Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Exploration of fouling propensity in an anaerobic membrane bioreactor treating municipal wastewater and… Pattanayak, Soubhagya Kumar 2007

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

Item Metadata

Download

Media
831-ubc_2007-0550.pdf [ 6.83MB ]
Metadata
JSON: 831-1.0063221.json
JSON-LD: 831-1.0063221-ld.json
RDF/XML (Pretty): 831-1.0063221-rdf.xml
RDF/JSON: 831-1.0063221-rdf.json
Turtle: 831-1.0063221-turtle.txt
N-Triples: 831-1.0063221-rdf-ntriples.txt
Original Record: 831-1.0063221-source.json
Full Text
831-1.0063221-fulltext.txt
Citation
831-1.0063221.ris

Full Text

EXPLORATION OF FOULING PROPENSITY IN AN ANAEROBIC MEMBRANE BIOREACTOR TREATING MUNICIPAL WASTEWATER AND COMPARISON TO THAT OF AN AEROBIC MEMBRANE BIOREACTOR by SOUBHAGYA KUMAR PATTANAYAK B.E. (Civil Engineering), Malaviya National Institute of Technology, India 2002 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA April 2007 © Soubhagya Kumar Pattanayak, 2007 / ABSTRACT Anaerobic biodegradation is a successful technology that has been used in industrial, food processing, and agricultural wastewater treatment for many decades. The operational costs associated with anaerobic systems are typically lower than with aerobic systems and anaerobic systems also generate less waste sludge. However, the application of anaerobic treatment systems is limited for low strength wastewaters in colder climates. In colder climates, the biomass growth yield and the growth rate are relatively low, resulting in a low net biomass production. To maintain adequate biomass concentration in an anaerobic bioreactor, membrane modules can be coupled to the reactor to effectively treat low strength wastewater in colder climates. One of the important advantages of the membrane bioreactors (MBRs) is that the membrane component of the system can retain virtually all of the biomass within the bioreactor. Membrane units in an anaerobic MBR can operate either as external units or as submerged units, depending on the requirements of the process. Currently, the application of submerged AnMBRs is limited as compared to external AnMBRs. However, the vacuum-driven submerged membrane process shows a lot of promise as compared to external membrane processes. This is because high energy consumption is one of the biggest limiting factors in external membranes. Also, the use of head space gas for reducing fouling in submerged anaerobic membranes can be a very successful technology in limiting energy consumption in bioreactors. However, the widespread use of membrane technology has been limited due to the fouling of the membrane fibers. Membrane fouling is an inherent problem with membrane processes, which not only affects the long term operational stability, but also leads to significant operational costs due to increased membrane replacement frequency and added energy consumption. Therefore, a considerable amount of research and engineering effort has been devoted to understanding the mechanisms of membrane fouling and to work out fouling prevention and control strategies. The broad objective of the present study was (1) to assess the treatment performance of a submerged membrane AnMBR treating low strength municipal ii wastewater at an ambient temperature and (2) to identify and characterize the fouling mechanism in the AnMBR. The anaerobic process was effective in removing chemical oxygen demand (COD) and volatile fatty acid (VFA) from the effluent VFA removal was essentially complete and 80% COD removal was achieved under acetate-supplemented conditions. Nonetheless, high concentration of effluent COD (i.e. 72 mg/L) indicated that aerobic post treatment is needed to achieve secondary quality effluent. On-line filtration studies were conducted simultaneously on both the anaerobic membrane and the aerobic membrane of membrane enhanced biological phosphorus removal (MEBPR) process. The on-line tests were conducted to compare the fouling mechanism in the anaerobic membrane process and the aerobic membrane process. The results from energy dispersive X-ray (EDX) analysis suggest that inorganic materials were not the prominent foulants in the AnMBR. Also, scanning electron microscopy (SEM) analysis indicated that very little microbial colonization was found in the anaerobic membranes. Mixed liquor characterization tests