF

Methanol synthesis in a membrane reactor

PDF

Stochastic and multiscale models for urban and natural ecology

PDF

Hybrid physics-based and data-driven computational approaches for multi-scale hydrologic systems under heterogeneity

PDF

Application of the adaptive river management approach to Ayamama River in Turkey

PDF

On the use of membrane reactors in biomass utilization

PDF

A high efficiency, ultra-compact process for pre-combustion CO₂ capture

PDF

Oxidative potential of urban atmospheric particles: spatiotemporal trends and associations with source-specific chemical components

Asset Metadata

Creator Kim, Woonhoe (author)

Core Title Optimizing biomembrane reactor systems for water reclamation and reuse applications

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Environmental Engineering

Publication Date 04/21/2016

Defense Date 03/22/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag graphene oxide,membrane bioreactor,membrane filtration,OAI-PMH Harvest,ultrafiltration membrane,water reclamation

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pirbazari, Massoud (committee chair), Hogen-Esch, Thieo (committee member), Lynett, Patrick (committee member)

Creator Email woonhoe.kim@gmail.com,woonhoek@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-239607

Unique identifier UC11278098

Identifier etd-KimWoonhoe-4341.pdf (filename),usctheses-c40-239607 (legacy record id)

Legacy Identifier etd-KimWoonhoe-4341.pdf

Dmrecord 239607

Document Type Dissertation

Format application/pdf (imt)

Rights Kim, Woonhoe

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA