USF Libraries
USF Digital Collections

Investigating the fouling behavior of reverse osmosis membranes under different operating conditions

MISSING IMAGE

Material Information

Title:
Investigating the fouling behavior of reverse osmosis membranes under different operating conditions
Physical Description:
Book
Language:
English
Creator:
Niriella, Dhananjaya P
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Scaling
Concentration polarization
Clay
Salt
Permeate
Dissertations, Academic -- Civil Engineering -- Doctoral -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: This dissertation describes the investigation of the fouling of a reverse osmosis membrane under different operating conditions. A mass transfer model to predict the permeate flux decline is defined. These studies used kaolin clay and bentonite clay as the fouling particles. As the membranes, thin film Low fouling Composite 1 polyamide reverse osmosis flat sheet membranes were used. Baseline experiments using only kaolin in D.I. water were conducted. At an operating pressure of approximately 1,380 kPa, no flux decline was observed. These results established the effects of a membrane-particle interaction. For the fouling experiments with kaolin clay, experiments show a linear relationship between the mass of the deposited foulant layer and total permeate flux decline. The increased concentration of scale forming salts such as calcium chloride and sodium carbonate combined with clay particles has been found to increase flux decline. It also leads to the formation of a less porous cake layer on the membrane surface, which may be due to the particle surface charge. The increase in transmembrane pressure leads to the formation of a well compacted, less porous, cake layer on the membrane surface. The reduced porosity results in the deterioration of the permeate quality, which is a direct result of reduced back diffusion of the salt solution.A fouling model that combines a resistance-in-series model and a simplified-mass-transport relationship were used to predict the transient stage permeate flux of a reverse osmosis membrane. This model contains a constant which is a function of the operating condition and the ionic species in the feed solution. It was found that the results from the model agreed with the experimental results.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Dhananjaya P. Niriella.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 152 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001914855
oclc - 179589532
usfldc doi - E14-SFE0001766
usfldc handle - e14.1766
System ID:
SFS0026084:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Investigating the Fouling Behavior of Reverse Osmos is Membranes Under Different Operating Conditions by Dhananjaya P. Niriella A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Civil and Environmental Engineering College of Engineering University of South Florida Major Professor: Robert P. Carnahan, Ph.D. Dean F. Martin, Ph.D. Stanley C. Kranc, Ph.D. Marilyn Barger, Ph.D. Michael VanAuker, Ph.D. Date of Approval: August 24, 2006 Keywords: scaling, concentration polarization, clay salt, permeate Copyright 2006, Dhananjaya P. Niriella

PAGE 2

Dedication To my beloved family

PAGE 3

Acknowledgements The author expresses his deep gratitude and endless appreciation to his major supervisor Dr. Robert P. Carnahan for his valued ad vice, guidance and encouragement throughout this study. The author wishes to express his sincere appreciation and thanks to all his committee members, Dr. Dean F. Martin, Dr. Stanley C. Kranc, Dr. M. Barger and Dr. Michael VanAuker for agreeing to serve in the d issertation committee and for their interest, useful suggestions and constant support i n this study. Special thanks go out to Ms. Elizabeth Hood and Mr. Brian Martin in carrying out the particle size analysis, Mr. Haito Li for assisting in carrying out BET analysis Mr. Jay Bieber for the EPS analysis. The author is also grateful to Mr. Rafael Urea for assisting him with instrumentation and troubleshooting on countless occasions, Mr. Robert R. Smith and Mr. Tom Gage of the engineering machine shop for their technical suppor t. The valuable discussions and useful suggestions fro m Mr. Jorge Agunaldo, Dr. Silvana Ghiu and Mr. Miles Beamguard are gratefully acknowledged. The author is also grateful to his wife for her constant help and supp ort in terms of mobilizing and running the experiments, and preparation of this report. Th e author deeply appreciates the financial assistance in the form of graduate resear ch and teaching assistantship given by office of the engineering research and department o f Civil and Environmental Engineering of the University of South Florida. Author also does not forget the opportunity extende d to me by Dr. Manjriker Gunaratne and Dr. Ram Pendyala to pursue a PhD at U SF and Dr. Sunil Saigal, Mr. Sean Gilmore, Ms. Catherine High and Mr. Paul Mulrenin f or assisting him administratively and financially at various times.

PAGE 4

i Table of Contents List of Tables List of Figures List of Symbols Abstract Chapter 1 Introduction 1.1 Scope and Significance 1.2 Research Objectives 1.3 Arrangement of the Dissertation Chapter 2 Literature Review 2.1 Introduction 2.2 Definition of a Membrane 2.3 Reverse Osmosis Membranes 2.3.1 Types of Reverse Osmosis Membranes 2.4 Clays 2.4.1 Mineralogy 2.4.2 Kaolinite 2.4.3 Bentonite 2.4.4 Clay Particle Surface Charge 2.5 Determining Surface Area of Particles 2.6 Electrokinetic Measurements and Zeta Potentia l Determination 2.7 Membrane Fouling 2.7.1 Introduction 2.7.2 Effects of Fouling 2.7.3 Previous Work on RO Membrane Fouling 2.7.4 Fouling Material 2.7.5 Clay Content in Water Sources 2.7.6 Fouling Mechanism in Microfiltration (MF) a nd Ultrafiltration (UF) Compared with Reverse Osmosi s Membranes 2.7.7 Membrane Surface Charge and Measurement Techniques 2.7.8 Fouling Tests v x xv xix 1 1 3 3 5 5 5 6 9 10 10 11 12 13 14 14 16 16 16 17 18 18 19 19 20

PAGE 5

ii 2.7.9 Effects of Fouling on Product Quality 2.7.10 Particle Deposition on Membrane Surface 2.8 Scaling 2.8.1 Introduction 2.8.2 Scale Formation in Membrane Systems and Its Effects 2.8.3 Concentration Polarization 2.8.4 Surface and Bulk Scaling 2.8.5 Factors Affecting Scaling 2.8.5.1 Supersaturation 2.8.5.2 Velocity and Shear Rate 2.8.5.3 pH and Ionic Strength 2.8.5.4 Nucleation 2.8.6 Cost of Scaling 2.8.7 CaCO3 Scale Structure 2.8.8 CaCO3 Scaling and Potential Determination 2.9 Fouling Models 2.9.1 Resistance in Series Model 2.9.2 Concentration Polarization Model 2.9.3 Gel Polarization Model 2.9.4 Inertial Migration Model 2.9.5 Shear Induced Hydrodynamic Convection Model 2.9.6 Shear Induced Hydrodynamic Diffusion Model 2.9.7 Scour Model 2.9.8 Turbulent Burst Model Chapter 3 Experimental Methodology 3.1 Electrokinetic Mobility and Zeta Potential Meas urement 3.1.1 Materials and Chemicals 3.1.2 Measuring Instrument and Technique 3.2 Membrane Characterizing, Fouling, Scaling and M odeling Experiments 3.2.1 Design Philosophy 3.2.2 Experimental Unit 3.2.3 Membrane Type and Specifications 3.2.4 Membrane Cell and Holder 3.2.5 Experimental Procedure for Membrane Characte rizing, Fouling, Scaling and Combined Fouling and Scaling Runs 3.2.5.1 Membrane Characterizing 3.2.5.2 Compaction of the Membrane and Pure Water Permeability 3.2.5.3 Kaolin Fouling Experiments 3.2.5.4 CaCO3 Scaling Experiments 21 21 23 23 23 23 24 25 25 25 26 26 27 27 28 29 29 31 33 34 34 35 36 37 39 39 39 40 43 43 44 45 46 50 50 51 52 52

PAGE 6

iii 3.2.5.5 CaCO3 Scaling and Kaolin Fouling Experiments 3.2.5.6 Membrane and System Cleaning after Operation Chapter 4 Results and Discussion 4.1 Characterization of Clay Particles 4.1.2 Surface Area 4.1.3 Particle Size Analysis 4.1.4 Zeta Potential Measurement 4.1.4.1 Kaolin in Distilled Water and Salt Solutions 4.1.4.2 Kaolin in Combined Salt Solution 4.2 Membrane Characterization 4.2.1 Purewater Permeation Tests for the LFC 1 Membrane 4.3 Membrane Performance 4.4 Membrane Fouling Runs with Kaolin 4.4.1 Flux-Time Relationship 4.4.2 Linear Flux vs Time Relationship 4.4.3 Comparison of Kaolin with Bentonite Clay Fou ling 4.4.4 Effects of Operating Variables on Flux 4.4.4.1 Applied Pressure 4.4.4.2 Particle Concentration 4.4.4.3 Crossflow Velocity 4.4.4.4 Occurrence of Critical Flux 4.4.4.5 Mass Deposited vs Flux Decline Relationship 4.4.5 Statistical Model 4.5 Membrane Scaling Runs 4.5.1 Preparation of Scaling Solution 4.5.2 Studies with 0.0005 M CaCl2 and Na2CO3 4.5.3 Studies with MgCl2 and Na2CO3 4.5.4 CaCO3 Scaled Membranes with Acetic Acid 4.5.5 Effect of Salt Concentration 4.5.6 Effects of Crossflow Velocity 4.5.7 Effect of Transmembrane Pressure 4.5.8 CaCO3 Scaling Runs at Different pH Values 4.5.9 Effect on Permeate Quality 4.6 Kaolin and CaCO3 Experiments 4.6.1 Permeate Flux vs Time 4.6.2 Permeate Quality with Time 4.6.3 Reversibility of the Fouling Layer Chapter 5 Modeling 53 54 55 56 56 56 57 57 57 58 58 60 62 62 63 64 65 65 67 68 70 70 71 72 73 74 75 76 77 79 80 81 83 84 85 87 88 90

PAGE 7

iv Chapter 6 Conclusion References Appendices Appendix A: Kaolin and Membrane Characterizing Data Appendix B: Kaolin Particles Size Distribution Appendix C: Permeation Data for Kaolin Runs Appendix D: SPSS Statistical Analysis Results Appendix E: Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs Appendix F: Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs Appendix G: Fouling Model Calibration Data and Resu lts About the Author 97 99 105 106 108 110 119 121 126 144 End Page

PAGE 8

v List of Tables Table 2.1 Types of Membrane Based on the Size of th e Material they Retained and their Driving Forces Table 3.1 Specific Conductance and the Recommended Maximum Applied Voltage Relationship for ZM -80 Table 3.2 Instruments and their Specifications Table 5.1 Summary Data from Model Analysis Table A.1 Zeta Potential Values of Kaolin Table A.2 Pure Water Permeability Data for LFC 1 Membrane Table A.3 Calculation of Supersaturation Factor for CaCO3 Table A.4 Pure Water Volumetric Flux (m3/m2/s) at Different CaCl2 Concentration and Crossflow Velocities for LFC 1 Me mbrane Table A.5 t-Test Results for Kaolin (Transmembrane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 1.62 and 4.04 cm/s) Table C.1 Permeation Data for Kaolin Runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 50 mg/l, Crossf low Velocity = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Table C.2 Permeation Data for Kaolin Runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Cross flow Velocity = 1.62 cm/s, pH = 6.7, Temperature = 24 oC Table C.3 Permeation Data for Kaolin runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Crossflow V elocity = 4.04 cm/s, pH = 6.7, Temperature = 24 oC 7 43 49 94 106 106 106 107 107 110 110 111

PAGE 9

vi Table C.4 Permeation Data for Kaolin Runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Cross flow velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table C.5 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 2,070 kPa, Kaolin Concentration = 150 mg/l, Cross flow Velocity = 4.04 cm/s, pH = 6.8, Temperature = 24 oC Table C.6 Permeation Data for Kaolin Runs. Transmem brane Pressure = 2,070 kPa, Kaolin Concentration = 150 mg/l, Crossf low Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table C.7 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 2,760 kPa, Kaolin Concentration = 150 mg/l, Cross flow Velocity = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Table C.8 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 2,760 kPa, Kaolin Concentration = 150 mg/l, Cros sflow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table C.9 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 3,450 kPa, Kaolin Concentration = 50 mg/l, Crossf low Velocity = 4.04 cm/s, pH = 6.8, Temperature = 24 oC Table C.10 Permeation Data for Kaolin Runs. Transm embrane Pressure = 3,450 kPa, Kaolin Concentration = 50 mg/l, Crossfl ow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table C.11 Permeation Data for Kaolin Runs. Transm embrane Pressure = 3,450 kPa, Kaolin Concentration = 150 mg/l, Cross flow Velocity = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Table C.12 Permeation Data for Kaolin Runs. Transm embrane Pressure = 3,450 kPa, Kaolin Concentration = 150 mg/l, Cross flow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table C.13 Table C.13 Permeation Data for Kaolin R uns. Transmembrane Pressure = 3,450 kPa, Kaolin Concentration = 250 m g/l, Crossflow Velocity = 4.04 cm/s, pH = 6.7, Temperat ure = 24 oC Table C.14 Permeation Data for Kaolin Runs. Transm embrane Pressure = 3,450 kPa, Kaolin Concentration = 250 mg/l, Crossf low Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC 112 112 113 113 114 114 115 116 116 117 118

PAGE 10

vii Table D.1 Univariate Analysis of Variance – Tests B etween – Subjects Effects Table D.2 Univariate Analysis of Variance – Paramet er Estimation Table E.1 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 1,380 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 8.9, Temperature = 24 oC Table E.2 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,070 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 9.1, Temperature = 24 oC Table E.3 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,760 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table E.4 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table E.5 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Crossflow velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Table E.6 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 5.5, Temperature = 24 oC Table E.7 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 4.0, Temperature = 24 oC 119 119 121 121 122 122 123 124 125

PAGE 11

viii Table F.1 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 1,380 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.2, Temperature = 24 oC Table F.2 Purewater Permeability Data at the Start Transmembrane Pressure = 1,380 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.3 Purewater Permeability Data at the End. Transmembrane Pressure = 1,380 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature= 24 oC Table F.4 Permeation Data for Kaolin and Ca Cl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,070 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.1, Temperature = 24 oC Table F.5 Purewater Permeability Data at the Start. Transmembrane Pressure = 2,070 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.6 Purewater Permeability Data at the End. T ransmembrane Pressure = 2,070 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.7 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,760 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.1, Temperature = 24 oC Table F.8 Purewater Permeability Data at the Start. Transmembrane Pressure = 2,760 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.9 Purewater Permeability Data at the End. T ransmembrane Pressure = 2,760 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.10 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.3, Temperature = 24 oC 126 126 127 128 128 129 130 130 131 132

PAGE 12

ix Table F.11 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Tempe rature = 24 oC Table F.12 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Tempe rature = 24 oC Table F.13 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4 .04 cm/s, pH = 9.2, Temperature = 24 oC Table F.14 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Tempe rature = 24 oC Table F.15 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.16 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 250 mg/l, Crossflow Velocity = 4 .04 cm/s, pH = 8.9, Temperature = 24 oC Table F.17 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.18 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.19 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4.0 4 cm/s, pH = 9.0, Temperature = 24 oC Table F.20 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.21 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC 132 133 134 134 135 136 136 137 138 138 139

PAGE 13

x Table F.22 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.3, Temperature = 24 oC Table F.23 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.24 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.25 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 250 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.2, Temperature = 24 oC Table F.26 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table F.27 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temper ature = 24 oC Table G.1 Calculated 1/V(t) and V(t)*t Values for Kaolin and CaCO3 at 1,380 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Table G.2 Calculated 1/V(t) and V(t)*t Values for Kaolin and CaCO3 at 2,070 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Table G.3 Calculated 1/V(t) and V(t)*t Values for Kaolin at 2,760 kPa, 4.04 cm/s, 0.0005 M CaCO3 and150 mg/l of Kaolin Table G.4 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Table G.5 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005M CaCO3 and 250 mg/l of Kaolin Table G.6 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005M CaCO3 and 50 mg/l of Kaolin Table G.7 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 50 mg/l of Kaolin 140 140 141 142 142 143 144 145 146 147 148 149 150

PAGE 14

xi Table G.8 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 150 mg/l of Kaolin Table G.9 Calculated1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 250 mg/l of Kaolin 151 152

PAGE 15

xii List of Figures Figure 2.1 Reverse Osmosis – Pressure Applied to Re verse the Normal Osmotic Flow of Water (After, Williams, 2003) Figure 2.2 Osmosis Solvent (Normal Water) Passes Through a SemiPermeable Barrier from Side of Low to High Solute Concentration (After, Williams, 2003) Figure 2.3 Asymmetric Membrane (After, Williams, 2 003) Figure 2.4 Thin – Film Composite Membrane (After, W illiams, 2003) Figure 2.5 Tetrahedral Structure (After, Stumm and Morgan, 1981) Figure 2.6 Octahedral Structure (After, Stumm and Morgan, 1981) Figure 2.7 Ideal Kaolin Structure (After, Rouquerol et al., 1999) Figure 2.8 Ideal Bentonite Clay Structure (After, R ouquerol et al., 1999) Figure 2.9 Electrical Double Layer (After, Sawyer e t al., 1994) Figure 2.10 Forces Acting on a Particle Near a Memb rane (After, Wiesner and Chellam, 1992) Figure 2.11 Concentration Polarization (CP) within a Membrane Module Figure 2.12 Concentration Polarization Model Figure 3.1 ZM-80 Apparatus Figure 3.2 GT-20 Electrophoresis Cell Figure 3.3 Zeiss DR Microscope Figure 3.4 Schematic Diagram of the Experimental Se t-up Figure 3.5 Membrane Cell 8 8 9 9 10 10 11 13 15 22 24 32 41 41 40 45 47

PAGE 16

xiii Figure 3.6 Reverse Osmosis Flat Sheet Membrane Cell – Bottom Side Figure 3.7 Reverse Osmosis Flat Sheet Membrane Cell – Top Side Figure 3.8 Reverse Osmosis Flat Sheet Membrane Cell – Assembled View Figure 4.1 Zeta Potential (mv) vs pH for Kaolin in Distilled Water and Scaling Solutions. Kaolin Concentration = 150 mg/l Figure 4.2 Pure Water Flux vs Transmembrane Pressur e for LFC 1 Membrane Figure 4.3 Purewater Volumetric Flux vs Transmembrane Pressure for 0.0005 M CaCl2 for LFC 1 Membrane Figure 4.4 Purewater Volumetric Flux vs Transmembrane Pressure for 0.001 M CaCl2 for LFC 1 Membrane Figure 4.5 Rel. Permeate Flux vs Time for Applied Pressure = 1,380 kPa and 3,450 kPa, and Kaolin Concentrations of 50 and 150 mg/l and Crossflow Velocity = 4.04 cm/s, pH = 6.7 Figure 4.6 Rel. Per. Vol. Flux vs Time for Kaolin and Bentonite. Applied Pressure = 1,380 – 3,850 kPa, Kaolin Concentrati ons of 150 mg/l, Bentonite Concentration 100 mg/l and Crossfl ow Velocity = 4.04 cm/s Figure 4.7 Permeate Flux vs Time for Applied Press ure = 1,380 3,450 kPa, and Kaolin Concentrations = 150 mg/l and Cross flow Velocity = 4.04 cm/s, pH = 9.0 Figure 4.8 Permeate Flux vs Time for Kaolin Concen tration. Transmembrane Pressure = 3,450 kPa and Crossflow Ve locity = 4.04 cm/s, pH = 6.4 6.8 Figure 4.9 Permeate Flux vs Time, Applied Pressur e = 1,380 kPa, Kaolin Concentration = 150 mg/l Figure 4.10 Flux Decline vs Mass of Cake Deposited for Bentonit e and for a LFC 1 Membrane Figure 4.11 Model Predicted Flux Decline vs Experimental Flux D ecline Values for Bentonite and for a LFC 1 Membranes 47 48 48 58 59 60 61 63 64 66 68 69 71 72

PAGE 17

xiv Figure 4.12 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 m/s, pH = 9.0 Figure 4.13 Permeate Flux vs Time for CaCl2 Plus Na2CO3 and MgCl2 Plus Na2CO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s and pH= 9.0 Figure 4.14 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 m/s, CaCO3 Concentration = 0.0005 M Figure 4.15 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 m/s, pH = 9.0. CaCO3 Concentration 0.0005 M and 0.0015 M Figure 4.16 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s and 25.8 cm/s, CaCO3 Concentration = 0.0005 M, pH = 9.0 Figure 4.17 Flux vs Time for CaCO3. Pressure = 1,380 – 3,450 kPa, Velocity = 4.04 cm/s, CaCO3 = 0.0005 M Figure 4.18 pH vs pC for a Carbonate System (After, Sawyer et a l., 1994) Figure 4.19 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Nominal CaCO3 Concentration = 0.0005 M, pH = 4.0, 5.5 and 9.0 Figure 4.20 Rejection vs Time for CaCO3. Transmembrane Pressure = 1,380 3,450 kPa, Crossflow Velocity = 4.04 cm/s, CaCO3 Concentration = 0.0005 M and 0.0015 M, pH = 9.0 Figure 4.21 Permeate Flux vs Time for Combined CaCO3 and Kaolin. Transmembrane Pressure = 3,450 kPa, Crossflow Veloc ity = 4.04 cm/s, Kaolin Concentration = 50 mg/l, CaCO3 Concentration = 0.0005 M and 0.0015 M, pH = 9.0 Figure 4.22 Permeate Flux vs Time for Combined CaCO3 and Kaolin. Transmembrane Pressure = 3,450 kPa, Crossflow Velo city = 4.04 cm/s, Kaolin Concentration = 50 – 250 mg/l, C aCO3 Concentration = 0.0005 M, pH = 9.0 74 76 77 79 80 81 82 83 84 85 86