were conducted to verify the role of extra-cellular polymeric substances (EPS) on membrane fouling. The tests concluded that the total bound EPS concentration was higher in the aerobic mixed liquor and the total soluble EPS concentration was higher in the anaerobic mixed liquor. The high soluble EPS concentration was probably the reason for rapid fouling of anaerobic membranes. In addition, the higher SRT in the case of the anaerobic membrane bioreactor might be responsible for the high soluble EPS production observed in some studies. iii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES , vii LIST OF FIGURES viii LIST OF ABBREVIATIONS AND SYMBOLS x ACKNOWLEDGEMENTS xi 1. INTRODUCTION 1 1.1 Anaerobic Process 1 1.2 Anaerobic Membrane Bioreactors (AnMBRs) 2 1.3 Fouling of Anaerobic Membrane Bioreactors 3 2. BACKGROUND AND LITREATURE REVIEW 4 2.1 Introduction.. 4 2.2 Anaerobic Membrane Bioreactor System Configuration 5 2.3 Factors Affecting the Treatment Performance of Anaerobic MBR Systems 7 2.4 Factors Affecting the Membrane Performance of Anaerobic MBR System 9 2.4.1 Membrane material 9 2.4.2 Membrane pore size 10 2.4.3 Hydrophobic nature and charge of membrane.. 10 2.5 Operational Parameters Governing the Membrane Performance of Anaerobic MBR System 11 2.5.1 Permeate flux....; 11 2.5.2 Trans membrane pressure (TMP) 12 2.5.3 Hydrodynamics 13 2.5.4 Mixed liquor suspended solids 15 2.5.5 Membrane cleaning 15 2.6 Membrane Fouling in Anaerobic MBR Systems 16 2.6.1 Biofouling 17 2.6.2 Organic fouling and inorganic fouling 19 2.7 Conclusions 20 iv 2.8 Research Objectives 21 3. MATERIALS AND METHODS 22 3.1 B E N C H S C A L E A N A E R O B I C S U B M E R G E D M E M B R A N E A T T H E UBC P I L O T P L A N T ..22 3.2 Initial Testing and Acclimatization of the Anaerobic Reactor 24 3.3 Membrane Cleaning 25 3.4 Monitoring 26 3.4.1 Routine monitoring 26 3.5 Analytical Methods 26 3.5.1 Chemical oxygen demand (COD) 27 3.5.2 Volatile fatty acids (VFA) 27 3.5.3 Gas composition 28 3.5.4 MLSS/MLVSS 28 3.5.5 BOD5 test 29 3.6 Periodic Monitoring 29 3.6.1 On-line filtration tests 29 3.6.2 Membrane sampling 32 3.6.3 Scanning electron microscopy (SEM)- Energy dispersive x-ray (EDX) analysis32 3.6.4 Extraction of membrane foulants 33 3.6.5 Chemical analysis of extracted fouling material 33 3.7 Mixed Liquor Characterization ; 36 4. RESULTS AND DISCUSSION 37 4.1 Assessment of the Operational Performance of Submerged AnMBR 37 4.1.1 UBC pilot plant influent wastewater characteristics 37 4.1.2 Description of operating conditions 38 4.1.3 Treatment performance 41 4.1.4 Biogas composition and production 45 4.1.5 Bioreactor suspended solids concentration 50 4.1.6 Membrane flux and trans-membrane pressure (TMP) 51 4.2 Assessment of Membrane Filtration Characteristics in Submerged AnMBR 52 4.2.1 Flux and permeability profile of anaerobic MBR and aerobic MBR 53 4.2.2 Scanning electron microscopy -Energy dispersive X-ray analysis 55 v 4.2.3 Chemical analysis of extracted foulant material 61 4.3 EPS Analysis of Anaerobic MBR and Aerobic MBR Mixed Liquors 64 5. CONCLUSIONS 73 6. RECOMMENDATIONS 75 7. REFERENCES 77 APPENDIX A: UBC WASTEWATER TREATMENT PILOT PLANT 83 APPENDIX B: CLEAN AND FOULED US FILTER MEMBRANE MODULES 85 APPENDIX C: BOD 5 DATA OF THE ANMBR INFLUENT AND EFFLUENT 87 APPENDIX D: DATA SUMMARY OF TRANS MEMBRANE PRESSURE AND FLUX IN AEROBIC MEMBRANE BIOREACTOR DURING ON-LINE FILTRATION STUDIES 89 APPENDIX E: DAILY MONITORING RECORDS 92 vi LIST OF TABLES Table 3.1. Description of membrane modules used in anaerobic submerged MBR 24 Table 3.2. Monitoring program to track ongoing process performance 26 Table 4.1. UBC pilot plant influent wastewater characteristics 37 Table 4.2. Measured COD removal efficiencies during the periods of operation with and without acetate supplementation 42 Table 4.3. Average influent COD, permeate COD, VFA-COD, "refractory" COD and removal efficiency 43 Table 4.4. Measurements of BOD5 in selected raw wastewater and permeate samples of the submerged membrane bioreactor 44 Table 4.5. Typical EDX analysis of clean membrane, anaerobic MBR membrane and aerobic MBR membrane 60 Table 4.6. TOC, TP and TKN analysis of extracted foulant material 62 Table 4.7. SMP analysis of extracted foulant material 63 Table C.l. BOD5 concentration of the influent and effluent 88 Table D.l. TMP and flux data of the aerobic MBR during on-line filtration studies 90 vii LIST OF FIGURES Figure 2.1. Schematic of a pressure driven external anaerobic MBR 6 Figure 2.2. Schematic of a vacuum driven submerged anaerobic MBR 7 Figure 3.1. Schematic of the submerged membrane anaerobic bioreactor system, 22 Figure 3.2. Schematic of the MEBPR system 31 Figure 4.1. Hydraulic retention times in the submerged membrane bioreactor 38 Figure 4.2. Organic loading rates in the submerged membrane bioreactor 39 Figure 4.3. pH of the submerged membrane bioreactor influent and effluent 40 Figure 4.4. Temperature of the submerged membrane bioreactor 40 Figure 4.5. Total and soluble COD concentrations in influent and permeates for the submerged membrane bioreactor. (w/A: influent with acetate supplementation) 41 Figure 4.6. Acetic acid concentrations in the submerged membrane bioreactor influent and permeate. (A: influent with acetate supplementation; NA: influent without acetate supplementation) 43 Figure 4.7. Measured percent composition of biogas from the submerged membrane bioreactor 46 Figure 4.8. Measured and expected total biogas production rates in submerged membrane bioreactor 47 Figure 4.9. Adjusted percentage composition of CH 4 and C O 2 in submerged membrane bioreactor biogas 48 Figure 4.10. Measured, Total measured (including dissolved methane in the effluent) and expected total methane production rates in submerged membrane bioreactor 49 Figure 4.11. Submerged membrane bioreactor sludge concentration 50 Figure 4.12. Flux of the submerged membrane bioreactor 51 Figure 4.13. Trans-membrane pressure (TMP) of the submerged membrane bioreactor.52 Figure 4.14. Flux of the aerobic MBR and the anaerobic MBR during the on-line filtration tests 53 Figure 4.15. Evolution of TMP of the aerobic MBR and anaerobic MBR during the on-line filtration tests 54 Figure 4.16. SEM images of virgin membrane surfaces at different magnifications (45, 2.5K, 10K and 25K respectively) 56 viii Figure 4.17. SEM images of anaerobic membrane surfaces at different magnifications (50, 2.5K, 10K and 25K respectively) 58 Figure 4.18. SEM images of aerobic membrane surfaces at different magnifications (50, 2.5K, 10K and 25K respectively) 59 Figure 4.19. Bound carbohydrates concentration in mixed liquor 64 Figure 4.20. Bound proteins concentration in mixed liquor 65 Figure 4.21. Bound humic substances concentration in mixed liquor 65 Figure 4.22. Total bound EPS concentration in mixed liquor 66 Figure 4.23. Soluble carbohydrates concentration in mixed liquor 67 Figure 4.24. Soluble proteins concentration in mixed liquor 68 Figure 4.25. Soluble humic substances concentration in mixed liquor 68 Figure 4.26.Total soluble EPS concentration in mixed liquor 69 Figure 4.27. Total EPS concentration in the aerobic mixed liquor 70 Figure 4.28. Total EPS concentration in the anaerobic mixed liquor 70 Figure 4.29. Total bound EPS flux in membranes 71 Figure 4.30. Total soluble EPS flux in membranes 72 Figure A. 1. UBC wastewater treatment pilot plant 84 Figure A.2. UBC pilot plant anaerobic membrane bioreactor 84 Figure B.l. Clean US filter membrane modules 86 Figure B.2. Fouled US filter membrane modules 86 ix LIST OF ABBREVIATIONS AnMBR Anaerobic Membrane Bioreactor BAP Biomass-Associated Products BOD Biochemical Oxygen Demand CER Cation Exchange Resin COD Chemical Oxygen Demand CSTR Continuous Stirred-Tank Reactor DDW Distilled Deionized Water EDX Energy Dispersive X-Ray EPS Extracellular Polymeric Substances HRT Hydraulic Retention Time MBRs Membrane Bioreactors MEBPR Membrane Enhanced Biological Phosphorus Removal MLSS Mixed Liquor Suspended Solids MLVSS Mixed Liquor Volatile Suspended Solids PVDF Polyvinylidene Fluoride SEM Scanning Electron Microscopy SMP Soluble Microbial Products SRT Solids Retention Time TMP Trans membrane Pressure TSS Total Suspended Solids UAP Utilization-Associated Products UASB Upflow Anaerobic Sludge Blanket UBC University of British Columbia VFA Volatile Fatty Acid X ACKNOWLEDGEMENTS I take this opportunity thank my teacher and research supervisor Dr. Eric R. Hall for his inspiration, guidance and patience during the past two year that I have worked under him. I am also thankful to Dr. Pierre R. Berube, for reviewing this thesis and for the invaluable information on membrane technologies he provided during our meetings. I greatly acknowledge Water Environment