Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Assessment of causes of irreversible fouling in powdered activated carbon/ultrafiltration membrane (PAC/UF)… Isabel, Londono Cristina 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2011_fall_londono_isabel.pdf [ 1.85MB ]
Metadata
JSON: 24-1.0063210.json
JSON-LD: 24-1.0063210-ld.json
RDF/XML (Pretty): 24-1.0063210-rdf.xml
RDF/JSON: 24-1.0063210-rdf.json
Turtle: 24-1.0063210-turtle.txt
N-Triples: 24-1.0063210-rdf-ntriples.txt
Original Record: 24-1.0063210-source.json
Full Text
24-1.0063210-fulltext.txt
Citation
24-1.0063210.ris

Full Text

 ASSESSMENT OF CAUSES OF IRREVERSIBLE FOULING IN POWDERED ACTIVATED CARBON/ ULTRAFILTRATION MEMBRANE (PAC/UF) SYSTEMS  by Isabel Cristina Londono B.Eng. (Civil Engineering), Instituto Politécnico Nacional, Mexico, 2004  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in The Faculty of Graduate Studies (Civil Engineering)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) August 2011  © Isabel Cristina Londono, 2011  ii  ABSTRACT .......... Powdered Activated Carbon (PAC) has been successfully used in conjunction with membrane ultrafiltration (UF) to reduce taste, odour, colour and other concerns caused by organic material present in raw drinking water sources. PAC addition also typically significantly reduces the extent of fouling in hybrid PAC/UF systems. However, in some cases, PAC addition can have a significant negative effect by increasing irreversible membrane fouling. The present study was developed to assess the cause of irreversible fouling in a system for which PAC addition had a negative effect. The first part of the study evaluated if irreversible fouling could potentially be due to the breakdown of PAC particles in a PAC/UF membrane system. Particle size analysis of the virgin PAC suggested that pore plugging was unlikely because the PAC was too large. However, when exposed to relatively high shear conditions typical of UF systems, the size of the PAC was observed to decrease to a range comparable to that of the size of the pores in the UF membranes used.  However results from the analysis of a bench scale PAC/UF system suggested that irreversible fouling was not solely due to the direct plugging of membrane. The decrease in the permeability for the irreversibly fouled system with PAC was observed to be linear over time, suggesting that irreversible fouling was possible due to the formation of a cake layer by PAC. This was confirmed with FESEM imaging. The second part of the study aimed to characterize the foulants in the cake layer on PAC/UF hollow fibres. FESEM and SEM-EDX were performed to obtain insight on the characteristics of the membrane surface of the hollow fibres. Sonication and a novel solubilisation technique were used to analyze the foulants in/on the membrane surface of the hollow fibres. Organic and inorganic material extracted from the fouled membrane suggested that these have great influence on the PAC cake layer formation on membrane fibres. The results from the study indicate that PAC does not cause irreversible membrane fouling in PAC/UF systems by itself, but it may facilitate the absorption of organic and inorganic material causing irreversible fouling.   iii  TABLE OF CONTENTS ABSTRACT .......... ....................................................................................................................................... ii TABLE OF CONTENTS ............................................................................................................................. iii LIST OF TABLES ... ................................................................................................................................... vi LIST OF FIGURES .................................................................................................................................... vii LIST OF ABBREVIATIONS AND SYMBOLS ........................................................................................ ix ACKNOWLEDGEMENTS ........................................................................................................................ xii CHAPTER 1 INTRODUCTION ......................................................................................................... 1 CHAPTER 2 BACKGROUND AND LITERATURE REVIEW ........................................................ 2 2.1. Powdered Activated Carbon ......................................................................................................... 2 2.1.1. Particle size of PAC .............................................................................................................. 2 2.1.2. Abrasiveness of PAC ............................................................................................................ 3 2.1.3. Abrasion resistance of PAC .................................................................................................. 3 2.2. Membrane filtration ...................................................................................................................... 3 2.3. Membrane fouling ......................................................................................................................... 5 2.3.1. Mechanisms of membrane fouling ........................................................................................ 5 2.4. Powdered Activated Carbon/ Ultrafiltration system ..................................................................... 7 2.4.1 Fouling in PAC/UF systems ................................................................................................. 8 2.5. Conclusions ................................................................................................................................... 9 2.6. Research objectives ....................................................................................................................... 9 iv  CHAPTER 3 MATERIALS AND METHODS ................................................................................. 11 3.1. PAC/UF systems under investigation ......................................................................................... 11 3.2. Materials ..................................................................................................................................... 13 3.2.1. Hollow fibres membranes ................................................................................................... 13 3.2.2. Powdered Activated Carbon ............................................................................................... 14 3.3. Methods....................................................................................................................................... 15 3.3.1. Study of membrane plugging due to breakdown of PAC particle ...................................... 15 3.3.2. Detection of foulants in PAC/UF hollow fibres .................................................................. 20 CHAPTER 4 RESULTS AND DISCUSSION .................................................................................. 26 4.1. Study of membrane plugging due to abrasion of PAC particle................................................... 26 4.1.1. Assessment of PAC particle size over time ........................................................................ 26 4.1.2. Assessment of PAC pore plugging potential over time ...................................................... 28 4.2. Detection of foulants in PAC/UF hollow fibres .......................................................................... 29 4.2.1. Detection of PAC in irreversibly fouled membranes .......................................................... 29 4.2.2. Assessment of mechanism of fouling in irreversibly fouled membranes ........................... 32 4.2.3. Detection of foulants in irreversibly fouled membranes ..................................................... 33 CHAPTER 5 CONCLUSIONS .......................................................................................................... 38 REFERENCES ......... ................................................................................................................................. 39 APPENDIX A WATER QUALITY OF WATER TREATMENT PLANT......................................... 44 APPENDIX B OPERATING CONDITIONS OF GE UNIT A ........................................................... 45 APPENDIX C EXPERIMENTAL BENCH SCALE SYSTEM .......................................................... 46 v  APPENDIX D PARTICLE SIZE ANALYSIS .................................................................................... 47 APPENDIX E CLEAN WATER FLUX TEST ................................................................................... 48 APPENDIX F CALCULATIONS ....................................................................................................... 49 APPENDIX G TOC AND ICP RESULTS FROM HOLLOW FIBRE EXTRACTIONS ................... 51  vi  LIST OF TABLES ... TABLE 2-1  SUMMARY OF CONSTANT FLOW FOULING MODELS ....................................................................................... 7 TABLE 3-1 CHARACTERISTICS OF THE UNITS INVESTIGATED BY GE WATER & PROCESS TECHNOLOGIES .................... 13 TABLE 3-2  PHYSICAL CHARACTERISTICS OF THE MEMBRANE FIBRES USED IN THE PRESENT STUDY ............................ 14 TABLE 3-3 NOMENCLATURE AND BACKGROUND OF MEMBRANE FIBRES ...................................................................... 14 TABLE 3-4 PHYSICAL CHARACTERISTICS OF PAC TESTED ............................................................................................ 15 TABLE 3-5 SIZE DISTRIBUTION PARAMETERS ............................................................................................................... 17  TABLE A-1 WATER QUALITY OF RAW WATER FOR THE FULL SCALE PAC/UF MEMBRANE SYSTEM UNDER INVESTIGATION .................................................................................................................................................... 44 TABLE D-1 PARTICLE SIZE DISTRIBUTION OF PAC, AFTER CENTRIFUGATION ............................................................... 47 TABLE G-1 TOC AND INORGANIC (ICP-OES ANALYSIS) MATERIAL EXTRACTED FROM SONICATED HOLLOW FIBRES .. 51 TABLE G-2 INORGANIC MATERIAL EXTRACTED FROM SOLUBILISED HOLLOW FIBRES (ICP-OES ANALYSIS) ............... 55 TABLE G-3 ORGANIC MATERIAL EXTRACTED FROM SOLUBILISED FOULED AND VIRGIN HOLLOW FIBRES (GC ANALYSIS) ........................................................................................................................................................... 56  vii  LIST OF FIGURES FIGURE 2-1 SCHEMATIC OF MEMBRANE FILTRATION PROCESS ....................................................................................... 4 FIGURE 2-2 CLASSIFICATION AND REMOVAL EFFICIENCIES OF MEMBRANE PROCESSES (ADAPTED FROM EPA, 2005 AND CRITTENDEN ET AL., 2005) ............................................................................................................................ 4 FIGURE 2-3 SCHEMATIC OF THE FOUR CLASSICAL FOULING MODELS .............................................................................. 6 FIGURE 2-4 ARRANGEMENT OF OUTCOMES OF THE PRESENT STUDY ............................................................................ 10 FIGURE 3-1 SCHEMATIC OF SYSTEMS INVESTIGATED BY GE WATER & PROCESS TECHNOLOGIES ............................... 11 FIGURE 3-2 PERMEABILITY OF PAC/UF SYSTEM, UNIT A (PROVIDED BY GE WATER & PROCESS TECHNOLOGIES) .... 12 FIGURE 3-3 SCHEMATIC OF A BENCH SCALE AIR SPARGED SYSTEM .............................................................................. 16 FIGURE 3-4 SCHEMATIC OF A BENCH SCALE SUBMERGED HOLLOW FIBRE MEMBRANE SYSTEM .................................... 19 FIGURE 3-5 SCHEMATIC AND PICTURE OF A MEMBRANE MODULE ................................................................................ 19 FIGURE 3-6 SCHEMATIC OF THE SETUP FOR MEMBRANE A) CLEAN WATER FLUX TEST AND B) INTEGRITY TEST........... 22 FIGURE 3-7 PHOTO OF VIRGIN MEMBRANE HOLLOW FIBRE A) BEFORE AND B) AFTER SOLUBILISATION ....................... 23 FIGURE 3-8 EXAMPLE OF ARRANGEMENT OF MEMBRANE SAMPLES IN SEM-EDX STUB .............................................. 25 FIGURE 4-1 PARTICLE SIZE DISTRIBUTION D10 OF PAC AFTER CENTRIFUGATION (PARTICLE SIZE FOR GENERAL CARBON AND CALGON 208 CP PAC NOT AVAILABLE FOR T PRODUCTS OF 3.1X10 9 , AND 4.5X10 9 ) .................. 26 FIGURE 4-2 PARTICLE SIZE DISTRIBUTION D50 OF PAC AFTER CENTRIFUGATION (PARTICLE SIZE FOR GENERAL CARBON AND CALGON 208 CP PAC NOT AVAILABLE FOR T PRODUCTS OF 3.1X10 9 , AND 4.5X10 9 ) .................. 27 FIGURE 4-3 PARTICLE SIZE DISTRIBUTION D90 OF PAC AFTER CENTRIFUGATION (PARTICLE SIZE FOR GENERAL CARBON AND CALGON 208 CP PAC NOT AVAILABLE FOR T PRODUCTS OF 3.1X10 9 , AND 4.5X10 9 ) .................. 27 FIGURE 4-4 RELATIONSHIP BETWEEN CHANGE IN PARTICLE SIZE (D10) AND PAC HARDNESS IN TERMS OF GOLD NUMBER .............................................................................................................................................................. 28 viii  FIGURE 4-5 PERMEABILITY OF PAC/UF BENCH SCALE SYSTEMS (DATA NOT COLLECTED ON DAYS 7 TO 11AND 22 TO 25 DUE TO MALFUNCTION OF THE DATA RECORDING SYSTEM; CLEANING BACKFLUSH WAS DONE FOR 4 MIN AND 15 MIN ON DAYS 13 AND 29, RESPECTIVELY) ....................................................................................................... 29 FIGURE 4-6 FESEM IMAGES OF VIRGIN AND FOULED MEMBRANE FIBRES .................................................................... 30 FIGURE 4-7 FREQUENCY ANALYSIS OF CARBON TO FLUORINE RATIO IN MEMBRANES .................................................. 31 FIGURE 4-8 PERMEABILITY OF PAC/UF SYSTEM IN WTP (PROVIDED BY GE WATER & PROCESS TECHNOLOGIES) WITH DATA FITTED ............................................................................................................................................... 33 FIGURE 4-9 INORGANIC MATERIAL RELEASED DURING SONICATION OF FOULED AND VIRGIN HOLLOW FIBRES ............. 34 FIGURE 4-10 TOC RELEASED DURING SONICATION OF FOULED AND VIRGIN HOLLOW FIBRES ...................................... 34 FIGURE 4-11 CLEAN WATER FLUX BEFORE AND AFTER ONE HOUR OF SONICATION OF FOULED AND VIRGIN HOLLOW FIBRES (ERROR BARS CORRESPOND TO THE STANDARD DEVIATION OF THE MEASUREMENTS) .............................. 35 FIGURE 4-12 INORGANIC MATERIAL RELEASED FROM SOLUBILISED FOULED AND VIRGIN HOLLOW FIBRES (ERROR BARS CORRESPOND TO THE STANDARD DEVIATION OF THE MEASUREMENTS) ............................................................... 36 FIGURE 4-13 ORGANIC MATERIAL RELEASED FROM SOLUBILISED FOULED AND VIRGIN HOLLOW FIBRES (TABLE G-3) 37  FIGURE C-1 PICTURE OF THE BENCH SCALE AIR SPARGED SYSTEM APPARATUS ........................................................... 46 FIGURE C-2 PICTURE OF THE BENCH-SCALE PAC/UF SYSTEM APPARATUS .................................................................. 46 FIGURE E-1 PICTURE OF CLEAN WATER FLUX TEST APPARATUS ................................................................................... 48 FIGURE F-1 FESEM PICTURE OF A VIRGIN HOLLOW FIBRE, AREA UNDER ANALYSIS = 0.9 ΜM 2  .................................... 49 FIGURE F-2 SCHEMATIC OF PORE PLUGGED BY A PAC PARTICLE (ASSUMPTION) ......................................................... 50  ix  LIST OF ABBREVIATIONS AND SYMBOLS Al Aluminum C Carbon Ca Calcium d10  Size of particles for which 10% of particles are smaller d50 Size of particles for which 50% of particles are smaller d90 Size of particles for which 90% of particles are smaller DBPs Disinfection by-products DMF N,N-Dimethylformamide DIwater Distilled water EDX Energy Dispersive X-Ray EPA Environmental Protection Agency F Flourine FM1 Fouled membrane 1 FM2 Fouled membrane 2 FM3 Fouled membrane 3 Fe Iron FESEM Field Emission Scanning Electron Microscopy FM Fouled Membrane   Velocity gradient GC-MS Gas chromatography-mass spectrometry GE GE Water & Process Technologies gfd Gallon per square feet per day GN Gold Number x  ICP-OES Inductively Coupled Plasma- Optical Emission Spectrometer K  Potasium Ka Fitted parameters of the adsorption model Kb Fitted parameters of the complete model Kc Fitted parameters of the cake model Ki Fitted parameters of the intermediate model L/m Litre per minute Lmh  Litre per metre per hour Mg Magnesium MGD Millions of gallons per day MLD Millions of litres per day MWCO Molecular weight cutoff NaClO Sodium hypochlorite NF Nanofiltration nm Nanometre NOM Natural organic matter Pa Pascal PAC Powdered Activated Carbon P0 Initial trans-membrane pressure ppm Parts per million psi Pound per square inch Pt Trans-membrane pressure at any time PVDF Polyvinylidene fluoride RO Reverse Osmosis r0 Pore radius SEM Scanning Electron Microscopy xi  Si Silica t Time TMP  Trans-membrane pressure TOC Total Organic Carbon UBC The University of British Columbia UF Ultrafiltration µm Micrometre V Specific volume related to membrane area VM Virgin membrane VM Virgin Membrane WTP Water Treatment Plant   xii  ACKNOWLEDGEMENTS I am immensely grateful to my supervisor Dr. Pierre R. Bérubé. His continuous encouragement made me see beyond the obstacles of the project and his contagious passion towards research kept me on track during my studies. All my professors at UBC have significantly influenced my career. I also thank Dr. Eric R. Hall for his suggestions and editing of this thesis. Besides professors, other people have been indispensable to make this work possible. I appreciate the time and patience that Paula Parkinson invested on my inquiries and her support during my several mess- ups in the lab. I thank Tim Ma and Scott Jackson for the dedication they spent fixing lab gadgets for my experiments, Derrick Horne and Bradford Ross for their time and assistance at the Bioimaging Facility- UBC. I have to thanks all those classmates and friends I had the pleasure to share memorable moments at UBC. They supported me in one way or another throughout my studies. I am extremely grateful to Margaret Parsotan for her patience and guidance in my first days in the lab. Her invaluable assistance, friendship and support were vital for the success of this project.  I am very grateful to Kerry Black and Soubhagya Kumar Pattanayak (Pattu) for reading and for their contribution in editing this thesis. I also thank the rest of the UBC membrane group: Syed Abdhullah (Zaki), Sepideh Jankhah, Dongying Ye and Joerg Winter for their support and constructive feedback. Special thanks also to all those other graduate friends, especially Sahar Kosari, Selina Yawson, Chad Novotny and Ryan Thoren for their moral support, friendship and valuable assistance. It has been a very rewarding and fruitful experience. I thank GE Water and Process Technologies for its collaboration and for providing the membranes for this project. I also thank the Natural Sciences and Engineering Research Council of Canada (NSERC) for the fund that made this project possible. Thanks to my beloved family, for the endless support; to Ricardo and Julietta, my two great loves, for their understanding, patience and endless love that inspire me to keep looking forward when life challenges cloud my mind. Thanks to all those here acknowledge to make unforgettable my experience at UBC. 1  CHAPTER 1     INTRODUCTION As drinking water quality regulations are becoming more stringent, development and improvement of treatment technologies are required. Ultrafiltration (UF) membranes are increasingly used for drinking water treatment due to their ability to effectively remove particulate contaminants such as microorganisms.  However, UF membranes alone cannot effectively remove soluble contaminants such as some disinfection by-product (DBP) precursors. On the other hand, activated carbon can be used to effectively remove soluble contaminants such as some disinfection by-product (DBP) precursors. Activated carbon is commonly used in form of Granular Activated Carbon (GAC) and Powdered Activated Carbon (PAC) for production of drinking water. When UF and PAC are used simultaneously, the characteristics of both systems complement each other, achieving stringent drinking water quality guidelines. This hybrid system is commonly known as PAC/UF system. It combines the adsorption capabilities of PAC and the removal of microorganisms and particles of the UF membranes. Although PAC has been extensively used in combination with UF membranes to reduce taste, odour, colour and other concerns caused by organics in drinking water, there are conflicting results regarding the effect of PAC on the membrane fouling. Some authors have reported the advantages of PAC/UF systems, such as improvement of permeate flux, longer filtration runs or reduced frequency of chemical washing (Adham, et al., 1991; Lee et al., 2007; Campinas and Rosa, 2010). Others have shown no effect or minimal effects (Yiantsios and Karabelas, 2001; Tomaszewska and Mozia, 2002; Matsui et al., 2006). Some other researchers have reported that PAC does have adverse effects on the efficiency of membranes performance (Lin et al., 1999; Zhang et al., 2003; Li and Chen, 2004; Seo et al., 2005; Saravia and Frimmel, 2008). Effects on membrane fibre integrity and efficiency of the treatment may potentially shorten the lifetime of the membrane, resulting in a significant increase of maintenance and operational costs.   2  CHAPTER 2     BACKGROUND AND LITERATURE REVIEW 2.1. Powdered Activated Carbon Powdered Activated Carbon (PAC) is extensively used in drinking water treatment to reduce the concentration of organics and odour, to improve taste and to treat low concentrations of pesticides (Crittenden et al., 2005). Any organic material with a high content of carbon, such as coal, wood, or coconut shell can be used as primary raw material to generate activated carbon. The raw material is ground and crushed to the correct size, and it is converted to activated carbon by thermal decomposition in a furnace. Activated carbon has a large surface area per unit volume, and a network of pores, whose walls provide an attractive surface where molecules can be adsorbed. Important physical properties of activated carbon are surface area, density, hardness, abrasiveness, particle size (mesh size), abrasion resistance and ash content. Particle size is especially important in water treatment as it affects conditions in membrane systems such as trans-membrane pressure (TMP) (Matsui et al., 2009), filtration capabilities, the rate of removal of organic matter (Matsui et al., 2006), and the characteristics of the fouling cake layer (Clark et al., 1996; Zhao et al., 2005). Abrasion resistance is also very important, it refers to the structural strength of the carbon and its ability to resist degradation during handling and is expressed in terms of an abrasion number. 2.1.1. Particle size of PAC The particle size of PAC is typically smaller than 0.297 mm (50-mesh sieve). The smaller the particle size of activated carbon, the better the access to the surface area (i.e., higher surface to volume ratio) and the faster the rate of adsorption kinetics. Careful consideration of particle size distribution can provide significant benefits in the operation and outcomes of a system. When PAC is used in conjunction with membranes, its particle size should be larger than the pore size of the membrane. Saravia et al. (2006) reported that a narrow particle size distribution and an average particle size of approximately 100 times larger than that of the membrane pore size helps to prevent membrane blocking. However, during its use, PAC characteristics can be modified by external factors, which may compromise the integrity and operation of the membrane system, caused by PAC scouring the membrane surface or plugging the membrane pores. 3  2.1.2. Abrasiveness of PAC The abrasiveness of materials, in form of slurry, can be expressed in terms of a parameter known as Gold Number (GN). Slurry abrasiveness is a complex relationship between chemical and mechanical action on a wear specimen. Mechanical wear relates to the properties of the solids and fluid as well as the solid concentration and load applied to the wear block or wear specimen. Particle mineral composition, hardness, size, shape, and friability are the main contributing wear factors for the solid components of the slurry. Rate of wear increases as hardness of the particle tested increases. The size of the particles in the slurry has a major affect on the degree of wear, similar to the action of sandpaper of different grits; hence larger particles tend to generate greater wear. Particle shapes ranging from spherical to sharp and angular also determine the degree of wear. The GN is a relative rate of wear index of the mechanical erosion on a 24K gold wear block; it is based on the ASTM G075 test procedure. A lower GN correlates to a less abrasive material (Miller, 2011). In a PAC/UF system, the continuous physical contact of PAC with the membrane can wear the membrane surface and may compromise the integrity of the fibres. Thus, GE Water & Process Technologies (2008) recommends PAC brands with low GNs, i.e. less abrasive. However, low abrasive PAC may be more susceptible to fracture into smaller particles that are similar in size to those of membrane pores, thus, able to enter the pores and plug them. 2.1.3. Abrasion resistance of PAC Abrasion resistance refers to the ability of carbon to resist degradation during handling. In the present study, abrasion resistance of PAC is also correlated to the susceptibility of the particles to break or fracture due to an external stimulus. Abrasion resistance will also be referred to as hardness. 2.2. Membrane filtration Membrane filtration is defined as a separation technique driven by pressure or vacuum. The contaminants are rejected by a barrier through a size exclusion mechanism, resulting in the production of product (permeate) and waste streams (retentate), as illustrated in Figure 2-1. Membranes are typically made of synthetic material less than 1 mm thick and act as selective barriers, where permeable material passes through the barrier and the impermeable material is retained on the feed side (Crittenden et al., 2005). Membranes are classified based on the type of material they reject, their operating pressure, and their nominal (i.e., average) pore size. 4   Figure 2-1 Schematic of membrane filtration process Microfiltration (MF) and ultrafiltration (UF) membranes are low pressure systems, which are able to separate suspended or colloidal particles from a liquid phase by passage of the fluid through a porous medium. MF membranes typically have a pore size of 0.1 to 0.2 µm. Pore size ranges of UF membranes are typically between 0.01 and 0.05 µm or less. Nanofiltration (NF) and Reverse Osmosis (RO) membranes are high pressure systems, characterized for their ability to remove dissolved solutes (such as ions and dissolved natural organic matter) by diffusion of water in response to a concentration gradient. Molecular weight cutoff (MWCO) for these types of membranes can range from 200 to 1000 Daltons for NF membranes and less than 100 Daltons for RO membranes (EPA, 2005). A summary of the classification of membranes processes is presented in Figure 2-2.  Figure 2-2 Classification and removal efficiencies of membrane processes (Adapted from EPA, 2005 and Crittenden et al., 2005) UF membrane systems have been in use for several years for drinking water treatment; their ability to remove particles, turbidity and pathogens has been well documented. However, several studies have Membrane Feed stream Permeate Retentate Size (µm) Water Components Membrane Filtration Process Molecular weight (Daltons) 100 200 20000 2000000 0.0001 0.001 0.01 0.1 1 10 Bacteria, algae, protozoa Giardia Viruses Cryptosporidium Nanofiltration (NF) Ultrafiltration (UF) Microfiltration (MF) Reverse Osmosis (RO) Reverse Osmosis Membrane filtration Monovalent species (Na+) Divalent ions (Ca2+) 5  demonstrated that most UF membranes are not capable of effectively remove dissolved organic matter (such as humic substances), taste- and odour- producing compounds, synthetic organic compounds, and pesticides (Clark et al., 1996).  Removal of dissolved organic matter is especially important given that it can contribute to the formation of disinfection by-products (DBPs) (Crittenden et al., 2005), some of which are suspected to be carcinogenic and are currently regulated by many jurisdictions (USEPA, 1998; Health Canada, 2010). Consequently every effort to eliminate organic matter can help to reduce the potential formation of disinfection by-products (Croue et al., 1999). UF systems by their own are not necessarily capable of meeting more stringent drinking water guidelines, especially when considering DBP regulations. Hybrid systems that combine UF membranes with other treatment technologies may be able to meet the increasingly demanding drinking water quality guidelines. PAC/UF systems are commonly utilized for this purpose. 2.3. Membrane fouling Membrane fouling is the gradual accumulation of material on a membrane surface or within a porous membrane structure which inhibits the passage of water, increasing the resistance to the permeate flow (EPA, 2005). Fouling can generally be classified into two categories: reversible fouling and irreversible fouling. Reversible fouling is the accumulation of material that can be controlled or removed through a backwashing and/or cleaning process. Irreversible fouling refers to any membrane fouling that is permanent and cannot be removed. 2.3.1. Mechanisms of membrane fouling In order to obtain insight on the causes of fouling in UF/PAC systems, it is necessary to identify the occurring fouling mechanism in the investigated system. Models can be used as tools to investigate the mechanism of fouling, which reduces the flux affecting the overall performance of the filtration process. Several studies have been conducted to attempt to explain the mechanisms that occur during membrane fouling. The four classical fouling models proposed by Hermans and Bredée (1935) (showed in Figure 2-3 ) are well accepted and broadly used to describe the interaction between foulants and membrane, and can help to predict membrane behaviour. 6   Figure 2-3 Schematic of the four classical fouling models Cake fouling assumes that the material that accumulates at a membrane surface forms a porous structure which increases the resistance to the permeate flow. The increasing thickness of the cake adds greater resistance to the flow. The adsorption blocking model assumes that the material is adsorbed on the walls of the pores, reducing the effective diameter and increasing the resistance to the permeate flow. Complete blocking assumes that material reaching the membrane can block the membrane pores, preventing water flow. Intermediate blocking, on the other hand, assumes that only a portion of the material reaching the membrane can block the membrane pores. These models have been extensively studied and used individually; however the combination of several fouling mechanisms occurring in parallel has been recently studied (Bolton et al., 2006; Katsoufidou et al., 2005). Equations describing the four classical models are summarized in Table 2-1. Ka, Kb, Kc, and Ki are the fitted parameters of the adsorption model, the complete model, the cake model, and the intermediate model, respectively. The trans-membrane pressure at any time (Pt) is described as a function of P0 (initial trans-membrane pressure), r0 (pore radius), V (specific volume related to membrane area) and the respective K (constant).         Cake fouling      Adsorption blocking               Complete blocking              Intermediate blocking 7  Table 2-1  Summary of constant flow fouling models (Adapted from Hermans and Bredée, 1935)  Model Equation Cake    Adsorption     Complete    Intermediate  2.4. Powdered Activated Carbon/ Ultrafiltration system Powdered Activated Carbon/Ultrafiltration (PAC/UF) systems integrate the adsorption capabilities of PAC and the particulate material removal capability of UF membrane systems. The system usually consists of a single tank, where the membrane unit is submerged together with the PAC. An aeration system keeps PAC in suspension. The compact configuration, the simplicity of operation, the low fouling potential, and the high quality of the treated effluent have contributed to the world wide application of this type of PAC/UF system for water treatment (Adham et al., 1996). Although PAC/UF systems have been used extensively, there are contradictory results regarding the effect of PAC on membrane fouling. Some authors have reported beneficial effects of the addition of PAC to UF systems, such as longer filtration runs, reduction of frequency of chemical washing, reduction of fouling layer, improvement of permeate flux, and even retardation of the membrane fouling (Adham, et al., 1991; Jacangelo et al., 1995; Clark et al., 1996; Kim et al., 1996; Saravia et al., 2006; Lee et al., 2007; Campinas and Rosa, 2010). Others have reported no effect or minimal effects of the addition of PAC to UF systems (Yiantsios and Karabelas, 2001; Tomaszewska and Mozia, 2002; Matsui et al., 2006). Other authors, on the other hand, have reported adverse effects of the addition of PAC to UF systems on the efficiency of membrane performance (Lin et al., 1999; Li and Chen, 2004; Zhang et al., 2003; Seo et al., 2005; Saravia and Frimmel, 2008). 8  2.4.1 Fouling in PAC/UF systems Fouling in PAC/UF systems has been studied by several researchers, but no consistent results have been achieved. The inconsistencies may be due to the type and the amount of PAC added (Jacangelo et al., 1995; Zhang et al., 2003; Matsui et al., 2006), the type of natural organic matter (NOM) used as well as the DOC concentration (Chang et al., 1998; Yiantsios and Karabelas, 2001), the solutions chemistry (Braghetta et al., 1998), pH (Chang et al., 1998), and characteristics of the membrane used in the different studies. Some studies have reported the presence of PAC particles deposited on the UF membrane surface (Lin et al., 1999; Yiantsios and Karabelas, 2001). The deposition of PAC was observed to be reduced in the presence of organic matter. This has been attributed to electrostatic and steric stabilization effects and reduced adhesion energy imparted on the particles and the membranes by the absorbed organic molecules (Yiantsios and Karabelas, 2001). Although PAC was observed to accumulate at the membrane surface, Saravia and Frimmel (2008) reported that the PAC layer built-up did not substantially increase the resistance to the permeate flow. Some researchers have reported that PAC alone does not affect the filtration performance in a PAC/UF system, but rather, PAC combined with organic and/or inorganic material affects the resistance to the permeate flow. Lin et al. (1999), Yiantsios and Karabelas (2001), Zhang et al. (2003), and Saravia and Frimmel (2008) attributed the increase in resistance to the permeate flow in PAC/UF systems to the combination of PAC and the organic material in the raw water source. They suggested that the NOM can bond strongly and simultaneously with both the PAC particles and the membrane surface and can thereby form a layer with high resistance to permeate flow. However, the removal of NOM from solution prior the addition of PAC does not necessarily reduce fouling as reported by Zhang et al. (2003). On the other hand, Zhao et al. (2005) attributed the increase in the resistance to the permeate flow in PAC/UF systems to the combination of PAC and inorganic material. Oh et al. (2006) observed that both organic and inorganic material contribute to the formation of a complex cake structure on the surface of membranes in a PAC/UF system. Some inorganic elements such as Si, Ca (Baker and Dudley, 1998; Plottu-Pecheux et al., 2003; Mosqueda-Jimenez et al., 2008; Arnal et al., 2009), Mg, Fe, Al, K, and S (Arnal et al., 2009) have been observed in the membrane foulant layer in PAC/UF systems. Zhao et al. (2005) suggested that metal ions can neutralize the charge on PAC particles, enabling them to form a more compact structure. Inorganic material may originate from the water source (Arnal et al., 2009) or from the inadequate use (dosage) of coagulant (Lee et al., 2007). 9  2.5. Conclusions Based on the literature review presented above, it can be concluded that the use of PAC in conjunction with membrane ultrafiltration has beneficial effects on fouling control. However, some studies have reported that PAC can also have negative effects on membrane fouling; therefore more investigation on the origin and the mechanisms governing the irreversible fouling in UF/PAC systems is required. Some studies have suggested that organic and/or inorganic material in the raw water can combine with PAC and increase its tendency to adhere to a membrane surface. It is also possible that fouling in UF/PAC systems is due to pore plugging by PAC particles, especially if these can break up into smaller particles in the process. 2.6. Research objectives The objective of this study was to investigate the origin and the mechanisms governing the irreversible fouling in UF/PAC systems and to understand the role of PAC in UF/PAC systems. An assessment of the potential causes of irreversible fouling in a UF membrane system for which PAC addition had a negative effect was done. As part of the present screening study, samples of hollow fibre membranes from a fouled PAC/UF water treatment system were examined and compared with new hollow fibre membranes that had not been used for filtration. The first part of the study assessed the contribution of PAC particles in the irreversible fouling in PAC/UF systems and whether the irreversible fouling is potentially due to the abrasion of PAC particles, which are subjected to the relatively turbulent conditions present in a submerged hollow fibre membrane system. The second part of the study explored the mechanism of fouling and the detection of foulants in an irreversibly fouled membrane of a PAC/UF membrane system. The sections of the thesis which describe the outcomes of the different parts of the study are summarized in Figure 2-4. Physical characteristics and composition of fouled (FM) and virgin (VM) hollow fibres were thoroughly analyzed in order to provide insight on the possible sources of membrane fouling. 10     Figure 2-4 Arrangement of outcomes of the present study  Section 4.1.1. •Assessment of PAC particle size over time Section 4.1.2. •Assessment of PAC pore plugging potential over time . •. Section 4.2.1. •Detection of PAC in irreversible fouled membranes Section 4.2.2. •Assessment of mechanism of fouling in irreversibly fouled membranes Section 4.2.3. •Detection of foulants in irreversibly fouled membranes Part 1 (Section 4.1). Study of membrane plugging due to abrasion of PAC particle Part 2 (Section 4.2). Detection of foulants in PAC/UF hollow fibres 11  CHAPTER 3      MATERIALS AND METHODS The present study was developed to investigate the cause of irreversible fouling in a submerged hollow fibre membrane system with PAC addition (PAC/UF system). Virgin and fouled membrane fibres obtained from the PAC/UF system described in Section 3.1 were subjected to various tests, described in Section 3.3. 3.1.PAC/UF systems under investigation The present study was done in collaboration with GE Water & Process Technologies (GE), who contributed information and samples of membrane fibres. Prior to the present study, GE conducted a series of tests investigating three ZeeWeed® UF membrane systems, including a full scale water filtration plant (Figure 3-1). The first part of the tests compared the performance of the full scale PAC/UF system (Unit A) to that of a UF system without PAC addition (Unit C) in order to determine the effect of PAC on the membranes. Operating conditions of Unit A are summarized in APPENDIX B    .  Figure 3-1 Schematic of systems investigated by GE Water & Process Technologies Raw water PAC additionUnit A. Full scale PAC/UF system Unit B. Pilot scale PAC/UF system Unit C. Mobile UF system Membrane fibres from Unit C used in Unit B 12  Unit A presented extensive irreversible fouling. The three chemical cleanings recovered the permeability to approximately 80%, 60%, and 52% of the initial flux, respectively (Figure 3-2). Overall, permeability decreased by approximately 40% over a 7 month period, from 8.5 Lmh (5 gfd/psi) to 3.4 Lmh (2 gfd/psi). Extensive hydraulic and chemical cleaning was not effective at reducing the extent of irreversible fouling. Unit C, on the other hand, presented limited signs of fouling over a parallel 12 month period. The results from the first part of the tests performed by GE, suggested that PAC had a significant negative effect on the UF system, causing low recovery of permeability and thus increasing irreversible membrane fouling.  Figure 3-2 Permeability of PAC/UF system, Unit A (Provided by GE Water & Process Technologies) The second part of the test by GE compared the performance of the full scale PAC/UF system (Unit A) to a pilot- scale PAC/UF system that contained relatively new membranes (Unit B). The membranes for Unit B were obtained from Unit C. This enabled the effect of membrane age/history on fouling to be considered. Unit B was shut down after approximately 6 months of operation due to extensive loss in permeability, even following extensive maintenance cleaning and recovery cleaning. All of the irreversibly fouled membranes used in the present research thesis were harvested from Unit B, following chemical cleaning. The characteristics of the units and the outcomes of the studies conducted by GE are summarized in Table 3-1. 0 1 2 3 4 5 6 7 0 30 60 90 120 150 180 210 240 P er m ea b il it y , g fd /p si Time, d System shut down PAC concentration increased from 1000 to 4000 mg/L Chemical cleaning 13  Table 3-1 Characteristics of the units investigated by GE Water & Process Technologies  Unit A. Full scale PAC/UF system Unit B: Pilot scale PAC/UF system Unit C: Mobile UF system System type ZeeWeed® 500C ZeeWeed® 500D ZeeWeed® 500D Membrane chemistry CP5 CP5 CP5 Total flow 4 MGD (15.14 MLD) 1 MGD (3.78 MLD) 1 MGD (3.78 MLD) Fouling characteristics Irreversible fouling in 12-18 months Irreversible fouling in 7 months No fouling in 12 months CP5: nomenclature used by GE to identify the membrane chemistry manufactured by the company 3.2. Materials The characteristics of the hollow fibre membranes and powdered activated carbon used during the present study are described in the following sections. 3.2.1. Hollow fibres membranes All membranes used in the present study were polyvinylidene fluoride (PVDF) membranes. Both irreversibly fouled and virgin membrane fibres were considered. Irreversibly fouled membranes (FM) were those that had been fouled in the PAC/UF system described in Section 3.1, in which PAC addition had a significant negative effect on membrane fouling. FM fibres were harvested from the system following hydraulic and chemical cleaning with sodium hydroxide and citric acid. Virgin membranes (VM) were new fibres that had not been used for filtration. All membrane fibres were shipped to the UBC Environmental Engineering Laboratory in sealed Ziploc® plastic bags. At UBC the membranes were stored in their packaging at 4°C until used. The physical characteristics of the membranes used in this study are presented in Table 3-2. 14  Table 3-2  Physical characteristics of the membrane fibres used in the present study Configuration Chemistry Outer surface diameter Surface properties Nominal pore diameter Typical operating pressure Outside-in hollow fibre CP5 1.77 mm Non-ionic and hydrophilic 0.04 µm 7-55 kPa (1-8 psi) The different FM fibres used in the present study were harvested at different times as presented in Table 3-3. Unfortunately, other than the age of the membrane at harvesting, no other information was available with respect to the exact operating conditions to which they were exposed. Table 3-3 Nomenclature and background of membrane fibres ID Description of fibres VM FM1 FM2 FM3 Virgin Membrane Fibre Fouled Membrane Fibre #1 Fouled Membrane Fibre #2 Fouled Membrane Fibre #3 New membrane fibres, original conditions. Substantially fouled membrane fibres. Age unknown Substantially fouled membrane fibres.  <1 year old Substantially fouled membrane fibres. 2 years old 3.2.2. Powdered Activated Carbon The characteristics of the different types of PAC used in the present study are presented in Table 3-4. All information was provided by the respective manufacturers, except for the abrasion resistance or hardness. This characteristic was based on Gold Number (GN), a parameter provided by GE Water & Process Technologies.  As previously discussed, a higher GN refers to a harder PAC.  The PACs mentioned in Table 3-4 are sorted from lowest to highest in terms of hardness. 15  Table 3-4 Physical characteristics of PAC tested No Brand Manufacturer Carbon source Mean particle size, µm Gold Number 1 Aqua Nuchar MeadWestvaco Wood 27.5 0.11 2 Norit SA UF Norit NA 13.6 0.45 3 Calgon 208CP Calgon Coconut shell 20.4 0.63 4 General Carbon (GCWDC) General Carbon Wood 29.4 0.74 5 Calgon WPH Calgon Bituminous coal 34.5 2.31 3.3.Methods Various experiments were designed and performed to assess the potential causes of irreversible fouling in a system for which PAC addition had a negative effect. A description of the experimental set-ups and analytical procedures used in the present study are presented in the following sections. 3.3.1. Study of membrane plugging due to breakdown of PAC particle The diameter of PAC particles is usually larger than that of the membrane pores. However, it is possible that the relatively turbulent conditions present in a submerged hollow fibre membrane system can break down the PAC into smaller particles that could potentially plug the membrane pores. The experiments were performed using PAC in a solution of distilled deionized water (MilliQ water, filtered through an ion exchange cartridge and 0.22 µm filter, with minimum resistance of 18.2 MΩ cm-1) to ensure that potential contributions of other material to PAC breakdown or membrane fouling were eliminated. 3.3.1.1. Assessment of the change of PAC particle size over time The objective of this analysis was to expose the PAC to relatively high shear conditions typical of submerged hollow fibre membrane systems and to determine if the size of PAC particles decreased with time. Changes in particle size distribution were assessed for all PACs listed in Table 3-4. The experiments were carried out in five air sparged systems operated in parallel, one for each PAC investigated (Figure 3-3). The system tank consisted of 1L- volumetric cylinder filled with 1 L of distilled water (DIwater) with PAC to achieve a concentration of 2 g/L.  Air sparging was applied with a stone diffuser (2 cm diameter and 2.5 cm long) located at the bottom of the cylinder. The air sparging flow was kept at 2 16  L/min at all times using an air flowmeter (Cole Parmer). Periodically the system tank was manually topped up to the original level with DIwater to compensate for water evaporation (ca. 2 mL/d) and to maintain a PAC concentration of 2 g/L.  Figure 3-3 Schematic of a bench scale air sparged system  liquots were collecte  from the s stem tan  at  ifferent times, an  the particle si e  istri ution of the aliquots was  etermine  imme iatel .    t pro uct was use  to reflect  oth the effect of time an tur ulence intensit  on the    .   is presente  in  quation (3-1).            (3-1 ), where   is the velocity gradient (s-1), Q is the air flow (m3/s), H is the depth of water (m), γ is the specific weight of water (N/m 3 ), V is the volume of system (m 3 ), and µ represents the viscosity of the liquid (Pa∙s .  liquots of the     solution were anal  e  at a  t pro uct of 0, 0.5x109, 3.1x109, and 4.5x109.  ignificantl  higher  t than that use  in the pilot scale        s stem   t = 4.2x104) was considered to accentuate the potential effect of time and turbulence intensity on the PAC. When considering a solution containing a range of particle sizes, the presence of smaller particles can be obscured by larger ones. Since the small particles of PAC were of main interest, this fraction was A Pressure transducer Membrane fibres Air Flowmeter Computer Air Diffuser Data logger PAC in suspension 1 L-cylinder Air Flowmeter Air Diffuser PAC in suspension 1 L-cylinder Compressed Air Compressed Air Scale B 17  separated from the rest of the aliquot by centrifugation prior to particle size analysis. Aliquots were placed into 20 mL-tubes and centrifuged at 2000 relative centrifugal force for 15 minutes. The supernatant of the aliquot which contained the smaller particles was then collected and analysed. The particle size distributions were determined using a laser particle size analyzer (Mastersizer Hydro 2000S, Malvern M).  Enough volume of the aliquot was introduced in the sampler tank of the instrument to reach a minimum obscuration rate of 1% to 3%. One drop of a dispersant (0.5% PCC-54, Thermo Scientific) was added to the sampler tank with every aliquot analyzed to wet the material, scatter the particles and prevent them from sticking to internal walls, cells and tubings of the instrument. Three replicate measurements per aliquot were performed and the results were averaged. In order to obtain a satisfactory background (determined by the manufacturer of the instrument), the sampler tank of the particle size analyzer was emptied after each run and filled again with DIwater and 2 mL of 0.5% of PCC- 54. This solution was pumped and circulated in the instrument for 5 min. Three subsequent cycles of internal cleaning were done before the next aliquot to ensure the removal of all material from the previous aliquot. Number distribution was selected for this study due to its better responsiveness to small particles. The values considered for the analysis for the number distribution were d10, d50, and d90 as defined in Table 3-5. At least two replicate measurements of particle size distribution per sample were done. All glassware was previously washed, rinsed with MilliQ water and baked at 400°C for 1 hour prior analysis. Every plastic container used was washed with soap, rinsed with 0.1 nitric acid and with MilliQ water prior use. All experiments were conducted at room temperature (20 o C ±2 o C). Table 3-5 Size distribution parameters Parameter Definition d10 Size of particles for which 10% of particles are smaller d50 Size of particles for which 50% of particles are smaller d90 Size of particles for which 90% of particles are smaller 3.3.1.2. Assessment of PAC pore plugging potential over time The objective of the experiment was to investigate the extent of the influence of PAC on irreversible fouling in a PAC/UF system. The experiments were carried out in three PAC/UF bench scale systems operated in parallel, one for each PAC investigated (Figure 3-4). 18  The system consisted of a system tank and an air sparger, a hollow fibre filtration system and a pressure monitoring system. The system tank and the air sparger system were similar to that described in Section 3.3.1.1. The hollow fibre filtration system consisted of a set of three virgin hollow fibre membranes (VM), similar to those used in the pilot scale PAC/UF system described in Section 3.1. The total operational length of each fibre was 30 cm, i.e. each module was 90 cm long (working length) making a total membrane surface area of 0.0113 m 2 per module. The hollow fibre was potted with Epoxy® to a piece of tubing (4 mm ID). The top end of the fibre was left open to allow permeate flow. The bottom end of the hollow fibre was sealed with Epoxy® to form a close ended membrane (Figure 3-5). The hollow fibre was immersed in the system tank and kept straight by a light weight attached to the bottom of the hollow fibre. Characteristics of the hollow fibre used are described in Section 3.2.1. The system was operated with a constant flux and variable pressure. A peristaltic pump (Masterflex L/S, model number 7524-50) provided the driving force to draw the liquid by suction. The flow was set at 6 mL/min (32 L/m 2∙hr), which was monitored to verify that the flow remained constant throughout the experiment. The permeate was recycled back into the system tank to maintain a constant system liquid volume. Periodically, the system tank was manually topped up to the original level with DIwater to compensate for losses due to evaporation. 19   Figure 3-4 Schematic of a bench scale submerged hollow fibre membrane system   Figure 3-5 Schematic and picture of a membrane module The pressure monitoring system logged the trans-membrane pressure throughout the duration of the experiment. A pressure transducer (Omega PX243A) connected to a data logger (National Instrument USB-6009) collected pressure measurements at a rate of 1 Hz.  The collected pressure measurements were recorded every ten minutes by a custom Labview application (Labview Version 7.0). The selection of the three different PACs for this experiment was based on their level of hardness, according to Table 3-4. PACs with a low, intermediate, and high hardness were selected. PAC was added A Pressure transducer Membrane fibres Air Flowmeter Computer Air Diffuser Data logger PAC in suspension 1 L-cylinder Air Flowmeter Air Diffuser PAC in suspension 1 L-cylinder Compressed Air Compressed Air Scale B Epoxy Hollow fibres Tubing 20  to the DIwater to achieve a concentration of 300 mg/L. The experiments were continuously conducted up to a  t product of 4.5x109. Membrane backflush (with MilliQ water) was only done when trans- membrane pressure (TMP) was observed to increase by 30% of the initial flux.  All experiments were conducted at room temperature (20 o C±2 o C). 3.3.2. Detection of foulants in PAC/UF hollow fibres A number of destructive and non-destructive techniques were used to detect PAC on/in the hollow fibres obtained from an irreversibly fouled pilot scale PAC/UF system (described in Section 3.1). Field Emission Scanning Electron Microscopy (FESEM) and Scanning Electron Microscopy Energy- Dispersive X-Ray (SEX-EDX) were performed to obtain insight on the characteristics of the membrane surface of the hollow fibres in a qualitative and semi-quantitative approach. Sonication and a novel solubilisation technique were used to analyze the foulants in/on the membrane surface of the hollow fibres using a quantitative approach. All techniques were applied to both virgin and fouled membrane fibres, so that the results for each could be compared. 3.3.2.1. Extraction of foulants from fouled membrane fibres Two techniques were used to extract the foulants from the membrane fibres harvested from the irreversibly fouled PAC/UF system under study. 3.3.2.1.1. Extraction of foulants by sonication The extraction of foulants accumulated in the irreversible fouled fibre from the fouled PAC/UF system was undertaken in order to gain insight into the characteristics of the material present on the irreversibly fouled fibre surface. Membrane fibres were subjected to sonication in order to extract the foulants accumulated in FM1, FM2, and FM3 hollow fibres (defined in Table 3-3). The foulants were extracted by sonication and subsequently analyzed for total organic carbon (TOC) and metal content. For comparison, VM fibres were also sonicated and any material released analyzed for TOC and metal content. Prior to sonication, VM fibres were rinsed with MilliQ water and soaked overnight in a 750 ppm sodium hypochorite (NaClO) solution, and FM fibres were soaked overnight in MilliQ water.  The 170 cm-long fibres were cut into small segments (~5 cm) and placed in an Erlenmeyer flask with DIwater (100 mL). The flasks containing fibre segments were then subjected to sonication (Aquasonic Model 550HT, VWR) at 40°C for 1 hour. An aliquot (15 mL) of the solution (i.e. extract) was collected from the solutions in the 21  sonicated flasks and the fibre samples were once again sonicated for 1 more hour, after which a second aliquot of the extract was collected. The aliquots were analyzed for both TOC and metal content. TOC analyses were based on 5310 B Standard Method ―High Temperature Combustion Method‖ (APHA et al., 2005). The aliquots were placed into glass vials, acidified to ~pH 3 with 10% HCl and analyzed in a TOC analyzer (IL 550 TOC-TN, Lachat Instruments). All glassware was washed with water and soap, rinsed with MilliQ water, and baked at 400°C for 1 hour. Blank samples and laboratory-made reference standards with known TOC concentrations were also analyzed. Metal content was analyzed through Inductively Coupled Plasma – Optical Emission Spectrometry (ICP-OES). The aliquots were placed into plastic test tubes and analyzed in an ICP- optical emission spectrometer (Perkin Elmer Optima 7300DV). All plasticware was rinsed with a nitric acid diluted 10 times (in DI water) and soaked in DIwater until use. DIwater and laboratory-made reference standards with known metal concentrations were also analyzed. Clean water flux tests were performed on the hollow fibres before and after sonication in order to determine if the resistance to the permeate flow decreased when the foulants were extracted during the sonication. The clean water flux test was carried out in a filtration system as illustrated in Figure 3-6A. The setup consisted of a system tank, a pressure gauge and a hollow fibre filtration system. The tank and the hollow fibre filtration systems are similar to the ones described in Section 3.3.1, except that only one hollow fibre was used.  The pressure and the flow were monitored during the experiment. During this flux test, MilliQ water was filtered through the hollow fibre at a fixed pressure of 39.3 kPa (5.7 psi) for a period of 5 minutes. The permeate flow was measured by weighing the volume of permeate collected. The system tank was periodically topped up to the original level with DIwater to maintain a constant liquid level in the system. 22   Figure 3-6 Schematic of the setup for membrane A) clean water flux test and B) integrity test Integrity tests were performed on representative samples of sonicated FM and VM in order to confirm that the fibres were not significantly damaged during sonication. The integrity test was carried out in a setup similar to the one used for the clean water tests, except that instead of negative pressure (i.e., vacuum), the system was subjected to positive pressure as illustrated in Figure 3-6 B. The hollow fibre was pressurized to negative testing pressure of 41.37 kPa (-6 psi) and isolated by closing the flow control valve for two minutes. Pressure was monitored during that time. The hollow fibre was considered to be un-breached if the pressure decline was less than 10% of the testing pressure. 3.3.2.1.2 Extraction of foulants by membrane solubilisation Although sonication proved to be effective at removing material from the membrane surface, the results suggested that not all of the material that caused fouling could be removed. To overcome this limitation, a new foulant extraction procedure was developed, where, by instead of removing foulants from the membrane fibre, the membrane fibre was removed from the foulants, and foulants present could be analyzed for organic and metal content. VM, FM1, FM2 and FM3 hollow fibres were used in this experiment. The total length for each fibre used was 190 cm. The hollow fibres were previously rinsed with MilliQ water overnight to remove any glycerine and grit from the virgin and fouled membranes, respectively. Additionally virgin hollow fibres were soaked in a 750 ppm solution of sodium hypochlorite (NaClO) (diluted from 6.0% Domestic Pressure gauge Vacuum Scale Membrane A Pressure gauge Compressed Air Membrane B 23  Miroclean Bleach) and rinsed with MilliQ water. Hollow fibre membranes were then cut into small pieces (~ 5 cm) and placed in an Erlenmeyer flask with 200 mL of N,N-Dimethylformamide (DMF, Acros). DMF is a strongly polar solvent that works very well as a solvent for PVDF (Brandrup et al., 1999). The solution was then placed in a hot water bath at 50-60°C for 20 minutes. It should be noted that the method was designed to solubilise the membrane components of the fibre, and not the internal support (see Figure 3-7).  Therefore, the internal support of the hollow fibre was intact and removed from the solution prior to analysis. Aliquots from the solubilised hollow fibres were analyzed for metal content and other organic compounds.  Figure 3-7 Photo of virgin membrane hollow fibre A) before and B) after solubilisation Inductively Coupled Plasma- Optical Emission Spectrometer (ICP-OES) analysis was carried out in order to analyze the extracted foulants for organic compounds. The aliquots were first diluted by 50% in 0.1% of nitric acid and centrifuged in order to remove the solid residues that might obstruct the tubes of the instrument. Organic compounds were analyzed through gas chromatography-mass spectrometry (GC- MS). Aliquots were placed into glass vials and analyzed in a GC-mass spectrometer (Agilent Technologies 5973). DMF was analyzed as a blank and laboratory-made reference standards with known metal concentrations were also analyzed in an ICP-OES as described in Section 3.3.2.2.1. All aliquots were analyzed in duplicate. 3.3.2.2. Analysis of membrane hollow fibre surface High resolution images of the surface of the hollow fibres from a fouled PAC/UF system were obtained in order to gain insight into the mechanism and cause of fouling. Membrane Internal support Membrane Internal support A B 24  3.3.2.2.1. Scanning Electron Microscope analysis Scanning Electron Microscopy (SEM) (Hitachi S-2600N) was used in conjunction with Energy Dispersive X-ray analysis (EDX).  The data acquisition was done by coupling the X-ray detector to a Quartz Imaging Systems XOne software. The elements that made up the membrane fibres were identified through full spectrum EDX. SEM-EDX analysis was exclusively done on VM and FM1 hollow fibres. FM fibres were rinsed with MilliQ water to wash off any particles loosely attached to the surface of the membranes. VM fibres were rinsed and soaked in MilliQ water overnight to eliminate any glycerine and other substances used by the manufacturer to protect the membranes until use. All fibres were air- dried prior to the analysis. Since carbon and fluorine are common elements found in fibre membranes, the ratio carbon-fluorine (C-F) was used as a parameter to detect the difference in carbon content between FM and VM fibres. It was assumed that the content of fluorine in all membranes analyzed was constant, given that fluorine is one of the main components of PVDF membranes. If PAC was present at substantial levels on the irreversible fouled fibres, the C-F ratio was expected to be higher than that in the virgin fibres tested. Sections of membrane fibres were selected randomly and cut into small pieces (3 to 5 mm long) with a PTFE-treated stainless steel razor blade.  