PAGE 18

xv Figure 4.23 Observed rejection vs Time for CaCO3 and Combined CaCO3 and Kaolin. Transmembrane Pressure = 3,450 kPa, Cro ssflow Velocity = 4.04 cm/s, Kaolin Concentration = 150 m g/l, CaCO3 Concentration = 0.0005 M, pH = 9.0 Figure 4.24 Final Kw (as % of initial Kw) For Kaolin = 150 mg/l, CaCO3 = 0.0005 M and Crossflow Velocity = 4.04 cm/s Figure 5.1 Particle Deposition Mechanism on a Memb rane Figure 5.2 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015M CaCO3 and 50 mg/l of Kaolin Figure 5.3 Specific Resistance of the Cake Layer () vs Transmembrane Pressure (P) for 0.0005 M CaCO3 and 150 mg/l of Kaolin Figure 5.4 Experimental and Model Results, Transmem brane Pressure = 3,450 kPa, CaCO3 Concentration = 0.0015 M, Kaolin Concentration = 250 mg/l Figure B.1 Particle Size Distribution. Kaolin in D istilled Water, Kaolin Concentration = 150 mg/l, pH = 6.7 Figure B.2 Particle Size Distribution. Kaolin in D istilled Water, Kaolin Concentration = 150 mg/l, pH = 9.0 Figure B.3 Particle Size Distribution. Kaolin in 0 .0005 M CaCO3, Kaolin Concentration = 150 mg/l, pH = 9.0 Figure B.4 Particle Size Distribution. Kaolin in 0 .0015 M CaCO3, Kaolin Concentration = 150 mg/l, pH = 9.0 Figure D.1 Relationship Between Experimental Data a nd Model Results Figure F.1 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 1,380 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.2, Temperature = 24 oC 87 89 90 93 95 96 108 108 109 109 120 127

PAGE 19

xvi Figure F.2 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,070 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4 .04 cm/s, pH = 9.1, Temperature = 24 oC Figure F.3 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,760 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4 .04 cm/s, pH = 9.1, Temperature = 24 oC Figure F.4 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.3, Temperature = 24 oC Figure F.5 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.2, Temperature = 24 oC Figure F.6 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 250 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 8.9, Temperature = 24 oC Figure F.7 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4.0 4 cm/s, pH = 9.0, Temperature = 24 oC 129 131 133 135 137 139

PAGE 20

xvii Figure F.8 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4 .04 cm/s, pH = 9.3, Temperature = 24 oC Figure F.9 Permeation and Purewater Permeability C oefficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 250 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.2, Temperature = 24 oC Figure G.1 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 1,380 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Figure G.2 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 2,070 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Figure G.3 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 2,760 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Figure G.4 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Figure G.5 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 250 mg/l of Kaolin Figure G.6 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 50 mg/l of Kaolin Figure G. 7 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 50 mg/l of Kaolin Figure G.8 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 150 mg/l of Kaolin Figure G.9 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 250 mg/l of Kaolin 141 143 144 145 146 147 148 149 150 151 152

PAGE 21

xv List of Symbols =2 ,1 a aK Carbonic dissociation constants mc = deposited mass of particles (kg) P = Transmembrane pressure (Pa) Am = Membrane area (m2) B = Constant (sm-2) C = Foulant concentration in the feed (mg/l) Cb = Bulk concentration (kg/m3) Cc = Cake layer concentration (kg/m3) Cf = Salt concentration in the feed (mol/m3) Cf = Foulant concentration in cake layer (mg/l) Cg = Gel layer concentration (mol/m3) Cm = Salt concentration on the membrane surface (mol/m3) Cpart = Particle concentration in the solution (mg/l) Cp = Salt concentration in the permeate (mol/m3) D = Diffusivity, Foulant diffusion coefficient (m2/s) dp = Diameter of the particle (m) D () = Di-electric constant dv = Volume equivalent diameter (m) dvg = Geometric mean diameter (m)

PAGE 22

xvi dv = Volume equivalent diameter (m) dvg = Geometric mean diameter (m) EM = Electrophoretic mobility (m.cm/s.v) Jcrit = Critical flux (m3/m2/s) Jf = Permeate flux of the fouled membrane (m3/m2/s) JSolvent= Solvent volumetric flux (m3/m2/s) J, V(t) = Permeate volumetric flux (m3/m2/s) k = Mass transfer coefficient (m/s) k2 = Mass transfer coefficient (m/d) K = Hydraulic resistance of the fouling layer per u nit mass (Nsm-1kg-1) K* = Solubility product (M2) KB = Boltzmann constant (JK-1) Ke (m) = the erosion coefficient Ksp = Thermodynamic solubility product at equilibrium (M2) Kw = Purewater Permeability Coefficient (gm/cm2/s/atm) LSI = Langelier Saturation Index m = Mass of foulant deposited per unit area (kg/m2) mp = Mass of the particles deposited (kg) m(t) = Foulant mass flux (mg/m3) pHR = pH of concentrate pHs = pH at CaCO3 saturation q = Permeate volumetric flux (m3/m2/s) rd = rate of deposition (kgm-2s-1)

PAGE 23

xvii re = rate of re-entrainment (kgm-2s-1) Re = Reynolds Number Rm = Membrane resistance (Pa.S/m) Rs, Rf = Fouling layer resistance (Pa.S/m) T = Temperature (K) TDS = Total Dissolved Solids (mg/l) t = time (s) t = Time interval (s) V = Applied Voltage (Volts) Ub = Bulk velocity (ms-1) V() = Viscosity of suspended liquid (poise) ZP = Zeta potential (mv) = Osmotic pressure (Pa) = Concentration boundary layer thickness (m) = Cake layer specific resistance (s-1) = Angstrom = Permeate viscosity (Pa.s) s = Supersaturation level c = Cake layer thickness (m) np = The density of the particles (kg/m3) = Fractional voidage of the cake layer g = Geometric standard deviation (m) = Dynamic shape factor

PAGE 24

xviii .g = Shear rate (s-1) = Activity coefficient I = Ionic strength (M) = constant (m2/kg-1) r = Specific resistance (mkg-1)

PAGE 25

xix Investigating the Fouling Behavior of Reverse Osmos is Membranes Under Different Operating Conditions Dhananjaya P. Niriella ABSTRACT This dissertation describes the investigation of th e fouling of a reverse osmosis membrane under different operating conditions. A ma ss transfer model to predict the permeate flux decline is defined. These studies use d kaolin clay and bentonite clay as the fouling particles. As the membranes, thin film Low fouling Composite 1 polyamide reverse osmosis flat sheet membranes were used. Baseline experiments using only kaolin in D.I. wate r were conducted. At an operating pressure of approximately 1,380 kPa, no f lux decline was observed. These results established the effects of a membrane-parti cle interaction. For the fouling experiments with kaolin clay, experiments show a li near relationship between the mass of the deposited foulant layer and total permeate flux decline. The increased concentration of scale forming salts such as calcium chloride and sodium carbonate combined with clay particles has been found to increase flux decline. It also leads to the formation of a less porous cake layer on the membrane surface, which ma y be due to the particle surface charge. The increase in transmembrane pressure lead s to the formation of a well compacted, less porous, cake layer on the membrane surface. The reduced porosity

PAGE 26

xx results in the deterioration of the permeate qualit y, which is a direct result of reduced back diffusion of the salt solution. A fouling model that combines a resistance-in-serie s model and a simplifiedmass-transport relationship were used to predict th e transient stage permeate flux of a reverse osmosis membrane. This model contains a con stant which is a function of the operating condition and the ionic species in the fe ed solution. It was found that the results from the model agreed with the experimental results

PAGE 27

1 Chapter 1 Introduction 1.1 Scope and Significance Water in the earth’s hydrosphere (water in the atmo sphere, earths’s surface and crust up to a depth of 2000 m) is usually considere d when calculating the earth’s water storage. This storage is approximately equal to abo ut 1,386 MCM. Of this volume, 97.5 % is saline and 2.5 % is freshwater. Out of the fre shwater volume, only 1 % is readily available for human consumption (Shiklomanov, 1999) Over the next 20 years, the average supply of water worldwide per person is exp ected to drop by one-third. Today, between 2 to 4 billion people suffer, annually, fro m diseases linked to contaminated water (http://www.lifetoday.org/partner ). Desalination is the most expedient means of increas ing the supply of freshwater in regions where water is scare. It is estimated that more than 50 % of the world’s population live within 50 miles of the sea (http:// www.solarsystem.nasa.gov ) Given the almost unlimited availability of seawater, desalina tion could provide a sustainable water supply to many municipalities and industries. Exper ts predict that the 21 st century belongs to seawater desalination.

PAGE 28

2 Reverse osmosis membranes used in water desalinatio n are capable of producing highly purified water by removing all the salts and some other contaminants from different water sources. During the past several de cades, tremendous strides were made in the research related to development of Reverse Osmo sis (RO) membranes, which has resulted in the production of new membranes capabl e of withstanding wide pH ranges, higher temperatures and pressures, increased flux a nd reduced solute concentration in the permeate. But unfortunately, with all these new fin dings, membrane fouling and scaling remain the two major operational and maintenance is sues faced by membrane water treatment plant operators. The short-term effects o f fouling and scaling are; reduction of treated water productivity, deterioration water qua lity combined with increase in energy consumption. The long term effect being membrane re placement. Clay is a major foulant (Ng and Elimelech, 2004), p resent in natural water and CaCO3 is the most common scale compound, which forms ten acious layers in domestic and industrial water systems. In membrane water tre atment units, feed water pretreatment is carried out in order to remove fouling and scale forming substances. Although 100 % removal of these substances is possi ble, it is not economically feasible. Also, in the past there had been instances (e.g., w atershed erosion taking place after a storm) where there had been heavy sediment flow to water bodies that act as water sources to membrane treatment units. These heavy se diment load in the feed water in turn has made the pre-treatment units ineffective (Rookl idge et al., 2002). However, understanding of the effects of physical o perating parameters (e.g. transmembrane pressure, crossflow velocity) on clay fouling of a R.O. membrane and the interaction between clay and scale forming salt wou ld assist in managing the fouling and

PAGE 29

3 scaling problem. Further, from a plant operational standpoint, a simple model to predict the transient stage permeate flux of a membrane for a given feed would be very useful. 1.2 Research Objectives The specific research objectives are: 1. To investigate the effects of operating paramet ers (transmembrane pressure, crossflow velocity) and solute concentrations (clay and CaCO3) on scaling of a Reverse Osmosis membrane, and clay-CaCO3 interaction on membrane performance 2. To describe a simple transport model based on t he data obtained to predict the transient permeate flux for a given feed solution. 1.3 Arrangement of the Dissertation Chapter 2 reviews the literature related to reverse osmosis membranes, current membrane fouling technology, clay structures, scali ng by inorganic salts in the feed water and fouling models. Chapter 3 outlines the experimental arrangement, me thodology and instrumentation used. The methodologies adopted by previous researchers have been considered as guidelines in this research.

PAGE 30

4 Chapter 4 presents experimental results including d ata analysis on characterizing of kaolin clay particles and Low Fouling Composite 1 (LFC1) reverse osmosis membrane manufactured by Hydranutics, Oceanside, CA effects of physical parameters (transmembrane pressure, crossflow velocity) and fo ulant concentration on both the clay fouling and CaCO3 scale formation and their interactions on R.O. mem brane. Chapter 5 describes a transport model for predictio n of transient stage flux in the presence of fouling feedwater and Chapter 6 summari zes experimental findings and concludes with suggestions for future work.

PAGE 31

5 Chapter 2 Literature Review 2.1 Introduction As world population is estimated to climb above si x billion by the end of 2005 (http://www.cia.gov), increasing demands for water from municipal, industrial, commercial, irrigation and environmental sectors wo uld impose additional stresses on the world’s limited fresh water resources. Further, unc ontrolled discharge of wastewater and effluent from various sources has led to the pollut ion of these limited water resources. It is estimated that over 50 % of the U.S population c urrently lives within 50 miles of the ocean (http://www.solarsystem.nasa.gov ) or other unusable water source. Finding solutions to the above issues are a major challenge facing the scientists and engineers. 2.2 Definition of a Membrane Membrane industry has made great strides during the past 50 years. Membranes have paved the way for new industries to emerge, co vering such wide-ranging applications as reverse osmosis, gas separation, co ntrolled-release pharmaceutical formulations and the artificial kidney. A combined knowledge of physical and polymer

PAGE 32

6 chemistry, electro-chemistry, process and mechanica l engineering is needed to produce membranes. A semi-permeable membrane is a very thin film that allows some types of matter to pass, while retaining others. Some membra nes such as micro-filtration and ultra-filtration membranes are porous. Others membr anes are dense and separate material based on differences in diffusion rates through the membrane (USBR, 1998). Membranes are divided into different categories bas ed on the size of the material they retain and their driving forces (USBR, 1998) a s shown in Table 2.1. There are four industrial membrane separation processes, which are micro-filtration (MF), ultrafiltration (UF), reverse osmosis (RO) and electro-d ialysis (Baker, 2000). 2.3 Reverse Osmosis Membranes Use of reverse osmosis membranes to treat high sali nity water is an ideal solution to solving water shortages. Reverse osmosis membran es can easily produce potable water from sea and brackish water (USBR, 1998).

PAGE 33

7 Reverse osmosis (RO) is a solution separation proc ess, in which, a solvent is passed through a semi permeable membrane while reta ining solutes (Williams, 2003). In reverse osmosis, pressure is applied to reverse the normal osmotic flow of water across a semi-permeable membrane (Figure 2.1). In the absenc e of applied pressure, until, osmotic Table 2.1 Types of Membranes Based on the Size of t he Material they Retain and their Driving Forces ED Dialysis Gas separation MF UF NF RO Driving force Voltage, typically 12 V/Cell Pair Concentration difference Pressure difference 1100 atm Pressure difference typically 10 psi Pressure difference typically 10 -100 psi Pressure difference 75 150 psi Pressure difference 100 -800 psi Materials retained Water, micro-organisms, uncharged molecules, suspended solids Dissolved and suspended material with molecular wt >1,000 Membrane-impermeable gases and vapors Particles larger than pore size, Suspended material (silica, Bacteria, etc) Molecules larger than the molecular mass cut off, mainly biologicals, colloids and macromolecules > 95% of multi-valent ions, 25-90% of monovalent ions, molecules and particles over 300 Daltons > 95% of all ions, most molecules and particles over 200, virtually all suspended and dissolved material Materials transported Dissolved salts Ions, and lowmolecular weight organics (Urea, etc.) Gases and vapors Dissolved salts, small particles, water Small molecules and ions, water Monovalent ions and very small molecules, water Very small uncharged molecules, water Water permeation (m3/m2/d) Practically None 800 4,000 0.4 2.5 1.0 0.8 (After, Baker, 2000; United States Department o f Interior Bureau of Reclamation, 1998) ED = Electro-dialysis, MF = Micro–filtration, UF = Ultra-filtration, NF = Nano-filtration, RO = Rev erse Osmosis

PAGE 34

8 equilibrium is achieved, the solvent flow through a semi permeable membrane is from the side with smaller concentration to that of greater concentration (Figure 2.2). When this occurs, pressure difference between the two sides o f the membrane is equal to the osmotic pressure (Byrne, 1995; Bhattacharyya and Wi lliams, 1992). Figure 2.1 Reverse Osmosis – Pressure Applied to Reverse the Normal Osmotic Flow of Water (After, Williams, 2003) Figure 2.2 Osmosis Solvent (Normal Water) Passes Through a Semi-Permeable Barrier from Side of Low to High Solute Concentration (After, Williams, 2003) In 1960, Loeb Sourirajan developed a cellulose ac etate (CA) membrane that gave an adequate flux so that it could be used in t he industry (Baker, 2000). They are currently used, successfully, in many industrial ap plications. One advantage of a CA membrane is its chlorine tolerance but they are pH sensitive and less stable in organic solvents than in polyamides. Aromatic polyamides ha ve a much higher solvent resistance and may be used over a wider pH range (4-11) and ar e less susceptible to hydrolysis (Sagle and Freeman, 2005). Cellulose acetate membranes were widely used from 1 960 to mid 1970’s until Cadotte from North Star Research developed the inte rfacial polymerization method of producing composite membranes. Interfacial composi te membranes are characterized by extremely high salt rejection, combined with high w ater fluxes. Fluid Systems was the first company to commercialize the composite membra ne (Sagle and Freeman, 2005).

PAGE 35

9 2.3.1 Types of Reverse Osmosis Membranes RO membranes fall into two groups. They are asymmet ric membranes and thin – film composite membranes. Asymmetric membranes have a very thin, skin layer, supported on a more porous sub-layer (Figure 2.3). In this membrane, the dense skin layer determines fluxes and selectivities of these membranes and the sub-layer serves as a support for the skin layer. The support layer has very little effect on the member separation capacity. On the other hand, thin-film composite membranes consist of a thin polymer barrier layer, formed on one or more porous support layers Figure 2.4 (Baker, 2000). Figure 2.3 Asymmetric Membrane (After, Williams, 2003) Figure 2.4 Thin – Film Composite Membrane (After, Williams, 2003)

PAGE 36

10 2.4 Clays 2.4.1 Mineralogy Clay minerals consists mainly of aluminum or magnes ium silicate layers. Each aluminum and magnesium layer lies in between a sil ica, gibbsite or brucite layer.. In the silica (Figure 2.5), silicon atoms are each linked to four oxygen atoms in a tetrahedral arrangement. On the other hand, the gibbsite and b rucite layers (Figure 2.6) consist of two layers of oxygen atoms (or hydroxyl groups) lin ked to aluminum or magnesium atoms, in an octahedral arrangement (Stumm and Morg an, 1981). Figure 2.5 Tetrahedral Structure (After, Stumm and Morgan, 1981) Figure 2.6 Octahedral Structure (After, Stumm and Morgan, 1981)

PAGE 37

11 2.4.2 Kaolinite Kaolinite is a 1:1 layer clay mineral consisting o f repeating layers of tetrahedral and octahedral sheets (Figure 2.7). The repeating l ayers are linked by sharing oxygen atoms between octaherdral and tetrahedral layers. T he C-spacing which is the distance between atom centers in two repeating layers is 7.2 Figure 2.7 Ideal Kaolin Structure (After, Rouquerol et al., 1999) The chemical composition of a kaolinite unit cell i s given by [Al2(OH)4(Si2O5)]2. Kaolinite platelets usually, contain 100 or more la yers and are usually thick. The kaolinite particles tend to be hexagonal shape with diameters of up to 1 micrometer (Rouquerol et al., 1999).

PAGE 38

12 According to Buchanan and Oppenheim, 1972, as kaoli nite is a binary oxide mineral, its stability in contact with an aqueous m edium is a function of pH. The study of electrophoretic behavior of kaolinite will recogniz e that the nature of the surface, and therefore, the electrical double layer, will be pHdependent. 2.4.3 Bentonite Bentonite is a 2:1 layer clay (Figure 2.8) belongin g to a group of expanding clays that include montmorillonite (Rouquerol et al., 199 9). The basal spacing between the layers varies between 9.6 and 21 The chemical composition of montmorillonite is given by (OH)4Si8Al4O20.H2O. Isomorphic substitution and pH dependent charges developed on the surface hydroxyls groups at broken edges provides a permanent negative charge to montmorillonite layers with Na+ and K+ counter ions (Tombcz and Szekeres, 2004).

PAGE 39

13 Figure 2.8 Ideal Bentonite Clay Structure (After, R ouquerol et al., 1999) 2.4.4 Clay Particle Surface Charge Rengasamy and Oades, 1977, said that when clay part icles interact with simple and complex cations, clay particles’ surface morpho logy and charge change. Thus, influencing particles’ ion exchangeability and stab ility. They also noted that when simple cations are adsorbed onto clay by ion exchange, the net charge remains negative. They also said that the particle electrophoretic mobilit y is an indication of the net charge. However, according to them, when complex cations ar e adsorbed in excess of the exchange capacity, it can reverse the charge of the clay particle from negative to positive.

PAGE 40

14 2.5 Determining Surface Area of Particles Many of the popular methods for determining the su rface area of powders and porous materials depend on the measurement of adsor ptive capacity of the adsorbed. The Brunauer – Emmett Teller (BET) method is a popula r method for determining the surface area of adsorbents and catalysts (Rouquerol et al., 1999). 2.6 Electrokinetic Measurements and Zeta Potential Determination Surfaces of most materials develop an electrical c harge when brought into contact with an aqueous medium and conversely this surface charge influences the distribution of ions in the aqueous medium (Shaw, 1970). The electr ical double layer consists of two regions (Figure 2.9): an inner region, which includ es adsorbed ions, and a diffuse region. Distribution of ions in the diffuse region is influ enced by electrical forces and random thermal motion. Electrokinetic behavior of a partic le depends on the potential difference at the surface of shear and electrolyte solution. T his potential is called electrokinetic or zeta potential.

PAGE 41

15 Figure 2.9 Electrical Double Layer (After, Sawyer e t al., 1994) Colloidal particles in liquid either strongly bind or do not bind with the liquid. Colloidal particles that bind strongly with the liq uid are stable and hard to separate from the liquid. The stability of the colloid particle i s a function of the charge that the colloid exerts on the diffuse layer. Size of the particle, the surface area, and the surface charge of the particle affects the stability of a colloidal p article. Surface charge in turn is a function of pH and dissolved salt concentration (TDS). Zeta potential (Figure 2.9) measures the electrical potential of the diffuse layer at the sh ear plane within the solution (Brunelle, 1980). When zeta potential is high, the particles are very stable due to electrostatic repulsion and when it is close to zero, the particl es coagulate very easily.

PAGE 42

16 2.7 Membrane Fouling 2.7.1 Introduction The literature (Potts et al. 1981; Song et al., 200 4) provides a range of definitions for membrane fouling varying from a simple to compl ex definitions. The simplest of all definitions is “the phenomenon where ‘foulants’ acc umulate on RO membranes leading to performance deterioration” (Potts et al. 1981). Fouling can severely deteriorate the membrane performance and is of major concern in des ign and application of membrane separation processes (Chen et al., 2004; Probstein et al., 1981). 2.7.2 Effects of Fouling Membrane fouling has several negative effects, incl uding a decrease in water production, required increase in applied pressure, increased operational cost, a gradual membrane degradation, and a decrease in permeate qu ality (Seidel and Elimelech, 2002; Boerlage et al., 1998; Probstein et al. 1981). Fur ther, membrane cleaning to remove foulants results in increased energy and chemicals and also, the wastewater produced in cleaning membranes increases the costs of treatment Zhu and Elimelech, 1995 found that the relative permeate flux (permeate flux at a nytime during the fouling runs divided by the initial water flux) vs time as a convenient way to compare the membrane fouling results at different operating conditions, in labor atory scale experimental studies.