Research Foundation (WERF) for having provided me with the financial assistance. I express my gratitude to Fred Koch who helped me a lot with my experimental set-up at the UBC Pilot Plant. Thanks are extended to Bill Leung, Derrick Home and Doug Smith who were always ready to lend a helping hand. I thank my colleagues, Alessandro Monti, Hemanth Shrinivas, Asif Qureshi and Parmeshwaree Bahadoorsingh for their help and support. Last but not the least I thank God Almighty and my family for having walked with me through each step that I took. xi INTRODUCTION CHAPTER ONE INTRODUCTION 1.1 Anaerobic Process Anaerobic biodegradation is a successful technology that has been used in industrial, food processing, and agricultural wastewater treatment for many decades. This is because anaerobic biological treatment systems offer a number of advantages over their aerobic counterparts. The operational costs associated with anaerobic systems are typically lower than with aerobic systems and anaerobic systems also generate less waste sludge. In addition, the biogas produced can be a potential energy source in an anaerobic treatment system. However, the application of anaerobic treatment systems is limited for low strength wastewaters in colder climates. In colder climates, the biomass growth yield and the growth rate are relatively low, resulting in a low net biomass production. The net biomass production must exceed the net biomass loss to the effluent for a biological treatment system to function properly. However, in a conventional anaerobic biological system, the net biomass loss to the effluent is governed by the relatively poor settling characteristics of the biomass. As a result, it is typically not possible to maintain a sufficiently large biomass concentration in an anaerobic system to effectively treat low strength wastewaters in colder climates. Also, the slower growth rates of the anaerobic microorganisms result in a relatively long hydraulic retention times (HRT) to prevent the biomass washout in a conventional completely mixed reactor (Wen et al, 1999) Therefore, an increase in reactor volume is required and the capital cost increases for the anaerobic reactor. Also, the anaerobic effluents, due to residual organic matter, usually need to be polished by a downstream aerobic biological reactor (Wen et al, 1999). 1 INTRODUCTION 1.2 Anaerobic Membrane Bioreactors (AnMBRs) In recent years, membrane technologies have been successfully implemented in aerobic biological treatment processes. As effluent discharge standards become more and more stringent and land use is becoming an increasing concern in many cities, membrane bioreactors are considered as plausible alternatives to conventional wastewater treatment processes (Ahn et al, 1999). The first full scale commercial aerobic MBR process appeared in North America in the late 1970s, and then in Japan in the early 1980s (Hu and Stuckey, 2006). Membrane bioreactors (MBRs) have some key advantages over conventional treatment processes that utilize secondary clarification. One of the important advantages is that the membrane component of the system can retain virtually all of the biomass within the bioreactor. Therefore, it may be possible to maintain an adequate biomass concentration by coupling a membrane to an anaerobic biological reactor to effectively treat low strength wastewater in colder climates. Also, the complete retention of biomass decouples the solids retention time (SRT) from the hydraulic retention time (HRT), allowing biomass concentrations to increase in the reactor and thus facilitating a smaller reactor and a higher organic loading rate (OLR) (Liao et al, 2006). However, over 98% of the membrane coupling has been done in aerobic processes and very limited application of MBRs has been seen in anaerobic processes (Hu and Stuckey, 2006). Conventional anaerobic biological treatment systems can effectively remove the bulk of the organic contaminants present in wastewater; however, they are typically not effective at removing residual levels of soluble and colloidal organic contaminants. Stuckey and Hu (2003) suggested that in an anaerobic membrane bioreactor (AnMBR), the residual organics could be retained in the system independently of the hydraulic throughput, enabling these contaminants to be hydrolyzed and biodegraded. The membrane modules can be used in the anaerobic process in different ways: as an "add-on" to existing anaerobic reactors such as the upflow anaerobic sludge blanket (USAB) reactor (Wen et al, 1999); as an external module with the biomass being 2 INTRODUCTION constantly recycled (side-stream) (Brockmann and Seyfried, 1996) or, with the membrane module submerged in the reactor itself. 