Longer pieces (~ 5 mm long) were placed on the instrument mount (surface view), while shorter pieces (~ 3 mm long) were placed with the cross section facing up (cross section view), as depicted in Figure 3-8.  The membranes fibres were sputtered with 10 nm of chromium using a High Resolution Sputter Coater (Cressington 208HR). This was done prior to the analysis to enhance the image contrast and reduce any potential charging of the sample under the electron beam. The microscope was at all times aimed to the external section of the fibres, which was the main area of interest and where the foulants were most likely to accumulate. Concentrations of elements found were quantified based on the obtained EDX spectra.  25      A Surface section B Cross section. Figure 3-8 Example of arrangement of membrane samples in SEM-EDX stub 3.3.2.2.2. Field Emission Scanning Electron Microscope analysis Field Emission Scanning Electron Microscopy (FESEM) (Hitachi S-4700) was used at high magnification to obtain 10 to 200k resolution images at 1 and 10kV.  FESEM analysis was exclusively done on VM and FM1. Fouled and virgin hollow fibres were cleaned and arranged in stubs as described in Section 3.3.2.2.1 and illustrated in Figure 3-8. The hollow fibres membranes were sputtered with 8-10 nm of gold palladium using a High Resolution Sputter Coater (Cressington 208HR). This was done prior to the analysis to enhance the image contrast and reduce charging of the sample under the electron beam.   A B 26  CHAPTER 4     RESULTS AND DISCUSSION 4.1. Study of membrane plugging due to abrasion of PAC particle 4.1.1. Assessment of PAC particle size over time The results from the particle size analysis are presented in Figure 4-1 to Figure 4-3. The particle size of the Norit SA UF PAC decreased by approximately 90% to a diameter of 0.038 μm, 0.068 μm, and 0.123 μm for d10, d50, and d90, respectively. General Carbon PAC also decreased by approximately 90% to a  iameter of 0.062 μm, 0.094 μm, an  0.164 μm for d10, d50, and d90, respectively. The particle size of the Calgon WPH PAC decreased significantly by approximately 35% to a diameter of 0.034 µm, 0.063 µm, and 0.114 µm for d10, d50, and d90, respectively (Figure 4-2). The particle size of the Aqua Nuchar PAC and Calgon 208CP PAC did not change significantly throughout the experiment, in fact, a slight increase in size was observed. This might be an indication of particle agglomeration.  Figure 4-1 Particle size distribution d10 of     after centrifugation   article si e for  eneral  ar on an  algon 208        not availa le for  t pro ucts of 3.1x109, and 4.5x109) 0.01 0.10 1.00 10.00 -1.00E+09 0.00E+00 1.00E+09 2.00E+09 3.00E+09 4.00E+09 5.00E+09 d 1 0 ,  u m  t Norit Calgon WPH Aqua Nuchar General Carbon Calgon 208 CP 27   Figure 4-2 Particle size distribution d50 of     after centrifugation   article si e for  eneral  ar on an  algon 208        not availa le for  t pro ucts of 3.1x109, and 4.5x109)  Figure 4-3 Particle size distribution d90 of     after centrifugation  particle si e for  eneral  ar on an  algon 208        not availa le for  t pro ucts of 3.1x109, and 4.5x109) As presented in Figure 4-4, no relationship was observed between the change in particle size and the PAC hardness. Therefore, hardness cannot be used to predict the likelihood of PAC decreasing in size in a 0.01 0.10 1.00 10.00 -1.00E+09 0.00E+00 1.00E+09 2.00E+09 3.00E+09 4.00E+09 5.00E+09 d 5 0 ,  u m  t Norit Calgon WPH Aqua Nuchar General Carbon Calgon 208 CP 0.01 0.10 1.00 10.00 -1.00E+09 0.00E+00 1.00E+09 2.00E+09 3.00E+09 4.00E+09 5.00E+09 d 9 0 ,  u m  t Norit Calgon WPH Aqua Nuchar General Carbon Calgon 208 CP 28  PAC/UF system. Although a reduction in size was observed for only three of the five PACs investigated, all PACs contained material of a size comparable to that of the pores in the UF membranes used (Figure 4-1).  Figure 4-4 Relationship between change in particle size (d10) and PAC hardness in terms of Gold Number 4.1.2. Assessment of PAC pore plugging potential over time Considering that PAC contains material of a size comparable to that of the pores in the UF membrane used, it is possible that PAC could physically plug the membrane pores, potentially leading to irreversible fouling. However, as presented in Figure 4-5, no significant sustained reduction in membrane permeability was observed when filtering solutions of PAC in milliQ water, indicating that PAC on its own does not appear to plug the membrane pores. Note that a slight reduction of permeability was observed for all PACs studied. However, this reduction was recovered by backwashing. -60% -40% -20% 0% 20% 40% 60% 80% 100% 120% 0 0.5 1 1.5 2 2.5 C h an g e in  p ar ti cl e si ze Gold's Number Norit Calgon WPH Aqua Nuchar General Carbon Calgon 208 CP 29   Figure 4-5 Permeability of PAC/UF bench scale systems (data not collected on days 7 to 11and 22 to 25 due to malfunction of the data recording system; cleaning backflush was done for 4 min and 15 min on days 13 and 29, respectively) Although PAC was not observed to irreversibly foul the UF membranes on its own, field operating data indicated that PAC significantly contributed to irreversible fouling, as presented in Section 3.1. Additional investigations, presented in Section 4.2, were completed to provide insight into the mechanisms and cause of membrane fouling. 4.2. Detection of foulants in PAC/UF hollow fibres Results from different methods used to explore the mechanisms of fouling and the detection of foulants in a PAC/UF membrane system are presented in the following sections. Physical characteristics and composition of fouled (FM) and virgin (VM) hollow fibres were thoroughly analyzed in order to provide insight on the possible sources of membrane fouling. 4.2.1. Detection of PAC in irreversibly fouled membranes As presented in Figure 4-6, an apparent cake layer of foulants was observed on the irreversibly fouled membrane (FM) obtained from the field. Although the results presented in Section 4.1 suggest that the irreversible fouling was likely not due to the direct plugging of membrane pores by PAC, PAC may have nonetheless contributed to fouling. 1.0 1.5 2.0 2.5 3.0 3.5 0 5 10 15 20 25 30 35 P er m ea b il it y , L m h /K P a Time, days Backflush Backflush  Aqua Nuchar x Calgon WPH Δ  Norit 30   A. Virgin membrane at 100k  B Fouled membrane at 100k  C. Fouled membrane at 200k Figure 4-6 FESEM images of virgin and fouled membrane fibres Fouling in PAC/UF systems has been attributed by others to the combination of PAC and the organic material and/or PAC and inorganic material (Lin et al., 1999; Yiantsios and Karabelas, 2001; Zhang et al., 2003; Zhao et al., 2005; Oh et al., 2006; Saravia and Frimmel, 2008). Some researchers have reported that PAC alone does not affect the filtration performance in a PAC/UF system, but rather, PAC combined with organic and/or inorganic material affects the resistance to the permeate flow. Lin et al. (1999), Yiantsios and Karabelas (2001), Zhang et al. (2003), and Saravia and Frimmel (2008) attributed the increase in resistance to the permeate flow in PAC/UF systems to the combination of PAC and the organic material in the raw water source. They suggested that the NOM can bond strongly and simultaneously with both the PAC particles and the membrane surface and can thereby form a layer with high resistance to permeate flow. However, they also indicated that the removal of NOM from solution prior to the addition of PAC does not necessarily reduce fouling as reported by Zhang et al. (2003). On the other hand, Zhao et al. (2005) attributed the increase in the resistance to the permeate flow in PAC/UF systems to the combination of PAC and inorganic material. Oh et al. (2006) observed that both organic and inorganic material contribute to the formation of a complex cake structure on the surface of membranes in a PAC/UF system. Si and Ca (Baker and Dudley, 1998; Plottu-Pecheux et al., 2003; Mosqueda-Jimenez et al., 2008; Arnal et al., 2009), Mg, Fe, Al, K, and S (Arnal et al., 2009) have been observed in the membrane foulant layer in PAC/UF systems. Zhao et al. (2005) suggested that metal ions can neutralize the charge on PAC particles, enabling them to form a more compact structure. Inorganic material may originate from the water source (Arnal et al., 2009) or inadequate use (dosage) of coagulant (Lee et al., 2007). SEM-EDX analysis was used to determine if PAC was a significant component of the irreversible cake layer. Since fluorine is present in the membrane materials and is not expected to be present in the influent A B C 31  of the PAC/UF system, a change in the carbon-fluorine (C/F) ratio could provide evidence of the presence of PAC in the irreversible cake layer. The C/F ratios of virgin and fouled membrane fibres are presented in Figure 4-7. Although the C/F ratio was greater for fouled membrane fibres, it was not significantly greater (based on a 90% confidence interval of the results). Therefore it was not possible to conclude that significant amounts of PAC were present in the irreversible cake layer. SEM-EDX analysis did, however, reveal that the irreversible cake layer contained aluminum and chlorine. From the 68 FM samples tested, almost all contained traces of aluminum (Al), whereas only 14% of VM contained traces of these compounds (out of 62 samples). Chlorine (Cl) was present in 56% of the FM samples and in 16% of VM.  Figure 4-7 Frequency analysis of carbon to fluorine ratio in membranes The presence of aluminum in FM fibres could be due to the pre-treatment with polyaluminium chloride at the water treatment plant, from which the samples were obtained. The Ca and Si present on the FM fibres may have originated from the raw water source (as observed by Mosqueda-Jimenez et al., 2008). The results are consistent with those from other studies that indicate that small particulate material and metal ions in the inflow can have a significant influence on the PAC cake layer formation on membrane fibres (Zhao et al., 2005; Arnal et al., 2009). 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 F re q u en cy C/F Ratio Virgin Membrane (VM), n = 62 Fouled Membrane (FM), n = 68 Confidence interval (90% confidence): VM = 2.37 – 3.87 FM = 3.14 – 4.14 32  The volume of foulants needed to plug all of the pores of a single hollow fibre was estimated based on the number of pores in a unit area of membrane surface, and assuming that all pores are plugged with a spherical PAC particle of 0.04 µm in diameter. The number of pores in a unit area of 0.9 µm 2  was calculated using a 100k FESEM image of a random virgin hollow fibre and the assistance of Matlab® software. The number of pores in the unit area was estimated to be 199. The amount of PAC needed to plug all pores in a 1.9 m long hollow fibre was estimated to be approximately 2.07x10 -5 g.  Even if present, such a small amount of PAC could not be detected with the methods used to assess if PAC was present in irreversibly fouled membrane fibres. Although the results presented in Section 4.1 suggested that a complete plugging of membrane pores with PAC is unlikely, the calculations (Appendix F) confirmed the difficulty to detect PAC particles that could be embedded within membrane pores. 4.2.2. Assessment of mechanism of fouling in irreversibly fouled membranes The presence of a cake layer on the irreversibly fouled membrane observed by FESEM images (Figure 4-6) is consistent with field operating data. Results of model fitting are presented in Figure 4-8, the irreversible decrease in membrane permeability in the irreversibly fouled PAC/UF system was linear, suggesting that fouling was predominantly due to the formation of a cake layer (see Section 2.3.1). Components of this cake layer were extracted and analyzed; the results are presented in Section 4.2.3 below. 33   Figure 4-8 Permeability of PAC/UF system in WTP (Provided by GE Water & Process Technologies) with data fitted 4.2.3. Detection of foulants in irreversibly fouled membranes Material extracted from VM, FM1, FM2, and FM3 during sonication was analyzed.  