PAGE 43

17 2.7.3 Previous Work on RO Membrane Fouling Although extensive research had been carried out on MF and UF membrane fouling, the same cannot be said about the hyperfil tration (RO) membranes. Only a few experimental studies on colloidal or particulates f ouling of RO membranes are available in the literature (Zhu and Elimelech, 1997). Cohen and Probstein, 1986, investigated the cellulose acetate membrane fouling by ferric hydrox ide in deionized water. Their work demonstrated a linear relationship between permeate flux and the foulant layer thickness during fouling. Zhu and Elimelech, 1997, investiga ted the RO membrane fouling by aluminum oxide colloids and found out that fouling was significant at high ionic strength and fouling was reversible. In another study, Vrij enhoek et al., 2001, found that flux decline in colloidal fouling of RO and NF membranes is primarily due to “Cakeenhanced osmotic pressure.” They suggested that fl ux decline observed during the experiments were due to colloidal deposit layer lim iting back diffusion of salt ions from the membrane surface to the bulk solution, thus, in creasing the salt concentration at the membrane surface. Winfield, 1979, investigated the fouling of cellulose acetate RO membranes by secondary wastewater effluents and fou nd that for this wastewater, dissolved organic matter (eg. humic acid) plays a s ignificant role than particulate matter in membrane fouling.

PAGE 44

18 2.7.4 Fouling Material Materials that cause membrane fouling could be broa dly classified as; sparingly soluble inorganic compounds, colloidal or particulate matter, dissolved organi c compounds and biological matter. Sparingly soluble inorganic compounds are discussed in detail under the section 2.8. Colloidal or particulate matter comprise both inorg anic and organic matter. Aluminum silicate clays, rangin g in size from 0.3 to 1.0 m in diameter comprise of the most common inorganic mat ter that cause membrane fouling. Common organic matter comprises of living and senes cent organisms, cellular exudates, partially to extensively degraded detrital material It has been found that natural organic matter (fulvic and humic acid) with a combination o f divalent cations such as Ca+2 present in feed water, also, contributes to membran e fouling (Seidel and Elimelech, 2002; Schfer et al. 1998) by forming a cake layer on the membrane surface. Biological matter, the last fouling category mentioned refers to micro -organism, living or dead, and pyrogens. Membrane fouling can be caused by biologi cal matter through the process of attacking and decomposing the membrane and through the formation of a fouling layer on the membrane surface. 2.7.5 Clay Content in Water Sources The environment of a region could contribute to an increase in clay content in water resources. For example, the geology of the co ast range-mountains of the Pacificnorthwest produce high levels of colloidal clay in the surface water runoff. In these

PAGE 45

19 areas, during rain storms, heavy clay loads are ex perienced, which exceed the ability of pre-treatment units such as filters. In fact, it ha s being found that during such an event, on average, the montmorillonite and kaolinite clay con centration in surface water could reach up to a value of around 100 mg/l (Rooklidge e t al., 2002). 2.7.6 Fouling Mechanism in Microfiltration (MF) and Ultrafiltration (UF) Compared with Reverse Osmosis Membranes Unlike the vast amount of literature on ultrafiltra tion (UF) and microfiltration (MF) membranes, published research on the fouling m echanism of RO membranes is rather limited. Fouling mechanisms of UF and MF me mbranes are not directly applicable to RO because of the substantial differences in mem brane pore size and permeation rates. Pore blocking is an important mechanism in the foul ing of MF and UF membranes by colloids and macromolecules, but its role is not th at important in RO membrane fouling (Zhu and Elimelech, 1997, Yiantsios and Karabelas, 1998). 2.7.7 Membrane Surface Charge and Measurement Techn iques According to Gerard et al., 1998, membrane surface charge plays an important role in the fouling mechanism because it can functi on as an absorption site for foulants. A polymeric membrane acquires surface charge when bro ught into contact with an aqueous medium. Charged membrane surface influences the dis tribution of ions at the membranesolution interface resulting in co-ions being repe lled from the membrane surface and

PAGE 46

20 counter ions being attracted to it. Consequently, a n electrical double layer forms at the membrane surface (Childress and Elimelech, 1996). Both thin-film composite and cellulose acetate memb ranes exhibit characteristics of amphoteric surfaces containing acidic and basic functional groups. The iso-electric points of the composite and cellulose acetate membr anes are found to be at pH of 5.2 and 3.5, respectively. The difference in the zeta pote ntial values of the composite and cellulose acetate membranes is attributed to differ ences in the membrane surface chemistry (Zhu and Elimelech, 1997). Membrane zeta potential can be determined by measur ing streaming potential, sedimentation potential, electro-osmosis or electro phoresis (Childress and Elimelech, 1996). Childress and Elimelech in 1996, found that the ionic composition and concentration of the solution have a marked affect on the characteristics of the surface charge of polymeric membranes. 2.7.8 Fouling Tests Silt Density index (SDI) and the Modified Fouling I ndex (MFI) are two tests that measure particulate fouling potential of feed water Both are quick tests to stimulate membrane fouling by passing pretreated water throug h a 0.45 micron micro-filter under dead-end flow mode and at constant pressure. Of the two, Silt density Index (SDI) is the most common criterion for characterizing the foulin g propensity of feed waters. However, the criticism of this method is that it is based on empirical character and its inability to represent the foulants and their inter actions with the membrane, under actual

PAGE 47

21 operating conditions (Yiantsios and Karabelas, 1998 ). Further, there is no linear relationship between the index and particle concent ration. In contrast, the MFI is based on cake filtration and is suitable to model flux de cline in membrane systems. In addition, the index shows a linear relationship with the part icle concentration in the feed water. However, it does not satisfactorily correlate to co lloidal fouling observed in practice. This is because particles less than 0.45 microns, w hich are responsible for membrane fouling, are not measured in this test (Boerlage et al., 1998). 2.7.9 Effects of Fouling on Product Quality Solute rejection from a membrane is a function of t he relative solute selectivity of the fouling layer and the membrane. When the membr ane has a higher solute rejection than the foulant layer, hindered back diffusion of solutes occur causing the solute to accumulate near the membrane surface. Thi s enhanced concentration polarization results in a decrease in solute reject ion. However, when the fouling layer has a higher solute rejection than the membrane surface solute rejection improves (Ng and Elimelech, 2004). 2.7.10 Particle Deposition on Membrane Surface Most research, on particle deposition on a membrane surface and forces acting on it (Figure 2.10) is based on micro-filtration studi es. According to Fischer and Raasch, 1986, there are only two forces acting on a deposit ed particle on a membrane. They are

PAGE 48

22 the drag force parallel to and the pressure force p erpendicular to the membrane surface. However, subsequently, Lu and Ju, 1989 pointed out that there Figure 2.10 Forces Acting on a Particle Near a Membrane (After, Wiesner and Chellam, 1992) are four forces acting on a deposited particle: tan gential drag force, normal drag force, lateral drag force and gravity force. Wiesner and C hellam, 1992 reported that the Brownian diffusion and inter-particle forces such a s van der waals attraction and double layer repulsion are significant and should be consi dered when modeling forces acting on a particle.

PAGE 49

23 2.8 Scaling 2.8.1 Introduction Surface and groundwater contain ions such as calciu m, sulfate and bicarbonate. When water containing such ions are used as water s ources in boilers and membrane treatment processes, it leads to deposition of mine ral salt on its surfaces commonly known as scaling. Depending upon the water source, the scale deposits may consist of salts such as CaCO3, CaSO4 and SiO2. 2.8.2 Scale Formation in Membrane Systems and Its E ffects While colloidal or particulate fouling leads to los s of performance in the lead elements of a membrane unit, scaling tends to cause loss of performance in the trailing elements because the feed water becomes more concen trated and the solubility of ionic species in solution are approached. 2.8.3 Concentration Polarization Concentration polarization (Figure 2.11) occurs bec ause the concentration of dissolved species rejected by the membrane increase s at the membrane surface and is greater than in the bulk feed stream. When concent ration polarization increases, the

PAGE 50

24 osmotic pressure of the solution next to the membra ne surface increases, resulting in higher applied pressure Figure 2.11 Concentration Polarization (CP) within a Membrane Module required to obtain the same permeate quantity. Als o, salt concentration in permeate will increase with increased concentration polarization due to increased diffusion of salt through the membrane. Concentration polarization i s a function of system hydrodynamics and therefore can be controlled by ma intaining the feed flow within the turbulent flow regime. 2.8.4 Surface and Bulk Scaling Membrane scaling could take place whenever the surf ace is exposed to a supersaturation solution, or when there is crystals generating material present in suspended form in the feed solution together with n ecessary conditions for nucleation (Lee and Lee, 1999). The formation of super-saturat ion level closer to a surface-feed interface would lead to salt precipitation on the s urface. If the super-saturation region

PAGE 51

25 forms away from the surface, crystals would be form ed in the bulk solution and move towards the solid surface to form scale (Hasson et al., 1996) 2.8.5 Factors Affecting Scaling Studies have shown that factors such as pH, tempera ture, velocity/shear rate, surface material, geometry, surface roughness, supe rsaturation of ions influence scale formation (Sheikholeslami and Ng, 2001; Hamrouni an d Dhahbi, 2001). Further, presence of foreign matter also influence scale for mation by offering nucleation sites for crystal growth (Nancollas and Reddy, 1971). 2.8.5.1 Supersaturation This is the primary cause for mineral scaling. When the solubility product of calcium/carbonate and calcium/sulfate (e.g., solubi lity product of calcium carbonate is defined as the product of concentration of calcium and carbonate ions in saturated solution of calcium carbonate. Usually these values are given in text books and are based on infinite solutions of ions) exceeds their satura tion values, calcium carbonate, calcium sulfate precipitates and forms scale (Lee and Lee, 1999).

PAGE 52

26 2.8.5.2 Velocity and Shear Rate The scale accumulation rate is determined by the fo rces acting to bind formed deposit on to the membrane surface and hydrodynamic shear forces opposing the binding process. Lee and Lee, 2000 have found that as the v elocity increases and as the boundary layer decreases, surface crystallization decreases. 2.8.5.3 pH and Ionic Strength Above a pH value of 8, and CO3 2and SO4 2are the dominant ions in the solution and, hence, CaCO3 and CaSO4 scaling takes place very easily. Increasing the io nic strength of the solution increases the solubility o f salts, this increases the level of supersaturation at which crystallization will occur (Sheikholeslami and Ng, 2001). 2.8.5.4 Nucleation The interaction between ions or molecules that form scale leads to a formation of clusters. These clusters further interact with the ions or molecules to form new nuclei on which, further deposition of material could take pl ace. For the nuclei to form, the activation energy barrier of the nuclei has to be s urpassed (Stumm and Morgan, 1981).

PAGE 53

27 2.8.6 Cost of Scaling Scale formation leads to operational and maintenanc e problems and/or loss of efficiency. In RO system, scaling of membranes resu lts in decreasing plant efficiency and does require in higher pumping pressures. The cost due to scaling may be equivalent to about 10 % of the capital cost of the plant (Hamrou ni et al., 2001). 2.8.7 CaCO3 Scale Structure Calcium carbonate crystallizes in three different p hases. They are calcite aragonite and vaterite (Dydo et al., 2003). Out of the three phases, calcite (rhombohedric structure) is theoretically the only stable phase a t atmospheric pressure and within the 090oC. However aragonite (orthorhombic) and vaterite (h exagonal) can be obtained as metastable forms in relation with the conditions of nucleation/growth. Sheikholeslami and Ng, 2001 found that both mineral and organic im purities can also have a major influence on the crystal growth process. For exampl e, magnesium ions, which are present in sea water have the tendency to inhibit calcite f ormation but promote the aragonite formation.

PAGE 54

28 2.8.8 CaCO3 Scaling and Potential Determination In a solution, CaCO3 supersaturation level is given as: s sp sK CO Ca K K } }{ {2 3 2 += =d----------------(2.1) Where K* is the solubility product (M)2, Ksp is the thermodynamic solubility product (M)2 at equilibrium for the CaCO3 at the considered temperature while, {Ca2+} and {CO3 2-} are the activities of these ions (Gabrielli et al ., 1999 ), M is the number of moles per liter of solution When s>1, scaling will take place. Dydo et al., 2003 has identified the Langelier Saturation Index (LSI) as the most suitable method to determine CaCO3 solution scaling potential. The LSI originally dev eloped by Langelier is a calculated number used to predict the calcium carbo nate stability of water; that is, whether water will precipitate, dissolve, or be in equilibrium with calcium carbonate. Langelier Saturation Index (LSI) is defined for a f eed water of a membrane system as follows: LSI = pHR pHs----------------------------(2.2) Where; pHR = pH of the concentrate pHs = pH at saturation in CaCO3 and is defined as; pHs = (9.3 + A + B) – (C+D)-----------------(2.3) where A = (log10(TDS) – 10)/10, B = -13.12 x log10(T) + 34.55, C = log10(Ca2+ as CaCO3) – 0.4 and D = log10(alkalinity as CaCO3) In the above set of equations T is measured in Kelv in and TDS (total dissolved solids) in mg/l, Ca2+ and alkalinity in mg/l.

PAGE 55

29 If the LSI of the concentrate in a membrane unit is negative, CaCO3 tends to dissolve and if positive, CaCO3 precipitation will take place. 2.9 Fouling Models 2.9.1 Resistance in Series Model Many models have been proposed during the last two or three decades for predicting fouling in R.O. membranes (Barger and Ca rnahan 1991; Schippers et al., 1981). Out of these, the resistance in series model also known as cake filtration model given eq. 2.4 is the most popular model. Although derived for predicting permeate volumetric flux through ultrafiltration membranes, later, resistance in series model has also been successfully applied to reverse osmosis b y Belfort and Marx 1979; Schippers, et al., 1981; Kimura and Nakao, 1975; Timmer et al. 1994; Van Boxtel and Otten, 1993; Vrijenhoek et. al, 2002; Ng and Elimelech, 2004 for colloidal fouling, and by Okazaki and Kimura, 1984 for slightly soluble salts. Accord ing to Okazaki and Kimura, 1984, the permeate flux (Jv) can be determined by the permeate resistance due to membrane and scale layer. When a scale layer is formed, it has a resistance to the flow of water in series to that of membrane and therefore, the permeate flu x could be written as; s m vR R p J + D = ----------(2.4) where p is the pressure difference across the membrane, a nd Rm and Rs are the resistance of the membrane and fouling layer. Accor ding to this model, the total

PAGE 56

30 resistance of a membrane consists of two parts, the resistance of the clean membrane (Rm) and resistance of the fouling layer (Rs). While Rm is a constant, the Rs increases with time (Chen et al, 2004). The value of membrane resi stance (Rm) is found by passing D.I. water through the membrane and monitoring permeate flux and pressure. Scaling layer resistance (Rs) can be found by using CarmenKozeny equation. Fane, 1984, defined the resistance in series model in a slightly different form than eq 2.4 as ) (s mR R P J + D =m---------(2.5) with m p sA m Rb=----------(2.6) Where Am = membrane area (m2), r = specific resistance (mkg-1), mp = mass of the particles deposited (kg) and = permeate viscosity (Ns/m2) In eq. 2.6, CarmenKozeny equation is used for def ining r and is given as (Fane, 1984): 3 2) 1( 180e r e bp pd =------------(2.7) and used the model to predict particulate filtrati on and colloidal ultrafiltration. The terms np (kg/m3) dp (m) is the density of the particles, porosity and diame ter of the particles. However, this equation is valid for uniformly sized spherical particles. Endo and Alonso, 2001 proposed the following relat ionship for the r to take into account particle polydispersity and particle shape.

PAGE 57

31 3 2 2) ln4 exp( ) 1( 180e s r e k bg vg pd =---------------(2.8) Where, dvg is the geometric mean diameter (m) of dv, dv is the volume equivalent diameter (m) and g the geometric standard deviation (m), is the dynamic shape factor, which is defined as the ratio between the drag forc e acting on a sphere of the volume equivalent diameter. Using the r factor to calculate the Rs requires information on physical and geometrical properties of foulants. Thus, in pract ice, due to the presence of numerous foulants in feed water, calculating the r factor is not possible. 2.9.2 Concentration Polarization Model Salt rejecting membranes, with time, accumulate sal t on their surfaces which result in a resistance to permeation of water. Wi th an increase in salt concentration at the membrane, there will be a tendency for these salts to diffuse back to the bulk solution, which can be described by using the Fick’s law as D dC/dy (Bowen and Jenner, 1995).

PAGE 58

32 Figure 2.12 Concentration Polarization Model At steady state, by applying the mass balance for t he salt (Figure 2.13), the salt concentration within the concentration boundary lay er can be written as )9.2(=dy dC D qC qCp Above equation can be integrated by using the bound ary conditions y=0, C=Cm and y=, C=Cf ) 10.2(0= f mC C pC C dC q D dyd Hence, permeate flux (q) could be finally, written as: =p f p mC C C C D q lnd----------------(2.11) where q = permeate flux (m3/m2/s), D = diffusivity (m2/s), = boundary layer thickness (m), C = salt concentration at a point wi thin the boundary layer (mol/m3), Cm = q Cp Cf Cm Uniform mixing Concentration Boundary Layer () membrane Feed Flow

PAGE 59

33 salt concentration at the membrane surface (mol/m3), Cf = salt concentration in the feed (mol/m3), Cp = salt concentration in the permeate (mol/m3), Brian, 1966, was the first person to derive the equ ation (2.11) for a reverse osmosis system. This model has been developed for f eed salt solutions and cannot be applied when particles are present. 2.9.3 Gel Polarization Model Blatt et al., 1970 said that for a feed that contai ns macrosolutes/colloidal particles, as the concentration at the membrane surface increa ses, the macrosolute deposit layer resembles a solid or thioxotropic gels. For colloid al particles, the gel layer resembles a close-packed sphere. The thickness of the gel layer on the membrane continues to increase until the steady state is reached. At the steady state, permeate flux is only a function of the back diffusion. According to this model for 100 % rejection, the permeate flux could be written as; = =b g b gC C k C C D J ln lnd---------------(2.12) Where the diffusivity (D) be calculated from Stokes -Einstein relationship p Bd T K Dpm3 = -------(2.13) Where J = permeate flux (m3/m2/s), D = diffusivity (m2/s), = boundary layer thickness (m), Cg = gel layer concentration (mol/m3), Cb = bulk feed concentration (mol/m3), T = temperature (K), = viscosity (Ns/m2), dp = particle diameter (m), k =

PAGE 60

34 mass transfer coefficient (m/s) and KB = Boltzmann constant (JK-1). According to the Gel Polarization Model, the permeate flux is indepe ndent of the operating parameters such as applied pressure and cross flow velocity. T his model can predict the permeate flux of macromolecular solutes with reasonable accu racy, but it has been reported that for ultrafiltration of colloids, the difference bet ween the Gel Polarization Model predicted and the experimental flux is often one to two order s of magnitude. 2.9.4 Inertial Migration Model Serge and Silberberg, 1962 observed that particles flowing through a non-porous tube were subjected to radial forces, which transpo rt them away from the tube wall and the longitudinal axis. These particles, irrespecti ve of their entry point to the tube, reached a certain equilibrium position at about 0.6 tube ra dii from the longitudinal axis, which is not time dependent. However, Altena et al., 1983 s aid that particles with a radius larger than 1.0 micron are affected by the lift velocity w hile submicron particles tend to be transported to the membrane wall by permeation drag 2.9.5 Shear Induced Hydrodynamic Convection Model Madsen, 1977 and Altena and Belfort, 1984 found tha t inertial lift velocity is less than the permeation velocity in cross-flow filtrati on. This results in the formation of a

PAGE 61

35 cake layer on the membrane surface. Blatt et al., 1 970 said that the cake thickness reaches an equilibrium when the convective particles transp orting towards the membrane are equal to the particles being removed by the shear f orce generated by the moving fluid. 2.9.6 Shear Induced Hydrodynamic Diffusion Model According to this model, the thickness of the cake layer will continue to increase until the cross-flow, which induces a shear stress on the cake layer, is adequate to cause shearing of the outer particles in the cake layer. The effect of this shear-induced motion is the net transport of particles in the direction of decreasing particle concentration. Schwinge et al., 2002, used shear induced diffusion model to predict foulant built up on a reverse osmosis membrane. According to him fouling took place only when J>Jcrit. He found the deposited mass on a membrane is equa l to t C J J mpart crit solvent cD = D) (----------(2.14) Where mc = deposited mass of particles (kg) during the time interval t (s), Jcrit (m3/m2/s) is the critical flux (critical flux is the perm eate flux at which no further fouling would take place on the membrane) and is a function of the feed velocity as given by the following equation 1 1Reb a Jcrit+ =----------(2.15) Where a1 and b1 are constants that have to be found experimentally for an experimental unit that has feed spacer in a diamond orientation. Jsolvent (m3/m2/s) is the solvent flux Cpart (mg/l) is the foulant concentration and Re, the Rey nolds Number. If this model is to be used, one needs to first determine t he Jcrit,, which is a function of the

PAGE 62

36 crossflow velocity. If the experimental unit is suc h that the crossflow velocity has very little impact on the permeate flux, then this model is of little use. 2.9.7 Scour Model Scour model for ultrafiltration of particles was su ggested by Fane, 1984. This model is based upon the scouring control of the cak e layer by the tangential flow of the feed solution. In this model, the rate of solids t ransported towards the membrane surface is balanced by the rate of scour. This relationship could be defined by the following expression: c c bC dt d JC =d---------(2.16) Where J is permeate flux (m3/m2/s), c is the cake thickness (m), Cc the cake concentration (kg/m3) and Cb the bulk concentration (kg/m3) and .g d e cK dt d =-----------------(2.17) where Ke (m) is the erosion coefficient and .g (s-1) is the shear rate. A drawback of this model is the non-inclusion of the particle size. Also, if the flow rate does not influence the permeate flux, so would the rate of s couring and under these circumstances the model will be of no use.

PAGE 63

37 2.9.8 Turbulent Burst Model Cleaver and Yates, 1973, 1976 suggested that partic les deposited onto a smooth impermeable wall can be removed by a “turbulent bur st”. According to them, the turbulent burst removes deposited material from a m embrane surface by seeping into the laminar sub-layer formed adjacent to the membrane a nd removing the fouling layer. Using this principle, Gutman, 1977 modeled the perm eate flux with time for reverse osmosis membranes fouled by suspended and colloidal material. He defined the net rate of fouling of the membrane surface to be equal to t he difference between the rate of deposition and re-entrainment as; e dr r dt dm =--------------(2.18) where m = deposited foulant mass (kg), rd and re are rate of deposition (kgm-2s-1) and re-entrainment (kgm-2s-1). The rate of deposition given in (2.18) is depen dent on the reynolds number and is given by b dC J k r)2/ (2+ =-------------------------(2.19) In applying this model, the hydraulic resistance of the fouling layer (Rf) is assumed to be proportional to the weight of the fou lant deposited. Rf=Km ------------------------(2.20) and then finally arriving at the following model n r + + =b b b f b b fBU C t JBU J J BU C J J* *1 1 exp 1 1a a ------------(2.21)

PAGE 64

38 When J>2k2 and n r n r + + + + + =+ -t C BU JBU k BU J k J k BU C J Jb b b b C BU k JBU f b b fb b b* 2 ) ( 2 2 2 *(2 exp 1 / 2 1 / 2 1 12a aa---(2.22) When J<2k2 where, J = permeate flux of the un-fouled membrane (m3/m2/s), Jf = permeate flux of the fouled membrane (m3/m2/s), Rm, Rf = hydraulic resistance of the membrane and the fouling layer (Nsm-3), K = hydraulic resistance of the fouling layer pe r unit mass (Nsm-1kg-1), Ub = bulk velocity (ms-1), Cb = bulk feed concentration (kg/m3), m = mass of foulants deposited per unit area (kg/m2), k2 = mass transfer coefficient (md-1), t = time (s) and B (sm-2) and = K/Rm (m2/kg-1) are coefficients. This model is dependent on the crossflow velocity and, therefore, if the crossflow velocity has very little impact on the permeate flux this model is of no use.

PAGE 65

39 Chapter 3 Experimental Methodology This chapter presents the technical procedure utili zed in performing the experimental part of this research. The chemical pr eparation and the instrumentation are described first, followed by the experimental proce dure. 3.1 Electrokinetic Mobility and Zeta Potential Meas urement 3.1.1 Materials and Chemicals All chemicals used in these experiments were ACS ce rtified or better. For practical purposes (reasons are described in the re sults section), it was decided to use kaolin concentrations of 50, 150 and 250 mg/l, resp ectively. For the CaCO3 solutions, concentrations of 0.0005 and 0.0015 M were used by mixing equimolar CaCl2 and Na2CO3 solutions. To prepare kaolin suspensions, initiall y a measured quantity of kaolin clay (Sigma-Aldrich Corporation, St. Louis, MO) was mixed in 1.0 L of distilled water and stirred. Next, 100 ml samples were selected for EM (Electrokinetic Mobility) measurements. Prior to carrying out EM measurements pH of the samples were measured with OAKTON 510 pH/Ion meter (Eutech Instr uments, Vernon Hills, IL) and

PAGE 66

40 then the pH of the samples were adjusted either by using 0.1 M NaOH or HCl solutions. This same procedure was for the kaolin clay samples mixed in 0.0005 or 0.0015 M CaCO3 solutions. 3.1.2 Measuring Instrument and Technique The EM measurements were carried out with a ZM-80 ( Zeta Meter Inc., Staunton VA) Zetameter (Figure 3.1) using the micro-electrop horesis principle, which measures the mobility of charged particles by determining th eir rate of movement in a DC voltage field. The clay suspension to be tested was first p laced in an electrophoresis cell (Figure 3.2). Electrodes were inserted at each end of the c ell and were connected to the power supply, and the cell then placed on the mirrored ce ll holder that permits a light beam to focus on the cylindrical glass tube of the cell. Th e voltage to be applied was determined by the specific conductivity of the solution and is given in table 3.1. The time taken by a particle to travel a particular micrometer distance was measured using model Zeiss DR microscope (Figure 3.3). To minimize reading error, a minimum of 10 particles were tracked and their average time of travel was calcul ated. In the ZM-80 electrophoresis cell, the distance tra versed for one full micrometer division is 160 microns. The voltage field strength is the overall applied voltage divided by the length of the cell tube which is 10 cm. The equation for the EM becomes (ZetaMeter Manual ZM -80); )1.3( 10 160 = v cm t microns EM

PAGE 67

41 Figure 3.1 ZM-80 Apparatus Figure 3.2 GT-20 Electrophoresis Cell

PAGE 68

42 Figure 3.3 Zeiss DR Microscope )2.3( V t 1600 EM = Where; t = time (in seconds) to traverse one full d ivision, V = applied voltage in volts and EM = Electrophoretic mobility of particles in m icrons per second/volts per centimeter. The Helmholtz-Smoluchowski equation is the most ele mentary expression for calculating Zeta potential from EM (Zeta-Meter Manu al ZM -80) and is given by: )3.3( ) ( ) ( 4 = EM D V ZPq q p

PAGE 69

43 Where; EM = Electrophoretic mobility at actual temp erature, V() = Viscosity of the suspending liquid in poises at temperature “”, D() = Dielectric constant of the suspending liquid at temperature “”, ZP = voltage in electrostatic units. Table 3.1 Specific Conductance and the Recommended Maximum Applied Voltage Relationship for ZM -80 Specific Conductance (Microhms/cm) Recommended Maximum Applied Voltage (DC) Less than 300 300 700 200 1,500 133 3,000 100 6,000 67 10,000 50 20,000 40 30,000 30 40,000 25 60,000 20 3.2 Membrane Characterizing, Fouling, Scaling and M odeling Experiments 3.2.1 Design Philosophy In the fouling studies, permeation quantity and qua lity data are required to determine the membrane performance. Therefore, when designing the experiments, focus should be paid on the proper selection of a membran e, feed solution and its concentration, and the membrane module. The operating conditions s hould cover a wide range and provide a stable system performance. The selected m embrane module must have a simple geometry.

PAGE 70

44 3.2.2 Experimental Unit A schematic diagram of the laboratory scale crossf low RO test unit is shown in Figure 3.4. All the experimental runs were carried out in complete recirculation mode. The membrane test unit consisted of a membrane cell a high pressure positive displacement pump/motor, a feed reservoir, a mixer and a temperature control system. In this unit, the test solution was held in a 100 lite r feed reservoir (1) and fed to the membrane cell by a constant-flow diaphragm pump (2) (Model M-03-E made by HydraCell, Wanner Engineering Inc. Mineapolis, Minnesota ), capable of providing a maximum pressure of 6,900 kPa and a maximum flow of 6.93 x 10-5 m3/s (1.1 gpm). A Blacoh H1020B pulsation dampener (CAT) (10) is installed a t the outlet port of the pump to eliminate vibrations created by pumping. This dampe ner is initially pressurized at 1,380 kPa using an air hand pump and, then, gradually, de pressurized so that the dampener bladder pressure is always maintained at half the s ystem operating pressure. The feed that contained clay particles or scaling salts or a mixt ure of both were held in suspension in the feed tank by continuous agitation with a mixer (12). The crossflow velocity in the membrane cell (4) is controlled by a bypass valve w hich split the flow such that only a portion of the test solution goes to RO membrane un it. The desired trans-membrane pressure is set by throttling the needle valve on t he concentrate side of the membrane. Temperature is maintained at 23oC by circulating water through a chiller (3) (ColeParmer Polystat model). All pipes and fittings on the high pressure side, which were up to the flow control valve on the concentrate side of the membrane and u p to the by-pass line flow control

PAGE 71

45 valve were of stainless steel type 316. In all othe r locations, braided Polyvinyl Chloride (PVC) hosings were used. Polyvinylidene Fluoride (P VDF) pipes were used in the permeate collection line. Figure 3.4 Schematic Diagram of the Experimental Se t-up 3.2.3 Membrane Type and Specifications Commercially available thin film composite LFC1 pol yamide reverse osmosis flat sheet membranes manufactured by Hydranutics, Oceans ide, CA were used in the experiments. The membrane’s operating pH range was between 3 to 10. According to manufacturer’s information, its NaCl rejection capa city is 99.4% and is capable of producing 29 GFD at a test pressure of 1,550 kPa an d at a temperature of 25o C when used with 1500 ppm NaCl solution. Surface property of Low Fouling Composite (LFC 1)

PAGE 72

46 membrane has been modified during casting process t o provide the membrane with low fouling characteristics. The membrane maintains a r elatively neutral surface charge over both acidic and basic pH environment. Its hydrophi licity is 47o (Gerard et al., 1998). Membranes were delivered in flat sheet forms to the laboratory. The membranes were cut into 7.5 x 5.5 inch coupons as required by the membrane filtration cell, and, necessary holes made using a template. Each membra ne coupon was washed with deionized water and then stored in fresh deionized water at 5o C, prior to using in scaling or fouling experiments. Each new membrane was place d in the membrane cell with the skin layer facing the high pressure side of the mem brane. Used membrane was replaced by a new membrane at the beginning of each new expe rimental run. 3.2.4 Membrane Cell and Holder The choice of the membrane module is important as it influences the hydrodynamics on the membrane surface. In order to keep the hydrodynamics uniform in all the experimental runs, a stainless steel recta ngular flat sheet membrane cell (Sepa CF II Membrane Cell System), manufactured by GE water Technologies was used. This cell consisted of two rectangular type 316 stainless ste el plates (Figs. 3.5, 3.6, 3.7 and 3.8) and the cell body dimensions of this unit were 16.5 cm x 21.3 x 5.2 cm. The membrane piece was held between these two plates and subject ed to high feed pressure by sealing the steel plates with two rubber rings. The feed wa ter channel dimensions of the cells were 14.6 and 9.5 cm for channel length (Lc) and wi dth (Wc), respectively. The channel

PAGE 73

47 Figure 3.5 Membrane Cell Figure 3.6 Reverse Osmosis Flat Sheet Membrane Cell – Bottom Side

PAGE 74

48 Figure 3.7 Reverse Osmosis Flat Sheet Membrane Cell – Top Side Figure 3.8 Reverse Osmosis Flat Sheet Membrane Cell – Assembled View height (Hc) was 0.2 cm. This gave a feed cross sect ional area of 1.9 x 10-4 m2. The active surface area of the membrane was 139 cm2. This membrane cell was held in a stainless

PAGE 75

49 steel holder, which had 6 steel bolts to seal the m embrane compartment. The bottom side of the membrane cell is referred to as high pressur e side or feed side and the top side is referred to as the low pressure or permeate side. S epa CF II was a lab scale cross-flow membrane filtration unit that provided fast and acc urate performance data with minimal amounts of membrane, product, expense, and time. It s design simulated the flow dynamics of larger, commercially available membrane elements, such as spirally-wound membrane elements. The other principal instruments used in the experim ents and their specifications, including manufacturer are given in table 3.2. Table 3.2 Instruments and their Specifications No: Instrument Manufacturer Model Calibration Required Range Accuracy 1 Analytical Balance Mettler Toledo Delta Range AE 260 Yes 0-81 g 0.1 mg 2 Electronic Top loading Balance Ohaus adventurer Yes 0-3100g 0.1g 3 do do 4 Conductivity meter & probe YSI Field YSI Conductivity meter Yes 0-2 S/cm 0-20 S/cm 0-200 S/cm 0-2000 S/cm 0.2% 0.15% 0.1% 0.15% 5 Flowmeter McMillan Model 111 Yes 20-200 mL/min 100-2000 mL/min 3% 6 Presssure Gauge Wika 13x.53 Yes 0-1000 psig 0.15% 7 Stopwatch Cole-Parmer Easy-Grip Stopwatch Yes 24 h 1.5% s/day 8 Volumetric Flask Kimble Serialized and Certified Class A with Stopper Yes 0-10 mL 0.2%

PAGE 76

50 3.2.5 Experimental Procedure for Membrane Character izing, Fouling, Scaling and Combined Fouling and Scaling Runs LFC 1 membrane is a brackish water membrane that is operated under low pressure conditions. Considering the LFC 1 membrane manufacturer recommended operating pressure (1,550 kPa), the R.O. membrane e xperiments for pure water, fouling and scaling runs were carried out at transmembrane pressures of 1,380, 2,070, 2,760 and 3,450 kPa values. These pressure values correspond to 200, 300, 400 and 500 psi. In all the experiments, crossflow velocity values of 4.04, 8.08, 12.9 and 26.26 cm/s were used. These values were based on the typical crossflow ve locity range that could be measured by the flow meter of the experimental set-up. In th e clay fouling runs, a kaolin concentration of 50, 150 and 250 mg/l were chosen t o cover the average concentration of clay particles that is present in surface water sou rces after a rainfall. In the CaCO3 scaling runs, an equimolar of 0.0005M CaCl2 plus 0.0005 M Na2CO3 and 0.0015M CaCl2 plus 0.0015 M Na2CO3 solutions were used. A low salt concentration was u sed to induce concentration polarization and surface cryst allization of CaCO3 but to prevent bulk crystallization from taking place in the feed container. 3.2.5.1 Membrane Characterizing At first, the membrane was characterized for a know n salt (CaCl2). This was done by varying transmembrane pressure from 1,380 to 3,4 50 kPa and crossflow velocity at 4.04 cm/s and 26.26 cm/s and by monitoring the stea dy state permeate flow and quality.

PAGE 77

51 The transmembrane pressure was measured by a precis ion pressure gauge (WIKA).The feed flow rate was measured with the digital flow m eter (McMillan, Model 111). The feed solution and the permeate salt concentration w as measured by a conductivity meter (YSI -35) and a conductivity probe (YSI -3417). The total permeate volumetric flux through the membrane was measured by collecting the permeate for a defined period of time and then weighing the permeate using a analyti cal balance. The collection time depended upon the permeation rate, but in general, 2-3 minutes were sufficient to collect around 20 ml of permeate. The collected amount was then used to measure the conductivity. The pH of the solution was monitored with a OAKTON 510 pH/Ion meter (Eutech Instruments, Vernon Hills, IL). The tempera ture of the feed solution was measured using a Thermometer (Flinn Sc. Co., USA). The temperature was maintained nearly constant in the range 231oC for all data sets. 3.2.5.2 Compaction of the Membrane and Pure Water P ermeability Before using the membrane coupons in any of the exp erimental runs that are described in latter sections of this chapter, disti lled water (conductivity 1.5 2.0 ohm/cm) was circulated within the unit at a transme mbrane pressure of 3,450 kPa and at a crossflow velocity of 26.26 cm/s for up to 12 hrs to dissociate flux decline due to membrane compaction and other unknown causes inhere nt in lab-scale recirculation system. Flux was monitored continuously for the dur ation of the experiment. The values, thus obtained, were used as baseline values for the fouling and scaling experiments.

PAGE 78

52 3.2.5.3 Kaolin Fouling Experiments For fouling runs with kaolin clay, for pre-compact ed membrane (3.2.5.2), first filtration was carried out with distilled water (co nductivity 2.0 ohm/cm) at the experimental pressures (1,380, 2,070, 2,760 and 3,4 50 kPa) and at cross flow velocity values (4.04, 8.08, 12.9 and 26.26 cm/s) appropriat e for that particular run until a steady state (none or very little variation of permeate fl ux with time) is reached. Later, this steady state value was used to calculate the pure w ater permeability of the membrane. Once the steady state value was reached, the kaolin particles, at the required concentration were quickly added to the system and mixed thoroughly, and, then, permeate flux was monitored again until the steady state was reached. Throughout the runs, clay particles were held in suspension by agi tating the contents of the feed tank. If the steady state was not reached by at least 6 hrs, the experiment was curtailed at the end of six hours. In addition to the permeate flux, fee d and permeate conductivity was also continuously monitored, in order to check the influ ence the addition of kaolin clay particles into feed solution, had on feed conductiv ity. 3.2.5.4 CaCO3 Scaling Experiments In the scaling runs, like in the fouling runs, firs t filtration was carried out with distilled water (conductivity 2.0 ohm/cm) at the experimental pressures (1,380, 2,070 and 3,450 kPa) and at cross flow rates (4.04, 8.08, 12.9 and 26.26 cm/s) appropriate for that particular run, until, a steady state (none or very little variation of permeate flux with

PAGE 79

53 time) was reached. The scaling solution was prepare d by first making individual CaCl2 and Na2CO3 salt solutions of desired concentrations, separate ly. Thereafter these solutions were transferred into the feed container and mixed further. Next, permeate flux and permeate and feed conductivity, together with p H values was continuously monitored with time until the steady state was reached. If t he steady state was not reached by at least 6 hrs, the experiments were curtailed at the end of six hours. 3.2.5.5 CaCO3 Scaling and Kaolin Fouling Experiments For combined scaling and fouling runs, as in the in dividual fouling and scaling runs, first filtration was carried out with distil led water at the experimental pressures (1,380, 2,070, 2,760 and 3,450 kPa) and cross flow rates (4.04, 8.08, 12.9 and 26.26 cm/s) appropriate for that particular run until a s teady state was reached. Next, kaolin and individual salt solutions were added separately to the feed container and permeate flux. Then, permeate and feed conductivity, together with pH, was continuously monitored until the steady state was reached. Throughout the experiment, clay particles were held in suspension by agitating the contents of the feed ta nk. If the steady state was not reached within 6 hrs, the experiment was curtailed at the e nd of six hours. At the end of this run, the system was cleaned, first with tap water, follo wed by deionized water. Next, in order to check the effect of kaolin fouling only, another run was carried out at a pH = 4.00 by adding 0.01 M CH3COOH to D.I. water. During these runs, the permeate flux was monitored until the flux recovery reached a steady state. Again the system cleaning

PAGE 80

54 procedure was repeated, first with tap water, follo wed by D. I. water. Finally, the cleaned membrane’s pure water flux was measured by running the unit with the D. I. water. 3.2.5.6 Membrane and System Cleaning after Operatio n At the end of each fouling and scaling run, fouled or scaled membrane in the membrane cell was replaced by the “washable” membra ne (membrane used only for cleaning the unit). The feed water in the feed con tainer was replaced by tap water, which was made to run through the unit to flush the entir e system. Thereafter, to further clean the system of fouling particles and/or salts, the tap water in the feed tank was replaced by distilled water (conductivity 1.5 2.0 ohm/cm), which was made to run through the system.

PAGE 81

55 Chapter 4 Results and Discussion In the first section of this chapter the absorbate particles (kaolin and bentonite) and the results of zeta potential measurements are characterized. Part of this research focused on the clay fouling of a Reverse Osmosis (R .O.) membrane and the effects of clay CaCO3 interactions on R.O. membrane fouling. The fouling could be influenced by the charge on a clay particle. Hence the zeta poten tial of kaolin particles both in distilled water and in CaCl2 and Na2CO3 solutions of varying concentration and under diffe rent pH conditions was investigated. Results of the purewater permeation experiment usi ng the LFC 1 (Low Fouling Composite 1) membrane are discussed. This data was used to determine the membrane resistance. In the same section, permeate data will be used to evaluate the performance of the LFC 1 membrane operating at two extreme crossfl ow velocities. The third section presents the results of kaolin c lay fouling on a LFC 1 reverse osmosis membrane. Then studies were conducted to de termine the effect of operating parameters (trans-membrane pressure, crossflow velo city, feed concentration) on flux decline. For comparison purposes, data from bentoni te clay fouling experiments were used.

PAGE 82

56 In the fourth section membrane scaling studies were conducted using CaCO3 solutions. In this section the effects of operating parameters, on membrane scaling were investigated, which included the effects of changes in pH. The final section discusses the interaction between salts and clay particles and th eir effects on membrane performance. 4.1 Characterization of Clay Particles 4.1.2 Surface Area The specific surface area of kaolin and bentonite p articles obtained from BET (Brunauer – Emmett – Teller) measurements using Mon osorb (Quantachrome Corporation, Boynton Beach, FL) were 19.2 m2/g and 11.76 m2/g respectively. The value for kaolin is within the range (10-20 m2/g) given for natural kaolinite. Although values for bentonite clay are not available, the values fo r montmorillonites clays which is the primary clay mineral (70%) present in bentonite are often around 30 m2/g (Rouquerol et. al, 1999). 4.1.3 Particle Size Analysis Particle size analysis for kaolin clay sample is gi ven in Appendix B. From the Figures B.2, B.3 and B.4 it can be seen that at pH= 9.0, as the CaCl2 and Na2CO3 concentration increases, the number of large partic les tend to increase. This characteristic

PAGE 83

57 is more pronounced in figure B.4. This could be a r esult of kaolin particles acting as nuclei for the CaCO3 crystallization. 4.1.4 Zeta Potential Measurement 4.1.4.1 Kaolin in Distilled Water and Salt Solution s The data presented in figure 4.1 shows the Zeta pot ential (ZP) of kaolin as a function of pH values between 3 to 10 in DI water a nd in combined salt solutions. In distilled water, the measurements show that Kaolin particles display a negative zeta potential that varies from 0 to -16 millivolt (mv) over the pH range between 3 to10. As the pH increases, although the negative zeta potent ial of clay particles increases, the rate of ZP increase decreases. The increase in negative ZP value with pH may be due to charge developing at edges by direct transfer of H+ from the clay particle to the water. 4.1.4.2 Kaolin in Combined Salt Solution Figure 4.1 also shows the behavior of kaolin in sal t solutions. As the concentration of salts increases from 0.0 M (no sal t) to 0.0015M, the ZP values decrease gradually and reach zero at 0.0015 M. At 0.0015 M s alt solutions, pH value of the solution has no effect on clay particle ZP. Also, a t the clay and salt concentrations that were tested, no charge reversal of kaolin particles was observed.

PAGE 84

58 y = -0.1078x3 + 2.4739x2 19.424x + 38.6 R2 = 0.9605 y = 0.0748x3 1.3137x2 + 5.1982x 5.4043 R2 = 0.9805-20 -16 -12 -8 -4 0 34567891011pH Zeta potential (mv) Kaolin in D.I Water Kaolin in 0.0005 M CaCl2 and Na2CO3Kaolin in 0.0015 M CaCl2 and Na2CO3 Figure 4.1 Zeta Potential (mv) vs pH for Kaolin in Distilled Water and Scaling Solutions. Kaolin Concentration = 150 mg/l 4.2 Membrane Characterization 4.2.1 Purewater Permeation Tests for the LFC 1 Memb rane As explained in chapter 3, permeability tests were carried out to determine the purewater transport coefficient (Kw) of the LFC 1 membrane. The regression analysis results are given in figure 4.2. Figure 4.2 shows t hat a linear regression has an R2 value of

PAGE 85

59 0.998. From the linear regression analysis it can b e seen that there is a perfect linear relationship between the purewater flux and transme mbrane pressure.The results also indicate that there is no transmembrane pressure ef fects on the water permeability of the membrane. Jw = 1.097*10-11 P R2 = 0.99788 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 01234 Transmembrane pressure (Pa x 10-6)Pure water flux (gm/cm2.s*105) Run 1 Run 2 Run 3 Run 4Figure 4.2 Pure Water Flux vs Transmembrane Pressu re for LFC 1 Membrane The slope of the plot gives the pure water coeffici ent value of 1.097*10-11 (gm/cm2.s.Pa) for the LFC 1 membrane. The clean membrane resistance, which is the membrane resistance to water flow, is the most impo rtant characteristics of RO membrane. The clean membrane resistance value, whi ch is the reciprocal of Kw, is 9.11 x 1010 Pa.s.cm2/gm. The result also shows that the clean membrane resistance is a constant for the LFC 1 membrane and the applied pressure doe s not cause any membrane compaction within the experimented transmembrane pr essure range.