1.3 Fouling of Anaerobic Membrane Bioreactors The widespread use of membrane technology has been limited due to the fouling of the membrane fibers. In general, fouling is a phenomenon which can be expressed as the increase in transmembrane pressure (TMP) or decrease in permeate flux due to accumulation of substances on membrane structures (Hong et al, 2002). The reduction in permeate flux is one of the most significant factors in deciding the economic feasibility of membrane processes. In wastewater treatment, membrane fouling in MBRs is referred as to the accumulation of the sludge foulants on the membrane structure due to the interactions between the membrane and the sludge to be. filtered (Liao et al, 2004). Membrane fouling is an inherent problem with membrane processes, which not only affects the long term operational stability, but also leads to significant operational costs due to increased membrane replacement frequency and added energy consumption. Therefore, a considerable amount of research and engineering effort has been devoted to understanding the mechanisms of membrane fouling and to work out fouling prevention and control strategies. Currently, most of the fouling research has been done in the drinking water area. Due to the limited use of membranes in wastewater treatment and the very complicated nature of membrane fouling, knowledge of fouling mechanisms, characteristics, and influencing factors in MBR processes is extremely limited. The fouling in an anaerobic MBR is composite fouling which can be further classified as biofouling, organic fouling and inorganic fouling (Liao et al, 2006). All these different types of fouling occur simultaneously, although the nature of fouling depends on membrane characteristics, sludge characteristics, environmental conditions, reactor design, and the operating strategy (Cho and Fane, 2002; Choo and Lee, 1996a, 1996b; Kang et al, 2002). Therefore, extensive research is required to study the fouling of membranes in anaerobic membrane bioreactors for long term stable membrane filtration and to benefit the increasing application of AnMBRs in the future. 3 BACKGROUND AND LITERATURE REVIEW C H A P T E R T W O B A C K G R O U N D AND L I T E R A T U R E REVIEW 2.1 Introduction The evolving health concerns and the development of upgraded and low cost membranes have increased the application of membrane technology dramatically in the field of environmental engineering (Tchobanoglous et al, 2003). In wastewater treatment processes, membrane technology has been used primarily in conjunction with aerobic biological treatment processes. In any membrane system, the permeate flux that can be maintained is one of the most significant factors that can affect the total cost of the membrane system and thus, the successful implementation of membrane systems. However, fouling is the major reason behind the permeate flux decline in MBRs and subsequent cleaning or replacement of the membranes increases the operational cost of the whole process. Therefore, considerable research has been done to identify, investigate and control the parameters that impact the permeate flux (or decrease in permeate flux) and fouling in the aerobic MBRs. Chang et al. (2002) have written an extensive review paper on the fouling in aerobic MBRs. On the other hand, very limited research has been done on fouling in anaerobic MBRs. The objective of this literature review is to summarize and critically evaluate the parameters that govern the successful implementation of anaerobic MBRs treating low strength municipal wastewater. In addition, this review will try to identify the causes behind the fouling of the anaerobic membrane bioreactors. 