Elements found in the extraction are presented in Figure 4-9 and Figure 4-10. Most of the material was released within the first hour of sonication. Aluminum (Al), silica (Si), calcium (Ca), potassium (K) and organic material (TOC) were released at highest concentrations from the irreversibly fouled hollow fibres. These results are consistent with those obtained in the present study by SEM-EDX analysis (Section 4.2.1). Material extracted from the irreversibly fouled membrane fibres could substantially affect the permeability, playing an important role in membrane fouling (Katsoufidou et al., 2005; Zhao et al., 2005; Oh et al., 2006).  Mosqueda-Jimenez et al. (2008), Baker and Dudley (1998), and Plottu-Pecheux et al. (2003) attributed the presence of calcium and silica on the membrane surface to the water source. Ionic material can promote intra- and inter-molecular forces between organic molecules by bridging free functional groups, thus promoting aggregation of those organics and deposition of a cake layer (Katsoufidou et al., 2005). 0 1 2 3 4 5 6 7 0 30 60 90 120 150 180 210 240 P er m ea b il it y , g fd /p si Time, d Chemical cleaning Data fitted to cake model (see Section 2.3.1) 34   Figure 4-9 Inorganic material released during sonication of fouled and virgin hollow fibres  Figure 4-10 TOC released during sonication of fouled and virgin hollow fibres Results from the clean water flux test performed on virgin and fouled hollow fibre membranes before and after sonication are illustrated in Figure 4-11. Clean water flux of virgin hollow fibre membranes was unaffected by sonication. However the clean water flux of fouled hollow fibre membranes increased significantly during sonication. These results indicate that the material released during sonication contributed to the observed irreversible fouling. The integrity of membrane fibres before and after sonication was also tested. No significant integrity breach (i.e. pressure drop or visible bubbles) was observed during the test, indicating that the sonication process did not affect the integrity of the membrane fibres. -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Al 308 Ca 315 Ca 317 K' 766 Si 212 Si 251 C o n ce n tr at io n , m g / m  o f fi b er Element, wavelength (nm) MQ water VM 1 FM 1 FM 2 FM 3 0.3 0.5 3.1 2.8 0.8 1.1 0.5 0.4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1 hr 2 hr C o n ce n tr at io n  T O C , m g / m  o f fi b er VM FM 1 FM 2 FM 3 35   Figure 4-11 Clean water flux before and after one hour of sonication of fouled and virgin hollow fibres (error bars correspond to the standard deviation of the measurements) A novel solubilisation technique was developed to dissolve the membrane material so that all foulants present on its surface could be analyzed. Once the hollow fibres were efficiently solubilised, the extract was collected and analyzed for organic compounds and metals. Solutions extracted from VM, FM1, FM2, and FM3 after the solubilisation were analyzed. Results from ICP and GC-MS analyses are illustrated in Figure 4-12 and Figure 4-13, respectively. Aluminum (Al), calcium (Ca), and sodium (Na) were identified in the solution. These results were consistent with those obtained by SEM-EDX analysis. Some organic compounds were found in fouled hollow fibres which were not found either in the virgin fibres nor in the blanks (DMF), confirming the presence of organic material in the irreversibly fouled hollow fibres.  A detailed list of the compounds identified is presented in APPENDIX G    . Organic and inorganic material extracted from the irreversibly fouled fibres suggests that these have great influence on the PAC cake layer formation on membrane fibres.  0 20 40 60 80 100 120 140 VM FM F lu x , L /m 2 ∙h Before Sonification After Sonification 36   Figure 4-12 Inorganic material released from solubilised fouled and virgin hollow fibres (error bars correspond to the standard deviation of the measurements)   0.00 1.00 2.00 3.00 4.00 5.00 6.00 Al 396 Ca 317 Na 588 C o n ce n tr at io n , m g / m  o f fi b er Element, wavelenth (nm) DMF Blank VM FM 1 FM 2 37  Figure 4-13 Organic material released from solubilised fouled and virgin hollow fibres (Table G-3) 38  CHAPTER 5     CONCLUSIONS The present project was designed as a preliminary screening study intended to gain insight into the mechanisms and cause of irreversible fouling in PAC/UF systems. Additional investigation is necessary to fully identify the causes of irreversible fouling in PAC/UF systems and determine the mechanism that facilitates the metal-PAC coalition. Analysis of the influent water source matrix and pretreatment processes should be explored in order to establish the relationship between influent constituents and the irreversible fouling potential of PAC/UF systems. Nonetheless, the following are the conclusions drawn from the results obtained from the present study. 1. The hydrodynamic conditions in PAC/UF systems can cause the breakage of PAC particles. Particle size of PAC can decrease to a range comparable to that of the size of the pores in the UF membranes used.  2. There was no apparent relationship between PAC hardness and the susceptibility of PAC particles to break into smaller particles.  3. Results from the bench scale PAC/UF systems suggest that the irreversible fouling of membranes was not solely due to the direct plugging of membrane pores by PAC.  4. The decrease in the permeability for the irreversibly fouled system with PAC was observed to be linear over time suggesting that fouling was due to the formation of a cake layer. This was confirmed by visual analysis of fouled membrane fibres. Although the results suggest that the irreversible fouling was likely not due to the direct plugging of membrane pores by PAC, PAC may have nonetheless contributed to fouling.  5. Organic and inorganic material extracted from the irreversibly fouled membrane suggests that these have great influence on the PAC cake layer formation on membrane fibres.  6. PAC does not cause irreversible membrane fouling in PAC/UF systems by itself, but it may facilitate the absorption of metals into membrane pores and this may induce organic material to get attached in/on membrane surface. 39  REFERENCES ......... Adham, S. S., Jacangelo, J. G., & Laine, J. M. (1996). Characteristics and costs of MF and UF plants. Journal American Water Works Association, 88(5), 22-31. Adham, S. S., Snoeyink, V. L., Clark, M. M., & Bersillon, J. L. (1991). Predicting and verifying organics removal by pac in an ultrafiltration system. Journal American Water Works Association, 83(12), 81- 91. APHA, AWWA, & WEF. (2005). In American Public Health Association (APHA), American Water Works Association (AWWA) and Water Environment Federation (WEF) (Eds.), Standard methods for the examination of water and wastewater (20th ed.). Washington, DC.: Arnal, J. M., Garcia-Fayos, B., Sancho, M., & Verdu, G. (2009). Ultrafiltration membrane cleaning with different chemical solutions after treating surface water. Desalination and Water Treatment, 7(1-3), 198-205. Baker, J. S., & Dudley, L. Y. (1998). Biofouling in membrane systems — A review. Desalination, 118(1- 3), 81-89. doi:DOI: 10.1016/S0011-9164(98)00091-5 Bolton, G. R., Boesch, A. W., & Lazzara, M. J. (2006). The effects of flow rate on membrane capacity: Development and application of adsorptive membrane fouling models. Journal of Membrane Science, 279(1-2), 625-634. doi:DOI: 10.1016/j.memsci.2005.12.057 Braghetta, A., DiGiano, F. A., & Ball, W. P. (1998). Nom accumulation at NF membrane surface: Impact of chemistry and shear. Journal of Environmental Engineering-ASCE, 124(11), 1087-1098. 40  Brandrup, J., Immergut, E. H., & Grulke, E. A. (Eds.). (1999). Polymer handbook. USA: Jhon Wiley & Sons Inc. Campinas, M., & Rosa, M. J. (2010). Assessing PAC contribution to the NOM fouling control in PAC/UF systems. Water Research, 44(5), 1636-1644. doi:10.1016/j.watres.2009.11.012 Chang, Y. J., Choo, K. H., Benjamin, M. M., & Reiber, S. (1998). Combined adsorption - UF process increases TOC removal. Journal American Water Works Association, 90(5), 90-102. Clark, M. M., Baudin, I., & Anselme, C. (1996). Membrane-powdered activated carbon reactors. In Mc Graw Hill (Ed.), Water treatment: Membrane processes Crittenden, J. C., Trussell, R. R., Hand, D. W., Howe, K. J., & Tchobanoglous, G. (2005). Water treatment, principles and design (2nd ed.) Jhon Wiley & Son Inc. Croue, J. P., & Korshin, G. V., Benjamin, M. M. (1999). In AWWARF, AWWA (Eds.), Characterization of natural organic matter in drinking water. Denver, CO.: EPA (Environment Protection Agency). (2005). Membrane filtration guidance manual. No. 815-R-06- 009). GE Water & Process Technologies. (2008). Powdered activated carbon - selection guidelines. No. WPG- 33-0702-1). Ontario, Canada: Health Canada. (2010). Guidelines for canadian drinking water quality. Hermans, P. H., & Bredée, H. L. (1935). Zur kenntnis der filtrationsgesetze. Rec. Trav. Chim. Pays-Bas, 54, 680. 41  Jacangelo, J. G., Laine, J. M., Cummings, E. W., & Adham, S. S. (1995). Uf with pretreatment for removing dbp precursors. Journal American Water Works Association, 87(3), 100-112. Katsoufidou, K., Yiantsios, S. G., & Karabelas, A. J. (2005). A study of ultrafiltration membrane fouling by humic acids and flux recovery by backwashing: Experiments and modeling. Journal of Membrane Science, 266(1-2), 40-50. doi:10.1016/j.memsci.2005.05.009 Kim, J. S., Lee, S. J., Yoon, S. H., & Lee, C. H. (1996). Competitive adsorption of trace organics on membranes and powdered activated carbon in powdered activated carbon-ultrafiltration system. Water Science and Technology, 34(9), 223-229. Lee, C. W., Bae, S. D., Han, S. W., & Kang, L. S. (2007). Application of ultrafiltration hybrid membrane processes for reuse of secondary effluent. Desalination, 202(1-3), 239-246. doi:DOI: 10.1016/j.desal.2005.12.059 Li, C., & Chen, Y. (2004). Fouling of UF membrane by humic substance: Effects of molecular weight and powder-activated carbon (PAC) pre-treatment. Desalination, 170(1), 59-67. doi:DOI: 10.1016/j.desal.2004.03.015 Lin, C., Huang, Y., & Hao, O. J. (1999). Ultrafiltration processes for removing humic substances: Effect of molecular weight fractions and PAC treatment. Water Research, 33(5), 1252-1264. doi:DOI: 10.1016/S0043-1354(98)00322-4 Matsui, Y., Sanogawa, T., Aoki, N., Mima, S., & Matsushita, T. (2006). Evaluating submicron-sized activated carbon adsorption for microfiltration pretreatment. Water Science and Technology: Water Supply, 6(1), 149-155. doi:10.2166/ws.2006.017 Matsui, Y., Hasegawa, H., Ohno, K., Matsushita, T., Mima, S., Kawase, Y., & Aizawa, T. (2009). Effects of super-powdered activated carbon pretreatment on coagulation and trans-membrane pressure 42  buildup during microfiltration. Water Research, 43(20), 5160-5170. doi:DOI: 10.1016/j.watres.2009.08.021 Miller, J. D. (2011). Gold number report for university of British Columbia. No. E1131. Texas: White Rock Engineering Services (WRES, Inc.). Mosqueda-Jimenez, D. B., Huck, P. M., & Basu, O. D. (2008). Fouling characteristics of an ultrafiltration membrane used in drinking water treatment. Desalination, 230(1-3), 79-91. doi:10.1016/j.desal.2007.09.014 Oh, H., Yu, M., Takizawa, S., & Ohgaki, S. (2006). Evaluation of PAC behavior and fouling formation in an integrated PAC–UF membrane for surface water treatment. Desalination, 192(1-3), 54-62. doi:DOI: 10.1016/j.desal.2005.07.034 Plottu-Pecheux, A., Houssais, B., Bele, C., Democrate, C., Gatel, d., & Cavard, J. (2003). Autopsies of membranes fouled on Mery-sur-Oise pilot units: Deposit characterization and cleaning efficiency assessment. Proc. WQTC Conference, Saravia, F., & Frimmel, F. H. (2008). Role of NOM in the performance of adsorption-membrane hybrid systems applied for the removal of pharmaceuticals. Desalination, 224(1-3), 168-171. doi:DOI: 10.1016/j.desal.2007.02.089 Saravia, F., Naab, P., & Frimmel, F. H. (2006). Influence of particle size and particle size distribution on membrane-adsorption hybrid systems. Desalination, 200(1-3), 446-448. doi:DOI: 10.1016/j.desal.2006.03.370 Seo, G. T., Jang, S. W., Lee, S. H., & Yoon, C. H. (2005). The fouling characterization and control in the high concentration PAC membrane bioreactor HCPAC-MBR. Water Science and Technology, 51(6- 7), 77-84. 