PAGE 86

60 4.3 Membrane Performance The comparison of permeate flux vs applied pressur e with solutions of CaCl2 at 0.0005 M and 0.001 M at two different velocities ar e shown in Figure 4.3 and 4.4. As the pressure increases, the rate of increase of the per meate flux for salt solutions tend to decrease, resulting in different behavior than for D.I. water run. Also, for the same feed concentration, permeate flux tends to increase with increasing cross flow velocity. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 01234 Transmembrane pressure (Pa*10-6)Permeate vol. flux (m3/m2/s*105) D.I 4.04 cm/s 26.26 cm/s Figure 4.3 Purewater Volumetric Flux vs Transmembrane Pressure for 0.0005 M CaCl2 for LFC 1 Membrane

PAGE 87

61 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 01234 Transmembrane pressure (Pa*10-6)Permeate vol. flux (m3/m2/s*105) D.I 4.04 cm/s 26.26 cm/s Figure 4.4 Purewater Volumetric Flux vs Transmembrane Pressure for 0.001 M CaCl2 for LFC 1 Membrane These results can be explained as follows. As the p ermeate flow through the membrane, due to the rejection of salt, salt concen tration at the membrane surface increases, which inturn increases the osmotic press ure at the membrane. Increased osmotic pressure reduces the net driving force thro ugh the membrane and hence a reduction in the permeate flux. However when the cr ossflow velocity is increased, it causes a decrease in the boundary layer thickness a nd salt concentration at the membrane surface. This results in a lower osmotic pressure l eading to a higher permeate flux. .Although the strength of the feed solution used is relatively small (0.0005 M and 0.001 M), still, there is concentration polarization taki ng place at the membrane surface.

PAGE 88

62 4.4 Membrane Fouling Runs with Kaolin Transmembrane pressure (1,380 – 3,450 kPa), kaolin concentration (50 250 mg/l) and crossflow velocity (4.04 and 26.26 cm/s) were selected as the independent variables in these studies. Permeate flux was consi dered the dependable variable and monitored with time. Further, feedwater and permeat e conductivity were constantly monitored over time to determine if any leaching of ions is taking place from kaolin particles. Low Fowling Composite (LFC 1) membrane was used an d operating pressure range between 1,380 – 3,450 kPa was selected for t he experimental work. The reasons for choosing 50-250 mg/l Kaolin concentration are, one, during heavy rain or watershed land slides, on average, the montmorillonite and ka olinite clay concentration in the water could reach a value around 100 mg/l, two, high part icle concentrations are used in laboratory scale membrane experiments so that fouli ng may be observed within a reasonable time frame. 4.4.1 Flux-Time Relationship Figure 4.5 shows the decline in relative permeate f lux vs time for various kaolin clay concentration and different operating pressures. Th e following empirical observations are noted from figure 4.5. 1. A linear permeate volumetric flux vs time relati onship 2. The existence of an operating condition at which no flux decline could be found.

PAGE 89

63 4.4.2 Linear Flux vs Time Relationship It was observed that for the operating pressure of 3,450 kPa, the permeate flux decline was linear with time within the duration of the experiments (6 hrs). A maximum of 20 to 32 % decline in relative flux (defined as the permeate volumetric flux at any time divided by the permeate volumetric flux at the star t of the experiment) was observed. According to Cohen and Probsetin, 1986, this linear behavior suggests that the foulant cake growth is linear in time. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0246 Time (Hrs.)Rel. Per. Vol. Flux 1,380 kPa, 150 mg/l 3,450 kPa, 150 mg/l 3,450 kPa, 50 mg/l 1,380 kPa, 50 mg/lFigure 4.5 Rel. Permeate Flux vs Time for Applied P ressure = 1,380 kPa and 3,450 kPa, and Kaolin Concentration s of 50 and 150 mg/l and Crossflow Velocity = 4.04 cm/s, pH = 6.7

PAGE 90

64 4.4.3 Comparison of Kaolin with Bentonite Clay Foul ing Figure 4.6 shows the relationship of relative perme ate volumetric flux vs time for kaolin and bentonite clay for 150 mg/l and 100 mg/l solution concentration and varying pressure. The most interesting feature in this graph is the flux at an operating p ressure of 1,380 kPa for both clays within the operating period with no flux decline being obs erved. As the pressure is increased to 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0123456 Time (Hrs.) Rel. per vol. flux Ben 100 mg/l, 1,380 kPa Ben 100 mg/l, 2,070 kPa Ben 100 mg/l, 3,450 kPa Kaolin 150 mg/l, 1,380 kPa Kaolin 150 mg/l, 2,070 kPa Kaolin 150 mg/l, 3,450 kPa Figure 4.6 Rel. Per. Vol. Flux vs Time for Kaolin a nd Bentonite. Applied Pressure = 1,380 – 3,850 kPa, Kaolin Con centrations of 150 mg/l, Bentonite Concentration 100 mg/l and Cro ssflow Velocity = 4.04 cm/s

PAGE 91

65 2,070 and 3,450 kPa, a higher flux decline was obse rved in the bentonite experiments, although they have lower clay concentration (100 mg /l) than in kaolin (150 mg/l). The permeate flux reduction occurs due to the depositio n of a fouling layer on the membrane. bentonite is a clay generated predominantly from sm ectite minerals, which displays swelling when exposed to water. The higher flux dec line observed from the bentonite experiments appears to be due to the swelling prope rty of bentonite when exposed to water, which does not occur with kaolin clays. When the fouling layer accumulates on the membrane surface, bentonite will start to expand an d reduce the permeability of the fouling layer. 4.4.4 Effects of Operating Variables on Flux 4.4.4.1 Applied Pressure Figure 4.7 shows the effects of applied pressure on flux decline. The results show that for the same kaolin concentration and flow vel ocity, application of a higher pressure yields a higher permeate flux at the beginning but leads to a greater flux decline. As the transmembrane pressure reduces, the rate of flux ch anging also reduces. However, at a pressure of 1,380 kPa, flux decline is not visible. This feature shows the importance of the applied pr essure force in filtration and cake formation. For the cake layer to be formed on the membrane, clay particles have to be transported near to the membrane surface. Then p articles have to get attached to the

PAGE 92

66 membrane surface, which needs a force that could ov ercome the particle -membrane surface repulsive force. Once the first layer is fo rmed, the particles have to y = -0.1765x + 3.8621, R2 = 0.9723 y = -0.0712x + 3.0473, R2 = 0.8238 y = -0.0104x + 2.2532, R2 = 0.442 y = 0.0001x + 1.4629, R2 = 0.00925 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0123456Time (Hrs)Per. flux (m3/m2/s*105) 3,450 kPa 2,760 kPa 2,070 kPa 1,380 kPa Figure 4.7 Permeate Flux vs Time for Applied Pressu re = 1,380 3,450 kPa, and Kaolin Concentrations = 150 mg/l an d Crossflow Velocity = 4.04 cm/s, pH = 9.0 overcome the subsequent particle-particles repulsiv e forces for subsequent foulant layer growth. In the case of application of 3,450 kPa tr ans-membrane pressure, a higher initial permeate flux (4.0 m3/m2/s*10-5) due to higher applied pressure transports brings more clay particles to the membrane. Further, the higher applied pressure also provides the necessary force to overcome the particle-membrane r epulsive force, while providing compressive force that is needed to compress the fo ulant layer. The result is the formation of a thicker cake layer on the membrane s urface leading to a lower permeate flux. During an operating time of 6 hrs, no flux de cline was observed for the membrane

PAGE 93

67 operating at 1,380 kPa producing 1.5 m3/m2/s*10-5. Visual inspection of the used membrane after the experimental runs showed no depo sition occurring on the membrane that was subjected to 1,380 kPa transmembrane press ure. Under the tested conditions (pH = 6.7), kaolin part icles have negative ZP values (Figure 4.1). Previous studies done on LFC 1 membra ne (Vrijenhoek et al., 2001) indicate that the membrane has negative ZP at the t ested conditions. The drag force exerted by the permeate on Kaolin particles at 1,38 0 kPa may not be adequate enough to overcome the electrical double layer repulsion betw een membrane – particle and particle – particle. This could be the reason for no fouling occurring at 1,380 kPa. Flux decline is important from an operational poin t of view. These results show that when the feed water contains fouling clay part icles, the advantage of operating a membrane at a higher pressure to obtain higher perm eate flux is lost due to higher flux decline. 4.4.4.2 Particle Concentration The effect of kaolin clay particles on the fouling rate is shown in Figure 4.8. It appears that the permeate flux decline is a functio n of kaolin clay concentration with the greater flux decline occurring at higher feed conce ntrations. The large decline of permeate flux at a higher feed particle concentrati on is the result of an increased particle transfer rate. As the convective particle transport towards the membrane increases, the overall rate of clay deposition on the membrane inc reases. This results in higher resistance to water flow through the membrane.

PAGE 94

68 y = -0.1381x + 3.8022, R 2 = 0.9822 y = -0.2036x + 3.7965, R 2 = 0.9771 y = -0.2244x + 3.6104, R 2 = 0.981 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0123456 Time (Hrs)Per. flux (m 3 /m 2 /s*10 5 ) 50 mg/l 150mg/l 250mg/l Figure 4.8 Permeate Flux vs Time for Kaolin Concent ration. Transmembrane Pressure = 3,450 kPa and Crossflow Velocity = 4.04 cm/s, pH = 6.4 6.8 4.4.4.3 Crossflow Velocity The effect of crossflow velocity on kaolin fouling is presented on Figure 4.9 (ttest results are given in Appendix A table A.5). Th is shows that for transmembrane pressure of 1,380 kPa, and kaolin concentration of 150 mg/l, decreasing the crossflow velocity from 4.04 to 1.62 cm/s did not influence t he permeate fouling. In order to establish the flow regime, Reynolds numbers were ca lculated and found to be 80.8 and 32.4 for the velocities of 4.04 and 1.62 cm/s, resp ectively. These values fall within the

PAGE 95

69 laminar flow regime for a rectangular channel. Var ying the crossflow velocity within the practical limits of the unit of the runs, which dis played permeate flux decline with time, failed to affect the permeate flux. These observati ons were same even for the runs that were carried for Bentonite clay containing feed wat er. These results show that increasing the crossflow velocity does not affect the fouling layer and could not make the permeate flux improve. y = 0.0009x + 1.5218 R2 = 0.0092 y = 0.0009x + 1.4629 R2 = 0.00930.0 0.4 0.8 1.2 1.6 2.0 0123456 Time (Hrs)Per. flux (m3/m2/s*105) 150,mg/l,1.62cm/s,pH=6.7 150mg/l,4.04cm/s,pH=6.7 Figure 4.9 Permeate Flux vs Time, Applied Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l

PAGE 96

70 4.4.4.4 Occurrence of Critical Flux The critical flux is defined as the lowest flux tha t creates a fouling layer on the membrane surface. When the membrane was operated at a pressure of 1,380 kPa and 4.04 and 26.26 cm/s crossflow velocities for a 150 mg/l of kaolin concentration, no flux decline was observed (Figure 4.9) an indication of non-occurrence of fouling. The flux decline is due to cake layer formation. The permeat e velocity transports particles onto the membrane, whereas, the crossflow velocity generates shear forces that could remove particles from the cake layer. Therefore, in orde r to verify the role of the crossflow velocity control of flux decline, an experiment was carried out at a reduced crossflow velocity of 1.6 cm/s. This experiment did not show any increased effect of fouling. The lack of any flux decline, observed at these testing conditions implies that these conditions were at or below the “critical flux” conditions for the LFC1 membrane. 4.4.4.5 Mass Deposited vs Flux Decline Relationship To investigate the effect of permeate flux decline from the mass of clay deposited on the membrane, clay deposit was carefully removed from the membrane surface with the help of a spatula, dried at 105oC temperature for 24 hrs and weighed. Figure 4.10 shows the relationship between the mass deposited v s flux decline relationship for several experimental runs with bentonite clays. Regression analysis shows reasonably good fit (R2= 0.8692) between the two parameters. The result co nfirms that the flux decline is attributed to the build up of a particle layer on t he membrane surface.

PAGE 97

71 4.4.5 Statistical Model From the data obtained from bentonite fouling runs carried out for three factor three level experimental design, an empirical model was derived. The general linear model function of the Statistical Package for the S ocial Sciences (SPSS) software (SPSS Inc., Chicago, Illinois) was used to determine this model (Output given in Appendix D) which predict flux decline at an end of 8-hour runs for the unit for a given transmembrane pressure, particle concentration and crossflow velo city. The expression for the flux decline (FD) is of the form; FD = -32.328 + 2.143LnP + 0.456LnC + 0.006LnV y = 9.4149x 2.1031 R2 = 0.86920.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 00.10.20.30.40.50.6 Mass deposited (mg)Flux decline (m3/m2/s*105)P value (intercept) = 1.6x10-3P value (X variable) = 2.9x10-3 Figure 4.10 Flux Decline vs Mass of Cake Deposited for Bentonite and for a LFC 1 Membrane

PAGE 98

72 The applicable units for flux decline (FD), pressur e (P), concentration (C) and crossflow velocity (U) are m3/m2/s*105, Pa, mg/l and cm/s, respectively. The relationship between the experimental values a nd the model values provides a good fit with a R2= 0.9877 and is given in figure 4.11. 4.5 Membrane Scaling Runs In this phase of the research, fouling of membranes by precipitation was studied. Calcium, sulfate and carbonate ions are present in abundance in most natural water sources that serve as feedwaters to many water trea tment units, including reverse osmosis. Calcium carbonate and calcium sulfate are the most scales found in water treatment units. Therefore, these scaling experimen ts were conducted using calcium y = 0.9951x + 0.0058 R2 = 0.9877 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.00.51.01.52.02.53.0 Exp. flux decline (m3/m2/s*105)Model predicted flux decline (m3/m2/s*105) Figure 4.11 Model Predicted Flux Decline vs Experimental Flux Decline Values for Bentonite and for a LFC 1 Membra nes

PAGE 99

73 carbonate. Due to calcium carbonate’s poor solubili ty in water (Ksp= 3.31 x 10-9 M2), combined CaCl2 and Na2CO3 solutions were used to generate a CaCO3 scale. Further, CaCl2 and Na2CO3 salts could be easily obtained from suppliers in h igh purity form (reagents grade) and also CaCl2 and Na2CO3 are individually soluble in water. In these experiments, equimolar concentration (0.00 05 M, 0.0015M) of CaCl2 and Na2CO3 salt solutions were to formulate different saturat ion levels. Supersaturated solutions were used to reflect the concentration, w hich is found on the retentate side of the reverse osmosis unit. Transmembrane pressure a nd crossflow velocities were varied from 1,380 kPa – 3,450 kPa and 4.04 cm/s – 26.26 c m/s, respectively. Additionally, MgCl2 salt was also used to prepare an osmotic feed solu tion that would provide the same osmotic pressure as in the CaCO3 solution. 4.5.1 Preparation of Scaling Solution The CO3 -2 ion concentration in a solution depends on the sol ution pH values. Before starting an experimental run, super-saturati on solutions were made with CaCl2 and Na2CO3 by dissolving a weighed salt quantity in distilled water. Detailed calculations are given in Appendix A table A.1.

PAGE 100

74 4.5.2 Studies with 0.0005 M CaCl2 and Na2CO3 In order to induce CaCO3 scaling, experimental runs with equimolar solution s of CaCl2 and Na2CO3 were carried out. Figure 4.12 gives the typical pe rmeate flux vs time relationship. The results show that by the fifth hour, there was a permeate flux decrease of around 43 %. Beyond 5 hrs, this trend tends to leve l off. The permeate flux decline is the result of additional resistance to flow. The resist ance to permeate flow is the result of membrane resistance, scale layer resistance and the change in osmotic pressure caused by the concentration polarization effect. Figure 4.12 raises some interesting questions. Fir st, was the flux decline function of the 0.0005 M equimolar solutions forming calcium carbonate that resulted in y = 0.281x2 1.0037x + 3.6325 R2 = 0.9234y = 0.0252x2 0.3419x + 3.2703 R2 = 0.99050.0 1.0 2.0 3.0 4.0 0123456 X = Time (Hrs)Y = Per. Flux (m3/m2/s*105) Figure 4.12 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 m/s pH = 9.0

PAGE 101

75 concentration polarization effect or CaCO3 precipitation on the membrane surface or both. The second issue was whether changes in opera ting parameters such as feed flow rate, transmembrane pressure or feed solution pH co uld have caused the flux decline. To confirm whether the salt precipitate contributed to the flux decline, additional studies were carried out using equimolar solution o f MgCl2 plus Na2CO3. 4.5.3 Studies with MgCl2 and Na2CO3 A sample of membrane that was used in the section 4 .5.2 experiment was subjected to a feed solution containing a mixture o f 0.0005 M MgCl2 and 0.0005 M Na2CO3 salt solutions. The 0.0005 M MgCl2 and 0.0005 M Na2CO3 solution provided the same osmotic pressure value as the mixture containi ng 0.0005 M CaCl2 and 0.0005 M Na2CO3. However, solubility of MgCO3 (Ksp # = 2.88 x 10-5 M2), was much higher than the CaCO3 (Ksp # = 3.31 x 10-9 M2 ) values. A plot of Permeate flux vs time curve is presented in Figure 4.13. Had the flux decline been purely due to concentration polarization, the curve for 0.0005 M CaCl2 plus Na2CO3 should have closely resembled that of 0.0005 M MgCl2 and 0.0005 M curve, which wa s not the case, indicating that scaling had some effect. Comparison of the slopes b etween the “osmotic solution” and the scaling solution revealed that flux from the sc aling solution decreased more rapidly than in the osmotic solution. This could also be du e to CaCO3 deposition. Also, in the combined CaCl2 plus Na2CO3 feed run, within few minutes of mixing the CaCl2 and Na2CO3 salt solutions, precipitation occurred.

PAGE 102

76 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0123456 Time (Hrs)Per. Flux (m3/m2/s*105)MgCl 2 & Na 2 CO 3 CaCl 2 & Na 2 CO 3 Figure 4.13 Permeate Flux vs Time for CaCl2 Plus Na2CO3 and MgCl2 Plus Na2CO3. Transmembrane Pressure = 3,450 kPa, Crossflow Vel ocity = 4.04 cm/s and pH= 9.0 4.5.4 CaCO3 Scaled Membranes with Acetic Acid The permeate flux vs time relationship for the CaCl2 plus Na2CO3 feed solution, when compared with MgCl2 plus Na2CO3 which gives the same osmotic pressure, showed that the scale layer would have contributed to this flux decline. In order to further verify this, after a 6.0 hr run with 0.0005 M CaCl2 and 0.0005 M Na2CO3, the permeate flux of the membrane was tested with a feed solution co nsisting of 0.01 M acetic acid (CH3COOH) followed by D.I. water run. The result show n in figure 4.14 clearly show that as the pH of the solution is lowered to around 4.0, the permeate flux begins to

PAGE 103

77 improve rapidly. Finally, using D.I. water as feed solution, the permeate flux stabilized around 3.58 m3/m2/s*105 which is about 96 % of the original permeate flux at the beginning of the scaling experiments. The use of a stronger acid such as HCl did not show further improvement on permeate flux. 0.0 1.0 2.0 3.0 4.0 0246810 Time (Hrs)Per. Flux (m3/m2/s*105) CH3COOH D.I. Water Figure 4.14 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 m/s, CaCO3 Concentration = 0.0005 M 4.5.5 Effect of Salt Concentration Experiments were conducted to determine the effects of solution concentrations on the rate of scale formation at a fixed crossflow velocity of 4.04 cm/s and trans-

PAGE 104

78 membrane pressure of 3,450 kPa. Figure 4.15 clearly shows that higher the concentration the more rapid the flux declines, but it appears to follow an exponential rate. The reason for this could be explained as follows; when salt s olution is subjected to RO membrane filtration, with time, due to permeate separation, the concentration of salts near the membrane surface increases. When the salt concentra tion exceeds supersaturation level, salt precipitation takes place on the membrane surf ace creating resistance to permeate flow. However, the decrease of permeate flux with t ime also reduces the transport of salt towards the membrane and, hence causes a decrease i n the rate of crystallization. To get an idea of the effect of supersaturation (f or calculations see Appendix A table A.3), batch experiments in a beaker were carr ied out with 0.0005 and 0.0015 M solution of CaCO3. In the case of 0.0015 M experiments, as soon as t he salt solutions of CaCl2 and Na2CO3 were mixed, the contents immediately turned into a cloudy solution. However, with 0.0005 M solution, the content took a bout 15 minutes to turn into a cloudy solution. These observations are an indication of t he formation of a white color CaCO3 precipitate.