4 BACKGROUND AND LITERATURE REVIEW 2.2 Anaerobic Membrane Bioreactor System Configuration An anaerobic MBR system can be defined as a biological wastewater treatment system operating without oxygen and coupled with membrane units for solids-liquid separation. Membrane units in an AnMBR can operate either as external units or as submerged units, depending on the requirements of the process. The basic difference in principle is that the external membranes operate under pressure and the submerged membranes operate under vacuum. In an external system, the mixed liquor to be filtered is pumped from the bioreactor to the membrane as illustrated in Figure 2.1. The treated wastewater is collected on the permeate side of the membrane and the retained biomass is returned to the bioreactor. The circulation rate (i.e. cross-flow velocity through the membrane system) and operating trans-membrane pressure used in external systems are relatively high as compared to the submerged membranes. For an AnMBR with an external membrane, the cross-flow velocity and operating trans-membrane pressure typically range from 1 to 5 m/s and 30 to 100 psi (207 to 690 kPa) respectively (Berube et al., 2006). The external cross-flow membranes use high liquid velocities to minimize fouling and maintain high fluxes. For example, Brockman and Seyfreid (1996) reported that for every 1 m3 of anaerobic membrane permeation, 40-80 m3 of wastewater must pass over the membranes. The high pressure liquid recirculation in the external membrane results in a substantial decrease of the mixed liquor particle size in the reactor (Cho and Fane, 2002; Choo and Lee, 1998; Elmaleh and Abdelmoumni, 1997). Also, there are some reports which suggest that circulation of biomass through the pumps is the cause for the reduced biomass activity in anaerobic bioreactors (Brockman and Seyfreid, 1996; Ghyoot et al., 1999). However, different types of pumps used in various projects might have different effects on the biomass. According to Kim et al. (2001), positive displacement pumps impose higher stress on sludge floes than turbine-type centrifugal pumps. 5 BACKGROUND AND LITERATURE REVIEW Process gas Influent wastewater Excess sludge Completely mixed anaerobic bioreactor Solids recycle Filtered effluent Pump Figure 2.1. Schematic of a pressure driven external anaerobic MBR The amount of research work conducted for submerged AnMBRs is limited, as compared to external AnMBRs, although submerged membranes are a successful technology in aerobic MBRs (Stephenson, et al, 2000). Unlike the external membranes, membrane modules in submerged AnMBRs are located inside the reactor as illustrated in Figure 2.2. In submerged AnMBRs, the membranes are provided with gas sparging at the base to reduce the effect of fouling. The sparged gas bubbles entrain liquid upwards, creating a cross-flow along the membrane surface. For an AnMBR with a submerged membrane, the operating trans-membrane pressure has been reported to range from 3 to 15 psi (21 to 103 kPa) (Berube et al, 2006). The vacuum driven submerged membranes do not require the biomass to pass through the pumps and therefore, the biomass activity is maintained inside the reactor (Hernandez, et al, 2002). Also, the power use is much lower in the submerged membranes because the suction pump operates at much less pressure. The difference in the hydrodynamic conditions between a submerged AnMBR 6 BACKGROUND AND LITERATURE REVIEW and an external AnMBR may have great impact on the mixed liquor of the AnMBR. Chen et al. (2005) reported that the size of most of the suspended material present in the mixed liquor in an external AnMBR ranged from approximately 0.1 to 0.4 um, while that for a submerged AnMBR from approximately 50 to 500 um. —• Excess Process Gas Influent wastewater -Excess sludge Process gas Membrane module(s) Completely mixed anaerobic bioreactor Gas handling system Filtered effluent Gas -recyclings-Figure 2.2. Schematic of a vacuum driven submerged anaerobic MBR 2.3 Factors Affecting the Treatment Performance of Anaerobic MBR Systems The parameters such as reactor temperature, hydraulic retention time (HRT) and solids retention time (SRT) can significantly affect the performance of AnMBRs. However, these parameters are usually optimized for the biological component of the system, rather than the membrane component. Therefore, an optimal set-point in terms of the biological component of an AnMBR may result in a non-optimal set-point in terms of the permeate flux. Anaerobic processes are usually operated at mesophilic (35 °C) and thermophilic (55 °C) temperatures. These temperatures are very important for treatment of high solids