43  Tomaszewska, M., & Mozia, S. (2002). Removal of organic matter from water by PAC/UF system. Water Research, 36(16), 4137-4143. doi:DOI: 10.1016/S0043-1354(02)00122-7 USEPA (United States Environment Protection Agency). (1998). National primary drinking water regulations: Disinfectants and disinfection byproducts rule. Federal Register, 63(241), 69389-69476. Yiantsios, S. G., & Karabelas, A. (2001). An experimental study of humid acid and powdered activated carbon deposition on UF membranes and their removal by backwashing. Desalination, 140(2), 195- 209. Zhang, M., Li, C., Benjamin, M. M., & Chang, Y. (2003). Fouling and natural organic matter removal in Adsorbent/Membrane systems for drinking water treatment. Environmental Science & Technology, 37(8), 1663-1669. Zhao, P., Takizawa, S., Katayama, H., & Ohgaki, S. (2005). Factors causing PAC cake fouling in PAC– MF (powdered activated carbon-microfiltration) water treatment systems. Water Science & Technology, 51(6–7), 231-240.   44  APPENDIX A     WATER QUALITY OF WATER TREATMENT PLANT  Table A-1 Water quality of raw water for the full scale PAC/UF membrane system under investigation Turbidity, NTU Alkalinity, mg/L pH Hardness, mg/L Iron, mg/L Manganese, mg/L Fluoride, mg/L 5.9 71 8.0 80 0.043 0.069 0.11  45  APPENDIX B     OPERATING CONDITIONS OF GE UNIT A  Permeation pressure 41.37 kPa (6 psi) Backpulse pressure 62.05 kPa (9 psi) Temperature 15 o C Frequency of  pressure decay test twice per day at 34.47 kPa (5 psi) Production cycle 15 min permeation (formerly 30 min) (permeate recycled); 60 sec backpulse Aeration 707.5 L/min (25 cfm) continuous    46  APPENDIX C     EXPERIMENTAL BENCH SCALE SYSTEM  Figure C-1 Picture of the bench scale air sparged system apparatus  Figure C-2 Picture of the bench-scale PAC/UF system apparatus 47  APPENDIX D     PARTICLE SIZE ANALYSIS Table D-1 Particle size distribution of PAC, after centrifugation  Aqua Nuchar  General Carbon  day 0 day 95 day 131  day 0 day 95 day 131 d(0.1) 0.037 0.035 0.057  NA NA 0.064 d(0.5) 0.068 0.065 0.088  NA NA 0.097 d(0.9) 0.133 0.122 0.158  NA NA 0.167 NA: Data not available due to technical difficulties  48  APPENDIX E     CLEAN WATER FLUX TEST  Figure E-1 Picture of clean water flux test apparatus  49  APPENDIX F     CALCULATIONS The amount of foulants needed to plug all pores of a single hollow fibre was calculated by visual analysis and Matlab® calculations. Calculation of the number of pores in a single fibre The number of pores in a fixed area was calculated using a 100k FESEM image of a random virgin hollow fibre (see Figure F-1) and the assistance of Matlab® software. The hollow fibre area analyzed was 0.9 µm 2 , and the number of pores calculated for that area was 179 (i.e., 199 pores per µm 2 ).  Figure F-1 FESEM picture of a virgin hollow fibre, area under analysis = 0.9 μm2 The approximate number of pores in the surface of a single membrane fibre was calculated as shown bellow. (OD) Outer diameter of a single fibre = 0.0018 m (L) Length of a single fibre = 1.90 m 1.                                   μ 2.   50  Calculation of the amount of foulants needed to plug all pores of a single fibre To be able to make the calculations, the assumptions listed below were taken into consideration (Figure F-2). 1) Pore plugging was the dominant mechanism 2) The diameter of all the pores of the hollow fibre were within the average calculated 3) The number of pores visually estimated was applied to the entire fibre 4) PAC particles (foulants) were spheres of the same size of the fibre pores  Figure F-2 Schematic of pore plugged by a PAC particle (assumption) 3.                 4.  5.  The volume of foulants needed to plug all pores of a single hollow fibre is 2.07x10 -5 g. 0.04 µm 0.04 µm PAC particle Membrane pore 51  APPENDIX G     TOC AND ICP RESULTS FROM HOLLOW FIBRE EXTRACTIONS Table G-1 TOC and inorganic (ICP-OES analysis) material extracted from sonicated hollow fibres Sample units Sonication time TOC mg/L IC Al 308.215 Ba 233.527 Ca 315.887 R Ca 317.933 R Cd 214.440 Cd 228.802 ID Replica #     mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L MQ1 1 mg/L 0 0.9 < 1 -0.027 0.000 -0.003 -0.005 0.000 0.000 MQ2 2 mg/L 0 - 0.086 -0.003 -0.017 -0.032 -0.006 -0.004 Mean: mg/L   0.9   0.030 -0.002 -0.010 -0.019 -0.003 -0.002 MQ1 1 mg/m of fibre 0 0.1 -0.00159 0.00000 -0.00018 -0.00029 0.00000 0.00000 MQ2 2 mg/m of fibre 0 - 0.00506 -0.00018 -0.00100 -0.00188 -0.00035 -0.00024 Mean:   mg/m of fibre   0.1   0.002 0.000 -0.001 -0.001 0.000 0.000 Virgin Fibre mg/L 1 hr 5.5 < 1 -0.025 0.001 0.003 0 -0.001 -0.001 Mean:  mg/L 2 hr 8.7 < 1 -0.021 0.001 0.091 0.086 0 0 Virgin Fibre mg/m of fibre 1 hr 0.3  -0.00147 5.88E-05 0.000176 0 -5.9E-05 -5.88E-05 Mean:   mg/m of fibre 2 hr 0.5  -0.00124 5.88E-05 0.005353 0.005059 0 0 Fouled 1  mg/L 1 hr 52.3 < 1 0.253 0 0.384 0.389 -0.003 -0.002   mg/L 2 hr 48.3 < 1 0.346 0 0.521 0.525 -0.002 -0.002 Fouled 1  mg/m of fibre 1 hr 3.1  0.014882 0 0.022588 0.022882 -0.00018 -0.000118     mg/m of fibre 2 hr 2.8   0.020353 0 0.030647 0.030882 -0.00012 -0.000118 Fouled 2  mg/L 1 hr 14.4 < 1 0.059 0 0.157 0.16 -0.002 -0.002   mg/L 2 hr 18.0 < 1 0.061 0.001 0.303 0.299 0 0 Fouled 2  mg/m of fibre 1 hr 0.8  0.003471 0 0.009235 0.009412 -0.00012 -0.000118     mg/m of fibre 2 hr 1.1   0.003588 5.88E-05 0.017824 0.017588 0 0 Fouled 3  mg/L 1 hr 9.2 < 1 0.167 0.001 0.425 0.428 -0.002 -0.001   mg/L 2 hr 7.2 < 1 0.140 0.001 0.645 0.648 -0.002 -0.002    52    Sample units Mn 257.610 Mn 259.372 Ni 221.648 Ni 231.604 P 213.617 P 214.914 Pb 217.000 Pb 220.353 Si 212.412 ID Replica # mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Fouled 3  mg/m of fibre 1 hr 0.5  0.009824 5.88E-05 0.025 0.025176 -0.00012 -5.88E-05     mg/m of fibre 2 hr 0.4   0.008235 5.88E-05 0.037941 0.038118 -0.00012 -0.000118 MQ1 1 mg/L 0.000 0.002 0.000 -0.020 -0.020 -0.090 -0.075 0.000 0.000 MQ2 2 mg/L -0.007 -0.002 0.002 -0.024 -0.026 -0.170 -0.163 0.988 0.025 Mean: mg/L -0.004 0.000 0.001 -0.022 -0.023 -0.130 -0.119 0.494 0.013 MQ1 1 mg/m of fibre 0.00000 0.00012 0.00000 -0.00118 -0.00118 -0.00529 -0.00441 0.00000 0.00000 MQ2 2 mg/m of fibre -0.00041 -0.00012 0.00012 -0.00141 -0.00153 -0.01000 -0.00959 0.05812 0.00147 Mean:   mg/m of fibre 0.000 0.000 0.000 -0.001 -0.001 -0.008 -0.007 0.029 0.001 Virgin Fibre mg/L 0.001 0.001 -0.001 -0.02 -0.019 -0.087 -0.071 1.843 0.001 Mean:  mg/L 0.001 0.002 -0.001 -0.018 -0.018 -0.09 -0.073 0.846 0.004 Virgin Fibre mg/m of fibre 5.882E-05 5.882E-05 -5.88E-05 -0.001176 -0.001118 -0.0051176 -0.0041765 0.1084118 5.882E-05 Mean:   mg/m of fibre 5.882E-05 0.0001176 -5.88E-05 -0.001059 -0.001059 -0.0052941 -0.0042941 0.0497647 0.0002353 Fouled 1  mg/L -0.001 0 0 -0.007 -0.006 -0.188 -0.162 4.958 0.037   mg/L 0 -0.001 0 -0.009 -0.009 -0.178 -0.152 4.753 0.065 Fouled 1  mg/m of fibre -5.88E-05 0 0 -0.000412 -0.000353 -0.0110588 -0.0095294 0.2916471 0.0021765     mg/m of fibre 0 -5.88E-05 0 -0.000529 -0.000529 -0.0104706 -0.0089412 0.2795882 0.0038235 Fouled 2  mg/L -0.001 -0.001 0 -0.011 -0.011 -0.207 -0.182 4.625 0.015   mg/L 0.001 0.003 -0.001 -0.018 -0.017 -0.075 -0.058 3.799 0.045 Fouled 2  mg/m of fibre -5.88E-05 -5.88E-05 0 -0.000647 -0.000647 -0.0121765 -0.0107059 0.2720588 0.0008824     mg/m of fibre 5.882E-05 0.0001765 -5.88E-05 -0.001059 -0.001 -0.0044118 -0.0034118 0.2234706 0.0026471 Fouled 3  mg/L -0.001 0.001 -0.001 -0.003 -0.002 -0.196 -0.173 5.925 0.036   mg/L 0 -0.002 0 -0.005 -0.005 -0.202 -0.178 5.623 0.064 Fouled 3  mg/m of fibre -5.88E-05 5.882E-05 -5.88E-05 -0.000176 -0.000118 -0.0115294 -0.0101765 0.3485294 0.0021176     mg/m of fibre 0 -0.000118 0 -0.000294 -0.000294 -0.0118824 -0.0104706 0.3307647 0.0037647   53  Sample units Mn 257.610 Mn 259.372 Ni 221.648 Ni 231.604 P    213.617 P     214.914 Pb 217.000 Pb 220.353 Si 212.412 ID Replica # mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L MQ1 1 mg/L -0.001 -0.001 -0.001 -0.001 -0.019 -0.028 -0.006 -0.013 -0.011 MQ2 2 mg/L -0.004 -0.004 -0.003 -0.012 -0.066 -0.078 0.012 -0.008 -0.507 Mean: mg/L -0.003 -0.003 -0.002 -0.007 -0.043 -0.053 0.003 -0.011 -0.259 MQ1 1 mg/m of fibre -0.00006 -0.00006 -0.00006 -0.00006 -0.00112 -0.00165 -0.00035 -0.00076 -0.00065 MQ2 2 mg/m of fibre -0.00024 -0.00024 -0.00018 -0.00071 -0.00388 -0.00459 0.00071 -0.00047 -0.02982 Mean:   mg/m of fibre 0.000 0.000 0.000 0.000 -0.003 -0.003 0.000 -0.001 -0.015 Virgin Fibre mg/L -0.001 -0.001 -0.002 -0.002 -0.015 -0.019 0.002 -0.016 0.009 Mean:  mg/L -0.001 -0.001 0 -0.002 -0.019 -0.029 0.01 -0.013 0.005 Virgin Fibre mg/m of fibre -5.88E-05 -5.88E-05 -0.000118 -0.000118 -0.000882 -0.001118 0.0001176 -0.000941 0.0005294 Mean:   mg/m of fibre -5.88E-05 -5.88E-05 0 -0.000118 -0.001118 -0.001706 0.0005882 -0.000765 0.0002941 Fouled 1  mg/L 0.007 0.006 0 0.002 0.006 -0.007 0.014 -0.022 0.271   mg/L 0.008 0.007 0.001 0.002 0.014 0.006 0.001 -0.022 0.447 Fouled 1  mg/m of fibre 0.0004118 0.0003529 0 0.0001176 0.0003529 -0.000412 0.0008235 -0.001294 0.0159412     mg/m of fibre 0.0004706 0.0004118 5.882E-05 0.0001176 0.0008235 0.0003529 5.882E-05 -0.001294 0.0262941 Fouled 2  mg/L 0.001 0 0 0.002 0.001 0.005 0.003 -0.022 0.121   mg/L 0.003 0.002 -0.001 -0.002 -0.001 -0.014 -0.01 -0.011 0.174 Fouled 2  mg/m of fibre 5.882E-05 0 0 0.0001176 5.882E-05 0.0002941 0.0001765 -0.001294 0.0071176     mg/m of fibre 0.0001765 0.0001176 -5.88E-05 -0.000118 -5.88E-05 -0.000824 -0.000588 -0.000647 0.0102353 Fouled 3  mg/L 0.116 0.104 0 0.004 0.012 0.011 0.005 -0.02 0.476   mg/L 0.138 0.123 0.002 0.004 0.004 0.002 0 -0.024 0.611 Fouled 3  mg/m of fibre 0.0068235 0.0061176 0 0.0002353 0.0007059 0.0006471 0.0002941 -0.001176 0.028     mg/m of fibre 0.0081176 0.0072353 0.0001176 0.0002353 0.0002353 0.0001176 0 -0.001412 0.0359412        54    Sample units Si        251.611 Zn 206.200 Zn 213.857 ID Replica #   mg/L mg/L mg/L MQ1 1 mg/L -0.007 0.000 -0.001 MQ2 2 mg/L -0.620 -0.009 -0.001 Mean: mg/L -0.314 -0.005 -0.001 MQ1 1 mg/m of fibre -0.00041 0.00000 -0.00006 MQ2 2 mg/m of fibre -0.03647 -0.00053 -0.00006 Mean:   mg/m of fibre -0.018 0.000 0.000 Virgin Fibre mg/L 0.012 0.003 0.002 Mean:  mg/L 0.008 0.004 0.003 Virgin Fibre mg/m of fibre 0.0007059 0.0001765 0.0001176 Mean:   mg/m of fibre 0.0004706 0.0002353 0.0001765 Fouled 1  mg/L 0.275 0.008 0.006   mg/L 0.446 0.005 0.004 Fouled 1  mg/m of fibre 0.0161765 0.0004706 0.0003529     mg/m of fibre 0.0262353 0.0002941 0.0002353 Fouled 2  mg/L 0.123 0.002 0.002   mg/L 0.175 0.013 0.012 Fouled 2  mg/m of fibre 0.0072353 0.0001176 0.0001176     mg/m of fibre 0.0102941 0.0007647 0.0007059 Fouled 3  mg/L 0.481 0.014 0.013   mg/L 0.612 0.015 0.014 Fouled 3  mg/m of fibre 0.0282941 0.0008235 0.0007647     mg/m of fibre 0.036 0.0008824 0.0008235 55  Table G-2 Inorganic material extracted from solubilised hollow fibres (ICP-OES analysis)  Description Al 396 As Be Ca 317 Cd  Co  Cr  Cu Fe  K mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L DMF Blank 0.254 0.000 0.000 0.000 0.086 0.020 0.072 0.000 0.000 0.000 Virgin Membrane, VM 0.318 0.000 0.000 0.203 0.072 0.022 0.062 0.002 0.130 0.030 Virgin Membrane, VM (duplicate) 0.338 0.000 0.000 0.175 0.068 0.022 0.065 0.000 0.000 0.057 Average VM = 0.328 0.000 0.000 0.189 0.070 0.022 0.063 0.001 0.065 0.043 Fouled Membrane, FM1 0.206 0.000 0.000 0.188 0.019 0.003 0.011 0.048 0.001 0.004 Fouled Membrane, FM1 (duplicate) 0.272 0.000 0.000 0.373 0.021 0.005 0.014 0.019 0.011 0.002 Average FM1 = 0.239 0.000 0.000 0.280 0.020 0.004 0.013 0.033 0.006 0.003 Fouled Membrane, FM2 0.318 0.000 0.000 0.892 0.028 0.004 0.022 0.004 0.068 0.003  Description Li Mg Mn  Mo Na 588 Ni P Pb Y Zn mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L DMF Blank 0.020 0.002 0.000 0.198 0.774 0.176 0.298 0.000 0.000 0.092 Virgin Membrane, VM 0.017 0.022 0.002 0.167 5.160 0.148 0.310 0.000 0.000 0.093 Virgin Membrane, VM (duplicate) 0.017 0.022 0.002 0.165 5.023 0.143 0.250 0.000 0.000 0.088 Average VM = 0.017 0.022 0.002 0.166 5.092 0.146 0.280 0.000 0.000 0.091 Fouled Membrane, FM1 0.005 0.012 0.143 0.049 3.165 0.045 0.069 0.000 0.000 0.137 Fouled Membrane, FM1 (duplicate) 0.007 0.019 0.255 0.051 3.511 0.052 0.079 0.000 0.000 0.102 Average FM1 = 0.006 0.016 0.199 0.050 3.338 0.049 0.074 0.000 0.000 0.119 Fouled Membrane, FM2 0.006 0.032 0.016 0.071 4.598 0.068 0.218 0.000 0.000 0.082   56   Table G-3 Organic material extracted from solubilised fouled and virgin hollow fibres (GC analysis) Retention time, min Component I.D. 5.29 Benzyl chloride 5.50 Benzyl Alcohol and 1-methyl-2-pyrrolidinone 6.3 Naphthalene 10.965 1-Dodecanamine,N,N-dimethyl 10.97 N.N-dimethyl-1-dodecanamine 13.83 N,N-Dimethyltetradecanamine 14.99 Phenanthrene 18.48 Pyrene 18.97 Butyl ester, hexadecanoic acid 1.53 2-methylpropyl ester, octadecanoic acid  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0063210/manifest

Comment

Related Items