PAGE 105

79 0.0 1.0 2.0 3.0 4.0 0123456Time (Hrs)Per. Flux (m3/m2/s*105) 0.0005 M 0.0015 M Figure 4.15 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 m/s, pH = 9.0. CaCO3 Concentration 0.0005 M and 0.0015 M In order to find answers to the second question r aised in section 4.5.2, several experiments, which are described below, were carrie d out by changing the crossflow velocity and transmembrane pressure. 4.5.6 Effects of Crossflow Velocity Increasing the crossflow velocity from 4.04 cm/s to 25.8 cm/s at a constant applied pressure of 3,450 kPa increased the permeat e flux. The results are given in figure

PAGE 106

80 4.16. The Reynolds number ranged between 83 and 53 0, which is within the laminar flow range for a rectangular channel. The larger th e axial velocity, larger the shear rate, thus, causing a decrease in the boundary layer thi ckness, which resulted in a decrease in the salt concentration at the membrane surface. Thi s also makes the conditions at the membrane surface unfavorable for salt precipitation to take place on the membrane. 4.5.7 Effect of Transmembrane Pressure The change in permeate flux was examined holding th e axial velocity constant and varying the applied pressure. These results are shown in figure 4.17. The rejection of 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0123456 Time (Hrs)Per. Flux (m3/m2/s*105) 4.04 cm/s 26.26 cm/s Figure 4.16 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s and 25.8 cm/s, CaCO3 Concentration = 0.0005 M, pH = 9.0

PAGE 107

81 salts at higher initial flux would lead to concentr ation polarization and scaling, resulting in a higher flux decline. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0123456 Time (Hrs)Per. Flux (m3/m2/s*105) 3,450 kPa 2,760 kPa 2,070 kPa 1,380 kPa Figure 4.17 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 1,380 3,450 kPa, Crossflow Velocity = 4.04 cm/s, CaCO3 Con. = 0.0005 M, pH = 9.0 4.5.8 CaCO3 Scaling Runs at Different pH Values For CaCO3 scaling to occur, the solubility product {Ca+2}{CO3 -2}has to exceed the Ksp value for CaCO3. Carbonate species concentration in the feed solut ion is pH dependent (see figure 4.18 for pH vs pC). In order to verify the above, experiments were carried out by adding acetic acid to lower the pH v alues to 4.0 and 5.5 respectively. The results are given in figure 4.18. At higher pH (9.0), in the presence of CO3 2-, flux decline

PAGE 108

82 due to CaCO3 is observable, whereas, at lower pH (4.0 and 5.5), flux decline due to CaCO3 scaling is not present. -25 -20 -15 -10 -5 0 0246810121416pHpCH2CO3HCO3 -CO3 2Figure 4.18 pH vs pC for a Carbonate System (After, Sawyer et al., 1994)

PAGE 109

83 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0123456Time (Hrs)Per. Flux (m3/m2/s*105)pH = 5.5 pH = 4.0 pH = 9.0 Figure 4.19 Permeate Flux vs Time for CaCO3. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Nomina l CaCO3 Concentration = 0.0005 M, pH = 4.0, 5.5 and 9.0 4.5.9 Effect on Permeate Quality Another consequence of fouling is changes in permea te quality, which is measured by salt rejection or salt passage through the membrane. In Figure 4.20, rejection as a function of time is shown for variou s experimental conditions. It may be observed that as the transmembrane pressure increas es, the salt rejection decreases. This change may be due to an increase in concentration p olarization and salt leakage through the membrane.

PAGE 110

84 80 84 88 92 96 100 0246 Time (Hrs.) Rejection (%) 1,380 kPa, 0.0005 M CaCO3 2,070 kPa, 0.0005 CaCO3 2,760 kPa, 0.0005 M CaCO3 3,450 kPa, 0.0005 M CaCO3 3,450 kPa, 0.0015 M CaCO3 Figure 4.20 Rejection vs Time for CaCO3. Transmembrane Pressure = 1,380 3,450 kPa, Crossflow Velocity = 4.04 cm/s, CaCO3 Concentration = 0.0005 M and 0.0015 M, pH = 9.0 4.6 Kaolin and CaCO3 Experiments Having investigated the individual effects of kaol in fouling and CaCO3 scaling on a LFC 1 membrane, separately, focus shifted to the interaction effects of clay fouling and CaCO3 scaling. These two mechanisms may take place concu rrently on the membrane surface.

PAGE 111

85 4.6.1 Permeate Flux vs Time Experiments were carried out using kaolin concentra tions of 50, 150 and 250 mg/l and CaCO3 concentrations of 0.0005 M and 0.0015 M at a pH of 9.0. Transmembrane pressure was varied from 1,380 – 3,450 kPa while ho lding the crossflow velocity constant at 4.04 cm/s. Figure 4.21 shows the results of these experiments For comparison purposes, the results obtained for 150 mg/l kaolin in distilled w ater and CaCO3 scaling in D.I. water results were also included in this figure. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0246 Time (Hrs)Permeate flux (m3/m2/s*105) Kaolin only CaCO3 Kaolin +0.0005 MCaCO3 Kaolin +0.0015 MCaCO3Figure 4.21 Permeate Flux vs Time for Combined CaCO3 and Kaolin. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/ s, Kaolin Concentration = 50 mg/l, CaCO3 Concentration = 0.0005 M and 0.0015 M, pH = 9.0

PAGE 112

86 The combined kaolin and CaCO3 run results show a higher flux decline for the experimental runs. This flux decline is seen to be more pronounced at a higher CaCO3 concentration. The figure 4.22 shows the effect of kaolin particle concentration on the flux decline at a 0.0005 M CaCO3 concentration. As expected, the increase in kaolin concentration tends to shift the curve down. These results were obtained for a crossflow velocity of 4.04 cm/s. The decrease in permeate flu x when the kaolin concentration increased, could be due to the dual effect of forma tion of an increasingly thick fouling layer comprising of kaolin particles and a concentr ation polarization layer. Further, an increase in the kaolin layer would have triggered a spin off effect by increasing the concentration polarization layer thickness. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0246 Time (Hrs)Permeate flux (m3/m2/s*105) CaCO3 50 mg/lKaolin 150 mg/lKaolin 250 mg/lKaolin Figure 4.22 Permeate Flux vs Time for Combined CaCO3 and Kaolin. Transmembrane Pressure = 3,450 kPa, Crossflow Veloc ity = 4.04 cm/s, Kaolin Concentration = 50 – 250 mg/l, CaCO3 Concentration = 0.0005 M, pH = 9.0

PAGE 113

87 4.6.2 Permeate Quality with Time Figure 4.23 shows the relationship between the reje ction with time. With time, the combined kaolin plus salt run displays a lower reje ction (measured in terms of conductivity) than the CaCO3 runs. The decrease in permeate quality could be at tributed to the increase in salt concentration on the membra ne surface, which may be attributed to a decrease in the salt back diffusion rate. The red uction in this rate could be a result of compaction of kaolin clay and reduction of cake por osity. 80 84 88 92 96 100 0123456 Time (Hrs.)Rejection (%) 3,450 kPa,0.0005 MCaCO3 3,450 kPa,0.0005 MCaCO3,150 mg/l Kaolin Figure 4.23 Observed rejection vs Time for CaCO3 and Combined CaCO3 and Kaolin. Transmembrane Pressure = 3,450 kPa, Crossflow Veloc ity = 4.04 cm/s, Kaolin Concentration = 150 mg/l, CaCO3 Concentration = 0.0005 M, pH = 9.0

PAGE 114

88 4.6.3 Reversibility of the Fouling Layer According to the Resistance in Series Model, the pe rmeate flux reduction is the result of the increased resistance from the deposit ed fouling layer. In the combined fouling studies, the fouling layer consists of a ka olin particle layer, plus, a CaCO3 scaling layer on the membrane surface. However, the individ ual contribution of kaolin layer and CaCO3 scale deposit towards the resistance cannot be fou nd directly. By lowering the pH of the feed water to a value of around 4.0, the CaC O3 scale on the LFC 1 membrane can be dissolved without causing any permanent damage t o the membrane surface. Hence, all fouled membranes were subjected to additional runs with D.I. (with pH = 4.00) water, immediately, at the end of each fouled study, to in vestigate the permeate volumetric flux recovery. Finally, as the last phase of the run, D. I. water runs were carried out on the same membrane to determine the pure water permeabil ity coefficient. Figure 4.24 shows the Kw values for a kaolin concen tration of 150 mg/l and CaCO3 concentration of 0.0005 M for different transmembr ane pressure. The results show that for the same CaCO3 and kaolin concentration, increasing the transmembr ane pressure could lead to a pure water permeate loss o f 1-35 %. This loss could be due to the change in the fouling layer’s characteristics (poss ibly a loss in porosity) under the increasing transmembrane pressure. Visual inspectio n of the membranes, except for the membrane that was subjected to a pressure of 1,380 kPa, displayed a cake layer on the membrane surface.

PAGE 115

89 94 78 65 99 0 20 40 60 80 100 1,3802,0702,7603,450 Transmembrane Pressure (kPa)Final Kw (as a % of initial Kw) Figure 4.24 Final Kw (as % of initial Kw) For Kaolin = 150 mg/l, CaCO3 = 0.0005 M and Crossflow Velocity = 4.04 cm/s

PAGE 116

90 Chapter 5 Modeling Figure 5.1 Particle Deposition Mechanism on a Membr ane Resistance in Series Model for reverse osmosis may expressed as )1.5( ) ( + DP D =f mR R P t V In the above equations, V(t) = Permeate volumetric flux (m3/m2/s), P = Transmembrane pressure (Pa), = Osmotic pressure (Pa), Rm = Membrane resistance (Pa.S/m), Rf = Fouling layer resistance (Pa.S/m) It is reasonable to neglect the osmotic pressure () term from the above equation for dilute salt solutions and write; )2.5( ) ( 1 D + D @ P R P R t Vf m

PAGE 117

91 The net rate of foulant particles deposited on the membrane surface can be described as the net effect of integration of parti cles brought to the membrane surface less foulant back diffused. For the combined clay-CaCO3 experimental runs, tested crossflow velocity range did not have any effects on the perm eate volumetric flux. Hence, the shearing effects of feed flow on the foulant layer removal was not included in developing the above relationship. Further, in the presence of CaCO3, the clay particle surface charge seems to be decreasing and finally reaching a value of zero. Therefore particle surface charge term was also neglected in this model. Consi dering all the above factors, the deposition of foulants on the membrane surface can be written as (Chen et al. 2004). )3.5( ) ( ) ( = dy dCf D C t V dt t dm Where m(t) = Foulant mass flux (mg/m2), V(t) = Permeate volumetric flux (m3/m2/s), C =Foulant concentration in the feed (mg/l), Cf = Fo ulant concentration in cake layer (mg/l), D = foulant diffusion coefficien t (m2/s) In the transient stage, steady state has not been r eached and, therefore, equation (5.3) may be approximated as follows. )4.5( ) ( ) ( C t V dt t dm The total mass of foulants deposited over a period of time is: )5.5( ) ( ) ( ) (0 0= = Cdt t V dt dt t dm t Mt t Based on the filtration theory, the fouling layer r esistance (Rf) is linearly proportional to the foulant mass deposited and coul d be expressed as:

PAGE 118

92 )6.5( ) ( = t M Rfa where is the cake layer specific resistance From (5.5) and (5.6). )7.5( ) (0=Cdt t V Rt fa )8.5( ) (0=dt t V C Rt fa By substituting the Rf value in eq. (5.2) )9.5( ) ( ) ( 10D + D =P dt t V C P R t Vt ma According to equation (5.9), the plot of ) ( 1 t V vstdt t V0) ( should provide a straight line from which the slope term (C) can be found. Figure 5.2 shows the results of ) ( 1 t V vstdt t V0) ( for the kaolin concentration of 50 mg/l and CaCO3 of 0.0015 M. The values give a good linear regression fit between ) ( 1 t V and tdt t V0) (with a R2=0.9952,

PAGE 119

93 y = 0.6244x + 0.2786 R2 = 0.9952 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.000.200.400.60 V(t)*t (m)1/ V (t) (s/m*10-5) Figure 5.2 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015M CaCO3 and 50 mg/l of Kaolin The same procedure was repeated for the other data sets as well (The results are given in Appendix G). The summary of these analysis is given in table 5.1. It is evident from the table that for all the cases with the exception of one, that a good linear fit exists between ) ( 1 t V and tdt t V0) (, with R2 varying between 0.9062 and 0.9952. This indicates that this model was very pre dictable for the used feed solutions. Although this model did not consider the crossflow velocity term and the particle surface charge in its derivation, it can be applied to any particulate and precipitation fouling cases with considerable accuracy. Alpha (), which is the specific resistance of the deposit layer, was plotted (Fig. 5.3) against the Transmembrane pressure (P). Results show that the value of alpha

PAGE 120

94 Table 5.1 Summary Data from Model Analysis Pressure (psi) Pressure (N/m2) CaCO3 Con. (M) Kaolin Con. (mg/l) Kaolin Con. (kg/m3) Slope (Kg/s.m3*105) Intercept (s/m*105) R2 C/p (s/m2) C {Kg/(s*m3)} 5003.45E+060.0005500.050.47700.28330.9865477001.643 3E+11 5003.45E+060.0015500.050.62440.27860.9952624402.151 1E+11 5003.45E+060.00051500.150.52960.29900.9860529601.82 45E+11 5003.45E+060.00151500.150.74230.33990.9062742302.55 72E+11 5003.45E+060.00052500.250.81910.23000.9288819102.82 18E+11 5003.45E+060.00152500.251.65450.21570.96511654505.6 998E+11 4002.76E+060.00051500.150.35340.33830.9811353409.73 97E+10 3002.07E+060.00051500.150.42420.45310.9665424208.76 82E+10 2001.38E+060.00051500.150.34330.70240.7557343304.73 07E+10

PAGE 121

95 increases exponentially as the transmemembrane pres sure increases. This trend may be the result of the higher applied pressure compressi ng the cake layer, which overcame the particle-particle interaction forces. This leads t o a less permeable cake layer with a higher resistance. Further, alpha () showed a strong dependence on the Zeta potential of the particles with resistance decreasing as zeta po tential increased. y = 1.4381e6E-07xR2 = 0.941 0 2 4 6 8 10 12 14 0.E+001.E+062.E+063.E+064.E+06 P (N/m2) (1/S*10-11) Figure 5.3 Specific Resistance of the Cake Layer () vs Transmembrane Pressure (P) for 0.0005 M CaCO3 and 150 mg/l of Kaolin Finally, experimentally obtained values were compar ed with the model results and the final results are given in Fig. 5.4

PAGE 122

96 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0123456 Time (Hrs)Permeate flux (m3/m2/s*105) ExperimentalModel Figure 5.4 Experimental and Model Results, Transmembrane Pressure = 3,450 kPa, CaCO3 Concentration = 0.0015 M, Kaolin Concentration = 250 mg/l

PAGE 123

97 Chapter 6 Conclusion For clay fouling experiments, the effect of operati ng parameters on permeate flux due to clay fouling on a LFC 1 membrane, installed in a laboratory test cell was discussed. No steady state was reached during the 6 .0 hr experiments. The permeate volumetric flux vs time curve for kaolin runs showe d a linear relationship, which indicated that the foulant cake growth was linear i n time. The rapid flux decline observed for the bentonite clay, even at a lower suspension concentration than kaolin, could be due to its swelling properties. Mass of the clay deposi ted on the membrane showed a linear relationship with permeate flux decline. For the kaolin clay runs, when the membrane was ope rated at a pressure of 200 psi and 4.04 and 29.9 cm/s cross flow velocities fo r a 150 mg/l of kaolin concentration, no flux decline was observed. The lack of any flux decline observed at these testing conditions implies that these operating conditions were at or below the “critical flux” conditions for the LFC1 membrane. Critical flux is defined as the flux level at which no noticeable fouling occurs and is very often discuss ed in MF and UF related research but not in R.O. research. These results also show that extensive pretreatment for the removal of suspended particles may not be required if the c lay content of the feed water is less

PAGE 124

98 than 150 mg/l. However, further verification of thi s is needed by conducting experiments at pilot and full scale levels for a longer duratio ns. In the combined kaolin and CaCO3 runs, with time, there was an increase in the permeate flux decline with an increase in both salt and kaolin concentrations. This is a direct result of fouling layer and concentration po larization taking place on the membrane surface. Simultaneously, permeate quality in these runs, relative to the pure CaCO3 runs, also decreased with increasing salt and kaolin conc entrations. The decrease in permeate quality in the presence of kaolin in feed solution is attributed to formation of a foulant layer and its subsequent hindrance to back diffusio n rate. The simple model used in this study by combining th e resistance in series model together with mass transport model that ignored the shearing action of the feed flow on fouling layer removal seems to well represent the t ransient stage permeate flux for the combined kaolin and CaCO3 system. Although not tested, the model could be ca librated for any data set of a feed solution that contains s cale forming salts and other types of clays such as bentonite or any other foulant matter Further, to use this method, no prior knowledge of particle size, particle surface charge or feed water chemical characteristics is required. Finally, this method could be used as a tool in a R.O. membrane water treatment plant fouling predictor with very little refinement, thus, saving money and time. Due to its complexity, microbiological fouling was not taken into consideration in this study. However, in future work, model will hav e to be calibrated in-order to take into account the biological fouling as well.

PAGE 125

99 References 1. Altena, F. W. and Belfort, G. (1984), “Lateral migr ation of spherical particles in porous flow channels: Application to membrane filtr ation,” Chem. Eng. Sci ., vol. 39, no. 2, pp. 343-355. 2. Altena, F. W., Belfort, G., Ortis, J., Fiessinger, F., Rovel, J. M., Nicoletti, J. (1983), “Particle motion in a laminar slit flow: A fundamen tal fouling study,” Desalination vol. 47, no. 1-3, pp. 221-232. 3. Baker, R.W. (2000). Membrane Technology and Applications New York: McGrawHill. 4. Barger, M. and Carnahan, R. P. (1991), “Fouling pre diction in reverse osmosis processes”, Desalination vol. 83, no. 1-3, pp. 3-33. 5. Belfort G., and Marx, B. (1979), “ Artificial parti culate fouling of hyperfiltration membranes – II analysis and protection from fouling Desalination vol. 28, no.1, pp. 13 -30. 6. Bhattacharyya, S. & Williams, M. (1992). Membrane Handbook Introduction In Winston, W. & Sirkar, K. (Eds.). New York, NY,Van Nostrand Reinhold. 7. Blatt, W. F., Dravid, A., Michaels, A. S., Nelson, L. (1970). Solute polarization and cake formation on membrane ultrafiltration: causes, consequence, and control techniques. In: Flinn, J.E. (Ed.), Membrane Science and Technology Plenum Press, New York. 8. Boerlage, F. E., Kennedy, M. D., Aniye, M. P., Abogrean, E. M., Galjaard, G., Schippers, J. C. (1998), “Monitoring particulate fo uling in membrane systems,” Desalination vol. 118, no. (1-3), pp. 131-142. 9. Bowen, W.R. and Jenner, F. (1995), “Theoretical des criptions of membrane filtration of colloids and fine particles: an assessment and r eview,” Adv. Coll. Interf. Sci., vol. 56, pp. 141-200. 10. Brian, P.L.T. (1966). Mass transport in reverse osm osis. In U. Merten (Ed.), Desalination by Reverse Osmosis MIT Press, Cambridge, MA.

PAGE 126

100 11. Brunelle, M.T. (1980), “Colloidal fouling of Revers e Osmosis membranes,” Desalination vol. 32 pp.127-131. 12. Byrne, W. (1995). Reverse Osmosis Practical Guide for Industrial User s. Littleton, CO: Tall Oaks Publishing. 13. Chen, K. L., Song, L., Ong, S. L., Ng, W.J. (2004), “The development of membrane fouling in full-scale RO processes,” J. Membr. Sci ., vol. 232, no. 1-2, pp. 63-72. 14. Childress, A. E. and Elimelech, M. (1996), “Effect of solution chemistry on the surface charge of polymeric reverse osmosis and nan ofiltration membranes,” J. Membr. Sci ., vol. 119, no. 2, pp.253-268. 15. Cleaver, J. W. and Yates, B. (1973), “Mechanism of detachment of colloid particles from a flat substrate in turbulent flow,” J. Coll. Interf. Sci ., vol. 44, no. 3, pp. 464473. 16. Cleaver, J. W. and Yates, B. (1976), “The effect of re-entrainment on particle deposition,” Chem. Eng. Sci ., vol. 31, no. 2, pp. 147-151. 17. Cohen, R.D. and Probsetin, R.F. (1986), “Colloidal fouling of reverse osmosis membranes,” J. Coll. Interf. Sci. vol. 114, no. 1, pp. 194-207. 18. Dydo, P., Turek, M., Ciba, C. (2003), “Scaling anal ysis of nanofiltration systems fed with saturated calcium sulfate solutions in the pre sence of carbonate ions,” Desalination vol. 159, no. 3, pp. 245-251. 19. Endo, Y. and Alonso, M. (2001), ” Physical meaning of specific cake resistance and effects of cake properties in compressible cake fil tration”, Filtr. Sepn, vol. 38, no. 7, pp. 43-46. 20. Espinasse, B., Bacchin, P., Aimar, P. (2002), “On a n experimental method to measure critical flux in ultrafiltration”, Desalination vol. 146, no. 1-3, pp. 91-96. 21. Fane, A. G. (1984), “Ultrafiltration of dispersions ,” J. Membr. Sci ., vol. 20, no. 3, pp. 249-259. 22. Faust, S.D. & Aly, O.M. (1987). Adsorption processes for water treatment Stoneham, MA: Butterworth Publishing. 23. Gabrielli, C., Maurin, G., Poindessous, G., Rosset, R. (1999), Nucleation and growth of calcium carbonate by an electrochemical scaling process, J. Cryst. Growth vol. 200, no. 1-2, pp. 236-250.

PAGE 127

101 24. Gerard, R., Hachisuka, H., Hirose, M. (1998), “New membrane developments expanding the horizon for the application of revers e osmosis technology”, Desalination vol. 119, no.1-3, pp. 47-55. 25. Gutman, R. G. (1977), “The design of membrane separ ation plantPart 2 Fouling of RO membrane”, Chem. Engineer July 1977, pp. 521-523. 26. Hamrouni, B. and Dhahbi, M. (2001), “Thermodynamics description of saline watersprediction of scaling limits in desalination proces ses,” Desalination vol. 137, no. 13, pp. 275-284. 27. Hasson, D., Bramson, D., Relis, B. L., Semiat, R. ( 1996), “Influence of the flow system on the inhibitory action of CaCO3 scale prevention additives,” Desalination vol. 108, no. 1-3, pp. 67-79. 28. Hasson, D. (1999). Progress in Precipitation foulin g research – A Review. In Bott T.R. (Eds.), Understanding heat exchanger fouling and its mitiga tion Belgell House, New York, pp. 67-90. 29. Vrijen Hoek, E. M. V., Kim, A. S., Elimelech, M. (2 002), “Influence of crossflow membrane filter geometry and shear rate on colloida l fouling in reverse osmosis and nanofiltration separation”, Environ. Engin. Scie. vol. 19, no. 6, pp. 357-372. 30. http://www.cia.gov/cia/publications/factbook/rankor der/2119rank.html. Accessed 11/03/2005. 31. http://www.lifetoday.org/partner. Accessed 11/02/20 05. 32. http://www.solarsystem.nasa.gov/Accessed 11/22/2005 33. Kimura, S and Nakao, S. (1975), “Fouling of cellulo se acetate tubular reverse osmosis modules treating the industrial water in To kyo”, Desalination vol.17, no. 3, pp. 267-288. 34. Lee, S. and Lee, C.H. (2000), “Effect of operating conditions on CaSO4 scale formation mechanism in NF for water softening,” Wat. Res. vol. 34, no. 15, pp. 3854-3866. 35. Lee, S., Kim, J., Lee, C. H.(1999), “Analysis of Ca SO4 scale formation mechanism in various NF modules, J. Membr. Sci ., 163, no. 1, pp. 63-74. 36. Lu, W.M. and Ju, S.C. (1989), “Selective particle deposition in crossflow filtration,” Sep. Sci. Technology ., vol. 24, no. 78, pp. 517-540.