content wastewaters like municipal wastewater sludges for which the SRT and the HRT 7 BACKGROUND AND LITERATURE REVIEW are usually the same and therefore, the increase in the reaction rates due to higher temperature results in reduction of size of reactor (Liao et al, 2006). The higher operating temperature in an AnMBR can have beneficial impact on the permeate flux by decreasing the viscosity (Berube et al, 2006). Hogetsu et al (1992) reported an increase in permeate flux of over 30% when the operating temperature was increased from 40 °C to 47 °C. Zoh and Stenstrom (2002) reported the same results while treating explosives process wastewater. The high heating requirements in AnMBRs can be met by using methane produced in the system. However, this is not possible in low organic content wastewaters like municipal wastewater due to low methane production and therefore, treatment must be carried out at ambient temperature (Liao et al, 2006). Consequently, the application of anaerobic treatment is more successful in warmer climates. The application of anaerobic processes is possible in ambient temperatures; though SRTs as much as double those of mesophilic operation may be required. Wen et al (1999) were able to operate an AnMBR successfully for an extended period of time with temperatures ranging from 14 °C to 25 °C. Kiriyama et al (1992) also reported successful operation of an AnMBR at a low temperature range of 20 °C to 25 °C. AnMBRs can retain solids at a different range of temperatures which increases the SRT in the anaerobic system and decreases the HRT of the system. For example, Wen et al (1999) were able to maintain an SRT of 150 d and an HRT of 4-6 h at 14-20 °C while treating screened municipal wastewater. The HRT of an AnMBR can significantly influence the capital cost of the whole process. This is because for a given influent composition, a higher organic loading rate (OLR) allows a shorter HRT and a smaller reactor. However, the HRTs used in AnMBRs have been higher than for non-membrane high rate anaerobic reactors (Liao et al, 2006). The complete solids retention possible in AnMBRs has not yet led to the expected decrease in HRTs (Liao et al, 2006). 8 BACKGROUND AND LITERATURE REVIEW 2.4 Factors Affecting the Membrane Performance of Anaerobic MBR System 2.4.1 Membrane material Membrane material plays a very important role in the performance of the membranes. Both inorganic and organic membranes have advantages and disadvantages. Ghyoot and Verstraete (1997) observed that the flux of a ceramic microfiltration membrane was 10-fold higher than the flux in a polymer ultrafiltration membrane when both membranes were producing permeate of similar quality. However, based on the life cycle analysis, the cost of an AnMBR with an inorganic membrane is almost twice that of an AnMBR with an organic membrane (Ghyoot and Verstraete, 1997). Also, the material used in the membrane can significantly affect the fouling mechanisms in an anaerobic MBR. According to Kang et al. (2002), fouling in an organic membrane can be attributed to the accumulation of cake layers on the membrane surface. In addition, Choo et al. (1996a) observed that the major constituents of the cake layer in an AnMBR were the biological/organic solids and struvite. However, Choo et al. (2000) reported that fouling of the organic membranes was predominately governed by the biological/organic interactions with the membrane, rather than by struvite deposition. Also, the resistance due to internal fouling, which is caused by the adsorption of soluble and/or particulate material within the pore structure of a membrane is significantly less than cake fouling in organic membranes (Choo and Lee , 1996a; Kang et al, 2002; Lee et al, 2001b) The fouling of inorganic membranes is mostly governed by internal fouling and a cake layer typically does not form on an inorganic membrane (Choo and Lee, 1996a; Kang et al, 2002). Both Choo et al. (2000) and Kang et al. (2002) reported the presence of struvite as a foulant in inorganic membranes. Yoon et al. (1999) used a scanning electron microscopy (SEM) technique to characterize fouling and did not observe any cake layer formation on the surface of an inorganic membrane. The SEM images also revealed the presence of white crystals, characteristic of struvite, in the pores of the inorganic membrane. However, Elmaleh and Abdelmounmni (1997) reported that the 9 BACKGROUND AND LITERATURE REVIEW formation of a cake layer was the principal mechanism behind the fouling of an inorganic membrane. 