PAGE 128

102 37. Madsen R.E (1977), Hyperfiltration and ultrafiltrat ion in plate and frame systems, Elsevier, New York. 38. Nancollas, G.H. and Reddy, M.M. (1971), “The crysta llization of calcium carbonate. II Calcium growth mechanism,” J. Collo. Interface Sci. vol. 37, no. 4, pp. 824-830. 39. Ng, H.Y. and Elimelech. M. (2004), “Influence of co lloidal fouling on rejection of trace organic contaminants by reverse osmosis,” J. Membr. Sci ., vol. 244, no. 1-2, pp. 215-226. 40. Okazaki, M. and Kimura, S. (1984), “Scale formation on reverse osmosis membranes,” J. Chemi. Engin. Japan vol. 17, no.2, pp. 145-151. 41. Potts, D.E., Ahlert, R.C., Wang, S.S. (1981). “A cr itical review of fouling or reverse osmosis membranes,” Desalination vol.36, pp. 235-264. 42. Probstein, R. F., Chan, K. K., Cohen, R. and Rubens tein, I. (1981) “Model and preliminary experiments of membrane fouling in reve rse osmosis”, Synthetic membranes and their applications (Edited by A. Turb ak), ACS symposium series, Am. Chem. Soc., Washington D.C. 43. Rengasamy, P. & Oades, J.M. (1977), “Interaction of monomeric and polymeric species of metal ions with clay surfaces. 11 change s in surface properties of clays after addition of iron (111),” Aust. J. Soil. Res ., vol.15, no.221, pp.235-242. 44. Rooklidge, S.J., Burns, P.C., Ketchum, L.H. (2002), “Clay removal in basaltic and limestone horizontal roughing filters,” Adv. Environ. Res. vol. 7, no. 1, pp. 231-237. 45. Rouquerol, F., Rouquerol, J., & Sing, K. (1999). Adsorption by Clays, Pillared Layer Structures and Zeolites. Adsorption by Powders and Porous Solids, Principles, Methodology and Applications London: Academic Press. 46. Sagle, A. and Freeman, B. (2005), “Fundamentals of membranes for water treatment”, http://www.twdb.state.tx.us/Desalinatio n.pdf, Accessed 11/25/2005. 47. Sawyer, C. N., MaCarty, P.L., Parkin, G. F. (Eds.). (1994). Chemistry for Environmental Engineers Singapore, McGraw-Hill. 48. Schfer, A.I, Fane, A.G. ,Waite, T.D. (1998), “Nano filtration of natural organic matter: removal, fouling and the influence of multi valent ions”, Desalination vol.118, no. 1-3, pp.109-122. 49. Schippers, J. C., Hanemaayer, J. H., Smolders, C. A ., Kostense, A. (1981), “Predicting flux decline of reverse osmosis membran es”, Desalination vol. 38, pp. 339-348.

PAGE 129

103 50. Schwinge, J., Neal, P. R., Wiley, D. E., Fane, A. G (2002), “Estimation of foulant deposition across the leaf of a spiral-wound module ”, Desalination vol. 146, no. 1-3, pp. 203-208. 51. Seidel, A. and Elimelech, M. (2002), “Coupling betw een chemical and physical interactions in natural organic matter (NOM) foulin g of nanofiltration membranes: implications for fouling control,” J. Membrane Sci vol. 203, no. 1-2, pp. 245-255. 52. Segre, G and Silberberg, A. (1962), “Behavior of ma croscopic rigid spheres in poiseuille flow. Part 2. Experimental results and i nterpretation”, J. Fluid Mech., vol. 14, pp.136-156. 53. Shaw, D. J (1970). Introduction to Colloid and Surface Chemistry. London: Butterworth and Co. 54. Sheikholeslami, R. and Ng, M. (2001), “Calcium sulf ate precipitation in the presence of nondominant calcium carbonate: Thermodynamics an d Kinetics,” Ind. Eng. Chem. Res ., vol. 40, no. 16, pp. 3570-3578. 55. Shiklomanov, I. A, (1999), “World water resources a nd their use”, http://espejo.unesco.org.uy, Accessed 09/05/06. 56. Song, L. (1998), “Flux decline in crossflow microfi ltration and ultrafiltration: mechanisms and modeling of membrane fouling”, J. Membr. Sci ., vol. 139, no. 2, pp.183-200. 57. Stumm, W. and Morgan, J. J. (1981). Aquatic Chemistry, An Introduction Emphasizing Chemical Equilibria in Natural Waters. New York: John Wiley & Sons. 58. Timmer, J. M. K., Kromkamp, J., Robbertsen, T. (199 4), “Lactic acid separation from fermentation broths by reverse osmosis and nanofilt ration ”, J. Membr. Sci. vol. 92, no. 2, pp. 185-197. 59. Van Boxtel, A. J. B., and Otten, Z. E. H. (1993), “ New strategies for economic optimal membrane fouling control based on dynamic o ptimization”, Desalination vol. 90, no.1-3, pp. 363-377. 60. Vrijenhoek, E.M., Hong, S., and Elimelech, M. (2001 ). “Influence of membrane surface properties on initial rate of colloidal fou ling or reverse osmosis and nanofiltration membranes,” J. Membr. Sci ., vol.188, no. 1, pp.115-128. 61. United States Department of Interior, Bureau of Rec lamation (USBR). The Desalting and water treatment membrane manual: A guide to mem brane for Municipal Water

PAGE 130

104 Treatment (Second Edition), Water Treatment Technol ogy Program – Report No. 29, July 1998. R98-5. 62. Wiesner, M.R. and Chellam, S. (1992), “Mass transpo rt considerations for pressure driven membrane processes”, J. Am. Water Works Assoc ., vol. 84, pp. 88 95. 63. Williams, M. E. (2003), “ A brief review of reverse osmosis membrane technolo gy”, http://www.wescinc.com/files/RO_Review.pdf, Accesse d 11/03/2005. 64. Winfield, B. A. (1979), “A study of the factors aff ecting the rate of fouling of reverse osmosis membranes treating secondary sewage effluen ts,” Water Res. vol. 13, no. 7, pp. 565-569. 65. Yiantsios, S.G. and Karabelas, A.J., (1998), “The e ffect of colloid stability on membrane fouling,” Desalination vol.118, no. 1-3, pp.143-152. 66. Zhu, X. and Elimelech, M. (1997). “Colloidal foulin g of reverse osmosis membranes: Measurements and Fouling Mechanisms,” Environmental Science and Technology vol. 31, no.12, pp. 3654-3662. 67. Zeta – Meter Manual, ZM 80, Zeta – Meter, Inc., S taunton, VA.

PAGE 131

105 Appendices

PAGE 132

106 Appendix A: Kaolin and Membrane Characterizing Data Table A.1 Zeta Potential Values of Kaolin Feed pHD.I water CaCl2 and Na2CO30.0005 M0.0015 M 30.0004-6.9005 -6-13.00-6.27-11.30 8 -9.19-16.00 10-15.40-10.13 Table A.2 Pure Water Permeability Data for LFC 1 Membrane Pressure (Pa) (x 10 -6 ) Run 1Run 2Run 3Run 4 1.381.601.481.511.452.072.272.202.302.252.763.033.053.013.023.453.753.813.793.83 Pure water flux (gm/cm 2 .s*10 5 ) Table A.3 Calculation of Supersaturation Factor for CaCO3 CaCl2 and Na2CO3 Conc. (M) 0.000010.00010.00050.000750.0010.00150.01 pH 9.09.09.09.09.09.09.0 Ka1 (Sawyer et al., 1994) 4.30E-074.30E-074.30E-074.30E-074.30E-074.30E-074.3 0E-07 Ka2 (Sawyer et al., 1994) 4.7E-114.7E-114.7E-114.7E-114.7E-114.7E-114.7E-11 [CO3 2] (M) See Note (a) 9.51E-069.51E-054.75E-047.13E-049.51E-041.43E-039.5 1E-03 [Ca2+] (M) 0.000010.00010.00050.000750.0010.00150.01 I (M) See Note (b)3.90E-053.90E-041.95E-032.93E-033. 90E-035.85E-033.90E-02 log10 CO3 2See Note ( c) -1.25E-02-3.95E-02-8.83E-02-1.08E-01-1.25E-01-1.53E -01-3.95E-01 log10 Ca2+-3.12E-03-9.88E-03-2.21E-02-2.70E-02-3.12E-02-3.82E -02-9.88E-02 CO3 29.72E-019.13E-018.16E-017.79E-017.50E-017.03E-014.0 3E-01 Ca2+9.93E-019.78E-019.50E-019.40E-019.31E-019.16E-017.9 7E-01 [Ca2-][CO3 2-](M2) ------(2) 9.17E-118.49E-091.84E-073.92E-076.64E-071.38E-063.0 5E-05 Supersaturation factor (2)/(1)0.0282.56455.688118.3 38200.485416.0739213.646 Ksp of CaCO3 (M2) 3.31E-09(Sawyer et al., 1994) Notes: (a) (Sawyer et al., 1994) (b) (Sawyer et al., 1994) ( c) (Sawyer et al., 1994) [][] 2 1 2 2 3 2 2 3. 10 10 1a a pH a pHK K K CO Na CO-+ + = ==n i i iC Z I1 22 1 5.0 2 105.0 log I zi i=g

PAGE 133

107 Appendix A: (Continued) Table A.4 Pure Water Volumetric Flux (m3/m2/s) at Different CaCl2 Concentration and Crossflow Velocities for LFC 1 Membrane 4.04 cm/s26.26 cm/s4.04 cm/s26.26 cm/s 1.38E+061.35E-051.442E-051.45E-051.442E-052.07E+062.02E-052.121E-052.03E-052.007E-052.76E+062.6E-052.824E-052.45E-052.716E-053.45E+062.92E-053.311E-052.67E-053.1E-05 CaCl2 Concentration 0.001 M 0.0005 M Transmembrane Pressure (Pa) Table A.5 t-Test Results for Kaolin (Transmembrane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 1.62 and 4.04 cm/s) Variable 1Variable 2 Mean1.5256751.464959MeanMeanVariance0.0010210.00024VarianceVarianceObservations1718ObservationsObservations Hypothesized Mean Difference 0 Hypothesized Mean Difference Hypothesized Mean Difference df23dfdft Stat7.088427t Statt StatP(T<=t) one-tail1.6E-07P(T<=t) one-tailP(T<=t) onetail t Critical one-tail1.713872t Critical one-tailt Cri tical one-tail P(T<=t) two-tail3.2E-07P(T<=t) two-tailP(T<=t) twotail t Critical two-tail2.068658t Critical two-tailt Cri tical two-tail

PAGE 134

108 Appendix B: Kaolin Particles Size Distribution Figure B.1 Particle Size Distribution. Kaolin in Di stilled Water, Kaolin Concentration = 150 mg/l, pH = 6.7 Figure B.2 Particle Size Distribution. Kaolin in Di stilled Water, Kaolin Concentration = 150 mg/l, pH = 9.0

PAGE 135

109 Appendix B: (Continued) Figure B.3 Particle Size Distribution. Kaolin in 0. 0005 M CaCO3, Kaolin Concentration = 150 mg/l, pH = 9.0 Figure B.4 Particle Size Distribution. Kaolin in 0. 0015 M CaCO3, Kaolin Concentration = 150 mg/l, pH = 9.0

PAGE 136

110 Appendix C: Permeation Data for Kaolin Runs Table C.1 Permeation Data for Kaolin Runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 50 mg/l, Crossflow Velocit y = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.001.551.000.501.530.991.001.551.001.501.530.992.001.551.002.501.571.013.001.571.013.501.551.004.001.530.994.501.551.005.001.551.005.501.530.996.001.551.00 Table C.2 Permeation Data for Kaolin Runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 1.62 cm/s, pH = 6.7, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.001.511.000.071.470.980.231.470.980.521.531.010.771.531.011.001.561.041.281.551.021.531.561.042.001.561.042.501.561.043.001.511.003.501.531.014.001.470.984.501.531.015.001.490.995.501.551.026.001.551.02

PAGE 137

111 Appendix C: (Continued) Table C.3 Permeation Data for Kaolin runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.001.481.000.051.481.000.221.481.000.351.460.990.581.460.990.781.440.981.001.440.981.251.440.981.581.460.992.031.460.992.501.481.003.001.481.003.501.440.984.001.440.984.501.481.005.001.481.005.501.481.006.001.460.99

PAGE 138

112 Appendix C: (Continued) Table C.4 Permeation Data for Kaolin Runs. Transmem brane Pressure = 1,380 kPa, Kaolin Concentration = 150 mg/l, Crossflow veloci ty = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.01.511.000.11.531.010.21.511.000.41.490.990.61.490.990.81.470.981.01.450.961.31.450.961.61.490.992.01.470.982.51.511.003.01.511.003.51.470.984.01.551.024.51.490.995.01.490.995.51.511.006.01.470.98 Table C.5 Permeation Data for Kaolin Runs. Transmem brane Pressure = 2,070 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 6.8, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105)Rel. Flux (V(t)/V(0) 0.002.251.000.102.251.000.252.301.020.532.271.011.002.271.011.502.220.992.002.220.992.502.190.973.002.190.973.502.150.964.002.140.954.502.140.955.002.070.925.502.040.916.001.990.89

PAGE 139

113 Appendix C: (Continued) Table C.6 Permeation Data for Kaolin Runs. Transmem brane Pressure = 2,070 kPa, Kaolin Concentration = 150 mg/l, Crossflow Velocit y = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.002.271.000.102.220.980.252.271.000.532.240.991.002.240.991.502.220.982.002.220.982.502.220.983.002.210.973.502.250.994.002.250.994.502.240.995.002.190.975.502.160.956.002.180.96 Table C.7 Permeation Data for Kaolin Runs. Transmem brane Pressure = 2,760 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.011.000.082.980.990.273.041.010.452.910.971.002.940.981.502.890.962.002.780.922.502.850.953.002.700.903.502.640.884.002.590.864.502.560.855.002.510.835.502.450.826.002.450.82

PAGE 140

114 Appendix C: (Continued) Table C.8 Permeation Data for Kaolin Runs. Transmem brane Pressure = 2,760 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.051.000.083.010.990.273.091.010.452.940.971.002.980.981.502.930.962.002.940.972.502.890.953.002.860.943.502.820.924.002.790.914.502.730.905.002.490.825.502.750.906.002.640.87 Table C.9 Permeation Data for Kaolin Runs. Transmem brane Pressure = 3,450 kPa, Kaolin Concentration = 50 mg/l, Crossflow Velocit y = 4.04 cm/s, pH = 6.8, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.811.000.053.821.000.183.790.990.383.700.970.503.740.980.773.750.981.003.660.961.253.670.961.503.660.961.753.540.932.003.500.922.303.440.902.503.420.903.003.350.883.503.270.864.003.260.864.503.180.835.003.120.825.503.040.806.003.040.80

PAGE 141

115 Appendix C: (Continued) Table C.10 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 3,450 kPa, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.751.000.053.761.000.183.730.990.383.640.970.503.680.980.773.690.981.003.600.961.253.610.961.503.600.961.753.540.942.003.480.932.303.430.922.503.400.913.003.360.903.503.320.884.003.280.884.503.250.875.003.200.855.503.160.846.003.130.83

PAGE 142

116 Appendix C: (Continued) Table C.11 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 3,450 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.801.000.083.780.990.273.891.020.533.720.980.923.630.961.003.520.931.503.470.912.003.370.892.503.210.843.003.140.833.503.070.814.002.920.774.502.800.745.002.860.755.502.720.726.002.650.70 Table C.12 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 3,450 kPa, Kaolin Concentration = 150 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.831.000.073.941.030.173.831.000.373.851.000.583.630.950.753.881.011.003.630.951.253.610.941.503.540.922.003.520.922.553.390.893.003.320.873.503.230.844.003.160.824.503.090.815.002.980.785.502.900.766.002.810.73

PAGE 143

117 Appendix C: (Continued) Table C.13 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 3,450 kPa, Kaolin Concentration = 250 mg/l, Crossflow Veloci ty = 4.04 cm/s, pH = 6.7, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.771.000.073.690.980.183.620.960.253.540.940.453.460.920.653.420.911.003.310.881.253.290.871.503.230.862.003.110.822.503.050.813.002.880.763.502.800.744.002.690.714.502.620.695.002.480.665.502.430.645.752.390.63

PAGE 144

118 Appendix C: (Continued) Table C.14 Permeation Data for Kaolin Runs. Transme mbrane Pressure = 3,450 kPa, Kaolin Concentration = 250 mg/l, Crossflow Velocit y = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Vol. Flux (V)(m3/m2/s*105) Rel. Flux (V(t)/V(0) 0.003.801.000.053.811.000.183.780.990.383.690.970.503.650.960.773.610.951.003.530.931.253.490.921.503.470.911.753.440.912.003.330.882.303.330.882.503.210.843.003.090.813.502.990.794.002.910.774.502.850.755.002.670.705.502.610.696.002.580.68

PAGE 145

119 Appendix D: SPSS Statistical Analysis Results Table D.1 Univariate Analysis of Variance – Tes ts Between – Subjects Effects Table D.2 Univariate Analysis of Variance – Para meter Estimation

PAGE 146

120 Appendix D: (Continued) Figure D.1 Relationship Between Experimental Data a nd Model Results

PAGE 147

121 Appendix E: Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs Table E.1 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 1,380 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 8.9, Tempe rature = 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.001.451.000.051.451.000.181.451.000.421.440.990.771.370.941.071.340.931.321.310.901.501.290.892.001.340.932.501.270.883.001.260.873.501.330.924.001.330.924.501.320.915.001.290.895.501.330.926.001.330.92 Table E.2 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,070 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 9.1, Tempe rature = 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.002.271.000.052.220.980.252.190.970.502.140.940.752.140.941.002.090.921.502.000.882.002.000.882.502.020.893.002.000.883.501.970.874.001.960.864.501.950.865.001.960.865.501.950.866.001.950.86

PAGE 148

122 Appendix E: (Continued) Table E.3 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,760 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 9.0, Tempe rature = 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.003.051.000.052.970.970.282.950.970.522.830.931.002.680.881.502.560.842.002.520.832.502.440.803.002.440.803.502.350.773.452.270.744.002.250.744.502.250.745.002.230.735.502.210.726.002.190.72 Table E.4 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 9.0, Tempe rature = 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.003.731.000.053.510.940.133.470.930.183.430.920.473.290.880.753.020.811.003.000.801.332.850.761.672.760.742.002.650.712.502.590.693.002.470.663.502.380.644.002.330.624.502.300.625.002.130.575.032.180.595.502.130.576.002.140.57

PAGE 149

123 Appendix E: (Continued) Table E.5 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Crossflow velocity = 4.04 cm/s, pH = 9.0, Tempe rature = 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.003.751.000.073.310.880.173.140.840.303.090.820.452.980.800.582.850.760.772.820.751.002.780.741.252.580.691.502.560.681.772.520.672.002.460.662.352.390.642.502.320.622.752.250.603.002.190.583.502.120.574.002.050.554.531.990.535.001.860.495.501.800.486.001.800.48

PAGE 150

124 Appendix E: (Continued) Table E.6 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 5.5, Tempe rature= 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.003.781.000.053.781.000.333.550.940.523.570.950.753.620.961.003.500.931.383.560.942.003.360.892.273.440.912.503.510.932.823.400.903.003.430.913.503.370.894.053.330.884.503.270.875.003.220.855.503.250.865.673.510.935.873.570.956.053.570.956.233.570.95

PAGE 151

125 Appendix E: (Continued) Table E.7 Permeation Data for CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Crossflow Velocity = 4.04 cm/s, pH = 4.0, Tempe rature = 24 oC Time (Hrs.) Per. Vol. Flux (V(t))(m3/m2/s*105) Rel. Flux (V(t))/V(0) 0.003.831.000.053.831.000.123.841.000.373.760.980.453.750.980.553.710.970.883.540.931.133.540.931.383.480.911.883.420.892.383.350.882.883.260.853.383.220.843.403.190.833.883.160.834.003.160.834.073.500.914.153.530.924.283.590.944.573.560.934.823.600.945.003.480.91

PAGE 152

126 Appendix F: Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs Table F.1 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 1,380 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 15 0 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.2, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.001.461.000.071.400.960.231.440.990.501.380.951.001.410.971.501.380.952.001.320.912.501.370.943.001.280.883.501.280.884.001.250.864.501.340.925.001.310.905.501.250.866.001.260.87 Table F.2 Purewater Permeability Data at the Start. Transmembrane Pressure = 1,380 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 0.0012.401.4910.960.0812.301.4810.870.2512.301.4810.870.5012.101.4510.690.7512.001.4410.601.0012.201.4710.781.2512.101.4510.691.5012.001.4410.602.0012.101.4510.69

PAGE 153

127 Appendix F: (Continued) Table F.3 Purewater Permeability Data at the End. T ransmembrane Pressure = 1,380 kPa, Crossflow Velocity = 4.04 cm/s, Temperature= 24 oC Time (From 6 hrs)Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0011.901.43 10.51 6.080.0811.801.42 10.43 6.250.2512.001.44 10.60 6.500.5012.001.44 10.60 6.750.7512.101.45 10.69 7.001.0011.801.42 10.43 7.251.2511.901.43 10.51 7.501.5012.001.44 10.60 8.002.0012.101.45 10.69 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw(gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.1 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 1,380 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9. 2, Temperature = 24 oC

PAGE 154

128 Appendix F: (Continued) Table F.4 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,070 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 1 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.1, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.002.281.000.052.150.940.222.210.970.422.170.950.602.120.930.802.010.881.002.040.891.502.000.882.001.910.842.501.880.833.001.860.813.501.810.804.001.730.764.501.760.775.001.690.745.501.570.696.001.630.71 Table F.5 Purewater Permeability Data at the Start. Transmembrane Pressure = 2,070 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm2/s*105) 0.0019.302.3211.370.0819.302.3211.370.2519.202.3111.310.5019.002.2811.190.7519.102.3011.251.0018.902.2711.131.2519.002.2811.191.5019.102.3011.252.0019.002.2811.19

PAGE 155

129 Appendix F: (Continued) Table F.6 Purewater Permeability Data at the End. T ransmembrane Pressure = 2,070 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs)Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0017.502.1010.316.080.0817.502.1010.316.250.2517.602.1110.376.500.5017.702.1310.436.750.7517.602.1110.377.001.0017.902.1510.547.251.2517.802.1410.487.501.5017.602.1110.378.002.0017.902.1510.54 0.0 0.5 1.0 1.5 2.0 2.5 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw(gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.2 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,070 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4. 04 cm/s, pH = 9.1, Temperature = 24 oC

PAGE 156

130 Appendix F: (Continued) Table F.7 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,760 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 1 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.1, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.003.011.000.072.730.910.252.920.970.532.870.951.002.700.901.502.580.862.002.490.832.502.330.773.002.310.773.502.240.754.002.120.714.502.110.705.002.040.685.501.950.656.001.930.646.001.930.64 Table F.8 Purewater Permeability Data at the Start. Transmembrane Pressure = 2,760 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) K w (gm/cm 2 /s*10 5 ) 0.0027.103.2611.970.0826.603.2011.750.2526.103.1411.530.5025.403.0511.220.7525.203.0311.131.0024.802.9810.961.2525.103.0211.091.5024.902.9911.002.0025.103.0211.09

PAGE 157

131 Appendix F: (Continued) Table F.9 Purewater Permeability Data at the End. T ransmembrane Pressure = 2,760 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs) Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0019.002.288.396.080.0819.302.328.536.250.2519.302.328.536.500.5019.602.368.666.750.7519.502.348.617.001.0019.202.318.487.251.2519.502.348.617.501.5019.602.368.668.002.0019.602.368.66 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.3 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 2,760 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9. 1, Temperature = 24 oC

PAGE 158

132 Appendix F: (Continued) Table F.10 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 1 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.3, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.003.801.000.033.300.870.203.190.840.383.070.810.582.960.780.952.790.741.252.630.691.832.460.652.122.390.632.502.290.603.002.240.593.502.150.564.002.020.534.501.960.515.001.900.505.501.840.496.001.810.48 Table F.11 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Rel. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 0.0032.203.8711.380.0832.103.8611.340.2532.003.8511.310.5031.703.8111.200.7531.803.8211.241.0031.703.8111.201.2531.903.8311.271.5032.003.8511.312.0031.903.8311.27

PAGE 159

133 Appendix F: (Continued) Table F.12 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs) Time (Hrs.) Per. Flow (ml/min) Rel. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0020.402.457.216.080.0820.702.497.326.250.2520.602.487.286.500.5020.602.487.286.750.7520.802.507.357.001.0020.902.517.397.251.2520.902.517.397.501.5020.802.507.358.002.0020.702.497.32 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.502468Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.4 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9. 3, Temperature = 24 oC

PAGE 160

134 Appendix F: (Continued) Table F.13 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 5 0 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.2, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.003.761.000.052.220.590.103.380.900.353.420.910.633.090.821.002.950.781.502.740.732.002.520.672.502.430.653.002.300.613.502.170.584.002.150.574.502.040.545.001.980.535.501.960.526.001.960.52 Table F.14 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 0.0033.203.9911.730.0833.003.9711.660.2532.703.9311.560.5032.503.9111.490.7532.103.8611.341.0031.703.8111.201.2531.303.7611.061.5031.403.7711.102.0031.703.8111.202.5031.403.7711.103.0031.303.7611.063.5031.403.7711.10

PAGE 161

135 Appendix F: (Continued) Table F.15 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs)Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0022.702.738.026.080.0822.802.748.066.250.2522.702.738.026.500.5022.602.727.996.750.7522.602.727.997.001.0022.802.748.067.251.2522.702.738.027.501.5022.602.727.998.002.0022.602.727.99 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.5 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.2 Temperature = 24 oC

PAGE 162

136 Appendix F: (Continued) Table F.16 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 2 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 8.9, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.003.851.000.083.881.010.323.630.940.653.220.840.833.240.841.082.860.741.502.820.732.002.550.662.582.500.653.002.320.603.502.110.554.001.850.484.501.720.455.001.600.415.501.410.376.001.410.37 Table F.17 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 0.0035.004.2112.370.0834.304.1212.120.2534.004.0912.020.5034.104.1012.050.7533.604.0411.871.0033.304.0011.771.2533.003.9711.661.5032.403.8911.452.0032.003.8511.312.5032.203.8711.383.0032.003.8511.313.3031.903.8311.27

PAGE 163

137 Appendix F: (Continued) Table F.18 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs)Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0017.402.096.156.080.0817.502.106.186.250.2517.802.146.296.500.5017.802.146.296.750.7517.602.116.227.001.0017.702.136.267.251.2517.802.146.297.501.5017.702.136.268.002.0017.802.146.29 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.6 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0005 M, Kaolin Concentration = 250 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 8. 9, Temperature = 24 oC

PAGE 164

138 Appendix F: (Continued) Table F.19 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 5 0 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.0, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.003.821.000.053.580.940.103.450.900.353.320.870.633.050.801.002.790.731.502.650.692.002.390.632.502.290.603.002.160.573.502.030.534.001.960.514.501.880.495.001.820.485.501.790.476.001.720.45 Table F.20 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 0.0035.104.2212.410.0835.004.2112.370.2534.504.1512.190.5034.004.0912.020.7533.604.0411.871.0033.003.9711.661.2533.003.9711.661.5032.603.9211.522.0031.903.8311.272.5031.503.7911.133.0031.603.8011.173.5031.703.8111.20

PAGE 165

139 Appendix F: (Continued) Table F.21 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs) Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0020.502.467.256.080.0820.702.497.326.250.2521.002.527.426.500.5021.002.527.426.750.7521.602.607.637.001.0022.002.647.787.251.2521.802.627.707.501.5021.902.637.748.002.0022.002.647.78 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 02468Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.7 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0005 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.0 Temperature = 24 oC

PAGE 166

140 Appendix F: (Continued) Table F.22 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 1 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.3, Temperature = 24 oC Time (Hrs.) Per. Flux (m 3 /m 2 /s*10 5 ) Rel. Flux 0.003.801.000.033.300.870.203.190.840.383.070.810.582.960.780.952.790.741.252.630.691.832.460.652.122.390.632.502.290.603.002.240.593.502.150.564.002.020.534.501.960.515.001.900.505.501.840.496.001.810.48 Table F.23 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 0.0032.203.8711.380.0832.103.8611.340.2532.003.8511.310.5031.703.8111.200.7531.803.8211.241.0031.703.8111.201.2531.903.8311.271.5032.003.8511.312.0031.903.8311.27

PAGE 167

141 Appendix F: (Continued) Table F.24 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs)Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0020.402.457.216.080.0820.702.497.326.250.2520.602.487.286.500.5020.602.487.286.750.7520.802.507.357.001.0020.902.517.397.251.2520.902.517.397.501.5020.802.507.358.002.0020.702.497.32 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.8 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 150 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9. 3, Temperature = 24 oC

PAGE 168

142 Appendix F: (Continued) Table F.25 Permeation Data for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 2 50 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9.2, Temperature = 24 oC Time (Hrs.)Per. Flux (m3/m2/s*105)Rel. Flux 0.003.801.000.073.570.940.283.360.880.523.050.800.752.790.731.032.690.711.532.310.612.052.000.532.501.840.493.001.610.423.501.490.394.001.430.384.501.320.355.001.170.315.501.070.286.001.070.28 Table F.26 Purewater Permeability Data at the Start Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm2/s*105) 0.0033.504.0311.840.0833.003.9711.660.2532.703.9311.560.5032.503.9111.490.7532.003.8511.311.0031.803.8211.241.2532.103.8611.341.5032.003.8511.312.0031.903.8311.27

PAGE 169

143 Appendix F: (Continued) Table F.27 Purewater Permeability Data at the End. Transmembrane Pressure = 3,450 kPa, Crossflow Velocity = 4.04 cm/s, Temperature = 24 oC Time (From 6 hrs) Time (Hrs.) Per. Flow (ml/min) Per. Flux (m 3 /m 2 /s*10 5 ) Kw (gm/cm 2 /s*10 5 ) 6.000.0015.601.875.516.080.0815.901.915.626.250.2516.001.925.656.500.5015.801.905.586.750.7516.001.925.657.001.0015.901.915.627.251.2516.101.935.697.501.5016.001.925.658.002.0015.901.915.62 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 02468 Time (Hrs)Permeate flux (m3/m2/s*105)0 5 10 15 20 25 30Kw (gm/cm2/s/atm*105) Clay+Salt Kaolin only CaCO3 Kw (Initial) Kw (Final) Figure F.9 Permeation and Purewater Permeability Co efficient (Kw) vs Time for Kaolin and CaCl2 Plus Na2CO3 Scaling Runs. Transmembrane Pressure = 3,450 kPa, CaCl2 Concentration = 0.0015 M, Na2CO3 Concentration = 0.0015 M, Kaolin Concentration = 250 mg/l, Crossflow Velocity = 4.04 cm/s, pH = 9. 2, Temperature = 24 oC

PAGE 170

144 Appendix G: Fouling Model Calibration Data and Resu lts Table G.1 Calculated 1/V(t) and V(t)*t Values for Kaolin and CaCO3 at 1,380 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m 3 /m 2 *s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 *s*10 5 ) (m)(m) 0.001.460.68493203.01000.071.400.7143912401.430.0034320.0034320.231.440.6920668401.420.0085340.0085340.501.380.72215618001.410.0135830.0255481.001.410.70679136001.400.0251960.0507451.501.380.72215654001.400.0251960.0759412.001.320.75498172001.350.0243840.1003252.501.370.73009290001.350.0242480.1245733.001.280.781628108001.320.0238420.1484143.501.280.781628126001.280.0230290.1714434.001.250.800462144001.260.0227580.1942014.501.340.746498162001.290.0233000.2175015.001.310.763659180001.320.0238420.2413435.501.250.800462198001.280.0230290.2643716.001.260.790933216001.260.0226220.286994 y = 0.3433x + 0.7024 R2 = 0.7557 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 00.10.20.30.4 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G.1 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 1,380 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin

PAGE 171

145 Appendix G: (Continued) Table G.2 Calculated 1/V(t) and V(t)*t Values for Kaolin and CaCO3 at 2,070 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.002.280.43859603.01000.052.150.4645661802.220.0039890.0039890.222.210.4526547802.180.0130850.0130850.422.170.46153015002.190.0157530.0328280.602.120.47076021602.150.0141600.0469880.802.010.49728228802.070.0148870.0618741.002.040.49037536002.030.0145810.0764551.502.000.50080954002.020.0363240.1127792.001.910.52306772001.950.0351770.1479562.501.880.53093390001.900.0341580.1821143.001.860.539038108001.870.0336480.2157623.501.810.551672126001.830.0330100.2487724.001.730.578804144001.770.0318630.2806354.501.760.569468162001.740.0313540.3119895.001.690.593395180001.720.0309710.3429605.501.570.636162198001.630.0293140.3722746.001.630.614035216001.600.0288040.401079 y = 0.4242x + 0.4531 R2 = 0.9665 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 00.10.20.30.40.5 V(t)*t (m)1/ V(t) (s/m*10-5) Figure G.2 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 2,070 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin

PAGE 172

146 Appendix G: (Continued) Table G.3 Calculated 1/V(t) and V(t)*t Values for Kaolin and CaCO3 at 2,760 kPa, 4.04 cm/s, 0.0005 M CaCO3 and150 mg/l of Kaolin Timet Per. FluxV(t) (m/s*10 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.003.010.33222591403.01000.072.730.3660833322402.870.006890.006890.252.920.3421135909002.830.018660.018660.532.870.34833383719202.900.029550.055101.002.700.37080698736002.780.046770.101871.502.580.38834515654002.640.047450.149312.002.490.40192365872002.530.045570.194882.502.330.42891853090002.410.043380.238263.002.310.432143482108002.320.041810.280073.502.240.445543280126002.280.041030.321094.002.120.471107238144002.180.039300.360404.502.110.475000686162002.110.038050.398455.002.040.491240026180002.070.037270.435725.501.950.513170384198001.990.035860.471576.001.930.517793541216001.940.034920.50649 y = 0.3534x + 0.3383 R2 = 0.9811 0 0.1 0.2 0.3 0.4 0.5 0.6 00.10.20.30.40.50.6 V(t)*t (m)1/ V(t) (s/m*10-5) Figure G.3 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 2,760 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin

PAGE 173

147 Appendix G: (Continued) Table G.4 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*tV(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.003.800.26315803.01000.033.300.3033011203.550.0042580.0042580.203.190.3139437203.240.0194470.0194470.383.070.32535913803.130.0206540.0443590.582.960.33763721003.020.0217270.0660860.952.790.35789534202.880.0379890.1040751.252.630.38073945002.710.0292710.1333461.832.460.40669965882.540.0530900.1864372.122.390.41810176322.430.0253200.2117572.502.290.43645790002.340.0320310.2437883.002.240.447368108002.260.0407380.2845273.502.150.466009126002.190.0394310.3239574.002.020.494330144002.080.0375190.3614774.501.960.511278162001.990.0358090.3972865.001.900.526316180001.930.0347030.4319895.501.840.542265198001.870.0336970.4656866.001.810.552307216001.830.0328920.498578 y = 0.5296x + 0.299 R2 = 0.986 0 0.1 0.2 0.3 0.4 0.5 0.6 00.20.40.6 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G.4 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 150 mg/l of Kaolin

PAGE 174

148 Appendix G: (Continued) Table G. 5 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005M CaCO3 and 250 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*tV(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.003.850.25974003.01000.083.880.2576463003.870.01159700.0115970.323.630.27541411403.760.03155120.0315510.653.220.31017523403.430.04112930.0842770.833.240.30867730003.230.02132990.1056071.082.860.35011638883.050.02706550.1326731.502.820.35497854002.840.04288990.1755632.002.550.39200172002.680.04831280.2238762.582.500.39935192882.530.05277500.2766513.002.320.431730108002.410.03644170.3130923.502.110.473304126002.210.03986160.3529544.001.850.541492144001.980.03563600.3885904.501.720.580874162001.780.03211460.4207055.001.600.626432180001.660.02986100.4505665.501.410.709957198001.500.02704390.4776096.001.410.709957216001.410.02535370.502963 y = 0.8191x + 0.23 R2 = 0.9288 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 00.10.20.30.40.50.6 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G.5 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 250 mg/l of Kaolin

PAGE 175

149 Appendix G: (Continued) Table G.6 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005M CaCO3 and 50 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.003.760.265957403.01000.053.500.28571431803.630.0065340.0065340.103.380.29573653603.440.0061930.0127270.353.420.292350212603.400.0306090.0433360.633.090.323188822803.260.0332250.0765611.002.950.338919236003.020.0398950.1164561.502.740.364741654002.850.0512300.1676862.002.520.396869172002.630.0473530.2150392.502.430.411805190002.470.0445330.2595713.002.300.4352031108002.360.0425350.3021063.502.170.4614201126002.230.0401850.3422914.002.150.4642166144002.160.0388930.3811844.502.040.4909984162002.100.0377180.4189015.001.980.5039194180002.010.0361900.4550915.501.960.5106383198001.970.0354850.4905766.001.960.5106383216001.960.0352500.525826 y = 0.477x + 0.2833 R2 = 0.9865 0 0.1 0.2 0.3 0.4 0.5 0.6 00.20.40.6 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G.6 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0005 M CaCO3 and 50 mg/l of Kaolin

PAGE 176

150 Appendix G: (Continued) Table G.7 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 50 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m/s*10 5 )(s/m*10 -5 ) (Sec) (m/s*10 5 ) (m)(m) 0.003.820.2603.010.000.000.053.580.281803.700.006660.006660.103.450.293603.510.006330.006330.353.320.3012603.380.030440.043430.633.050.3322803.180.032470.075901.002.790.3636002.920.038520.114421.502.650.3854002.720.048940.163362.002.390.4272002.520.045360.208722.502.290.4490002.340.042140.250863.002.160.46108002.230.040110.290973.502.030.49126002.100.037720.328704.001.960.51144002.000.035930.364634.501.880.53162001.920.034620.399255.001.820.55180001.850.033310.432555.501.790.56198001.800.032470.465026.001.720.58216001.760.031630.49666 y = 0.6244x + 0.2786 R2 = 0.9952 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.000.200.400.60 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G. 7 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 50 mg/l of Kaolin

PAGE 177

151 Appendix G: (Continued) Table G.8 Calculated 1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 150 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.003.790.26385203.01000.032.220.4508031203.000.0036050.0036050.202.820.3545847202.520.0151150.0151150.382.790.35883913802.800.0185030.0372230.582.660.37535521002.730.0196230.0568470.952.420.41340934202.540.0335480.0903951.252.320.43129745002.370.0255830.1159771.832.170.46005065882.250.0468990.1628772.122.130.46968576322.150.0224600.1853372.502.130.46968590002.130.0291260.2144633.001.970.506835108002.050.0369190.2513823.501.810.553764126001.890.0340100.2853924.001.790.557203144001.800.0324050.3177964.501.640.610270162001.720.0309000.3486965.001.630.614450180001.630.0293950.3780915.501.530.654816198001.580.0283920.4064826.001.460.684807216001.490.0268870.433369 y = 0.7423x + 0.3399 R2 = 0.9062 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 00.10.20.30.40.5 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G.8 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 150 mg/l of Kaolin

PAGE 178

152 Appendix G: (Continued) Table G.9 Calculated1/V(t) vs V(t)*t Values for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 250 mg/l of Kaolin Time tPer. Flux V(t) 1/V(t) t Average V(t)V(t)*t V(t)*t (Hrs.) (m 3 /m 2 /s*10 5 )(s/m*10 -5 ) (Sec) (m 3 /m 2 /s*10 5 ) (m)(m) 0.003.800.26315803.8000.073.570.2803992403.680.0088400.0088400.283.360.29786010203.460.0270020.0358420.523.050.32788618603.200.0269100.0627520.752.790.35821927002.920.0245340.0872861.032.690.37130537082.740.0276430.1149291.532.310.43253155082.500.0450470.1599762.052.000.49887073802.160.0404020.2003782.501.840.54210590001.920.0311780.2315573.001.610.620731108001.730.0311010.2626573.501.490.672031126001.550.0278910.2905494.001.430.700998144001.460.0262310.3167804.501.320.759961162001.370.0246820.3414615.001.170.855956180001.240.0223570.3638195.501.070.934664198001.120.0201440.3839626.001.070.934664216001.070.0192580.403221 y = 1.6545x + 0.2157 R2 = 0.9651 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 00.10.20.30.40.5 V(t)*t (m)1/ V (t) (s/m*10-5) Figure G.9 1/V(t) vs V(t)*t for Kaolin and CaCO3 at 3,450 kPa, 4.04 cm/s, 0.0015 M CaCO3 and 250 mg/l of Kaolin

PAGE 179

1 About the Author Dhananjaya Niriella obtained a B.Sc (Hons.) in civi l engineering from the University of Moratuwa, Sri Lanka and a M. Eng in w ater and wastewater engineering from Asian Institute of Technology, Thailand. He started his career as a civil engineer at Mihin du Keerthiratne Associates, an architectural partnership. Then he joined Central E ngineering Consultancy Bureau in Sri Lanka, where, he worked as a civil construction eng ineer and later as a design engineer in water resources, wastewater, structural planning an d designing. In 2000, he joined Engineering Consultants Ltd. In 1998, Mr. Niriella was selected by the water di vision of the United Nations Food and Agriculture Organization, under its young professionals program to carry out country water resources study for the Asian region. Before joining University of South Florida, he worked at International Water Managemen t Institute. Besides Sri lanka, Mr. Niriella has worked in seve ral countries including, Thailand, Laos PDR and Italy.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001914855
003 fts
005 20071029113744.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071029s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001766
040
FHM
c FHM
035
(OCoLC)179589532
049
FHMM
090
TA145 (ONLINE)
1 100
Niriella, Dhananjaya P.
0 245
Investigating the fouling behavior of reverse osmosis membranes under different operating conditions
h [electronic resource] /
by Dhananjaya P. Niriella.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: This dissertation describes the investigation of the fouling of a reverse osmosis membrane under different operating conditions. A mass transfer model to predict the permeate flux decline is defined. These studies used kaolin clay and bentonite clay as the fouling particles. As the membranes, thin film Low fouling Composite 1 polyamide reverse osmosis flat sheet membranes were used. Baseline experiments using only kaolin in D.I. water were conducted. At an operating pressure of approximately 1,380 kPa, no flux decline was observed. These results established the effects of a membrane-particle interaction. For the fouling experiments with kaolin clay, experiments show a linear relationship between the mass of the deposited foulant layer and total permeate flux decline. The increased concentration of scale forming salts such as calcium chloride and sodium carbonate combined with clay particles has been found to increase flux decline. It also leads to the formation of a less porous cake layer on the membrane surface, which may be due to the particle surface charge. The increase in transmembrane pressure leads to the formation of a well compacted, less porous, cake layer on the membrane surface. The reduced porosity results in the deterioration of the permeate quality, which is a direct result of reduced back diffusion of the salt solution.A fouling model that combines a resistance-in-series model and a simplified-mass-transport relationship were used to predict the transient stage permeate flux of a reverse osmosis membrane. This model contains a constant which is a function of the operating condition and the ionic species in the feed solution. It was found that the results from the model agreed with the experimental results.
502
Dissertation (Ph.D.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 152 pages.
Includes vita.
590
Adviser: Robert P. Carnahan, Ph.D.
653
Scaling.
Concentration polarization.
Clay.
Salt.
Permeate.
690
Dissertations, Academic
z USF
x Civil Engineering
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1766