2.4.2 Membrane pore size The membrane pore size or molecular weight cut-off plays a significant role in the performance of an anaerobic MBR. Elmaleh and Abdelmounmni (1997) studied the effect of pore size on the steady state permeate flux in an anaerobic MBR. The highest permeate flux was achieved with a pore size of 0.45 um while filtering anaerobic mixed liquor. However, Elmaleh and Abdelmounmni (1997) observed that the optimal pore size was changed to 0.15 um when they filtered a mixed microbial population of methanogens. Therefore, it can be said that the optimal membrane size is dependent on the specific mixed liquor which is to be filtered through the membranes. Choo and Lee (1996b) reported that the optimal pore size for an anaerobic MBR was 0.1 um. Also, larger pore sizes or higher molecular cut-offs usually lead to increased permeate flux. Hernandez et al. (2002) found that the steady state membrane flux was about seven times higher for a pore size of 100 jum than for 10 jum using immersed membranes with a granular sludge. However, Wen et al. (1999) observed that the rate of fouling in an AnMBR was greater at a higher operating flux. 2.4.3 Hydrophobic nature and charge of membrane The hydrophobic nature of the membranes has been reported to have important impact on the performance of an AnMBR. Choo et al. (2000) observed that higher flux can be maintained when the membrane is hydrophilic in nature. According to Sainbayar et al. (2001), the permeate flux in a hydrophobic membrane can be improved through graft polymerization, in which hydrophilic functional groups are introduced on a membrane surface. However, Choo and Lee (1996b) reported contradictory results, as they found the fouling was lower in membranes that were more hydrophobic in nature. 10 BACKGROUND AND LITERATURE REVIEW The membrane surface charge is strongly impacted by the pH and the ionic strength of the mixed liquor. Therefore, the charge of the membranes may significantly impact the performance of AnMBR. Shimizu et al. (1989) observed that negatively charged inorganic membranes fouled less rapidly than neutral or positively charged membranes during the filtration of anaerobic broth. They attributed this phenomenon to a stronger electrical repulsion between negatively charged colloids in the mixed liquor and the membrane surface. However, when filtering protein solutions, Fane et al. (1983) reported that the impact of the membrane surface charge becomes negligible when the ionic concentration of the solution being filtered is high. 2.5 Operational Parameters Governing the Membrane Performance of Anaerobic MBR System 2.5.1 Permeate flux The permeate flux is the most important parameter for successful implementation of anaerobic membrane technology. Higher permeate flux in a smaller membrane surface area significantly reduces the cost of the whole process. Theoretically, if the membranes operate below critical flux, then the trans membrane pressure (TMP) remains constant throughout the operation. Nevertheless, fouling should always be expected in membrane operations, and there should be a balance between a high permeate flux and the long filtration runs that must be achieved to maximize the total permeate volume prior to membrane cleaning. The critical permeate flux that can be achieved in a wastewater treatment application is dependent on the sludge characteristics and the sludge concentration (Liao et al, 2006). The typical membrane flux for an AnMBR is 10-40 L/m2hr in a temperature range of 20 °C to 50 °C (Liao et al, 2006). However, the design membrane flux is dependent on membrane composition, pore size, porosity, hydrophobicity, surface charge, anaerobic sludge properties, and operational and environmental conditions (Liao et al, 2006). 11 BACKGROUND AND LITERATURE REVIEW 2.5.2 Trans membrane pressure (TMP) The functional relationship between the permeate flux and the TMP is governed by different mechanisms at a high or low TMP. At a relatively low TMP, the permeate flux is governed by the TMP. Under such conditions, the permeate flux increases linearly with the TMP and the permeate flux is not significantly impacted by the cross flow velocity (Beaubien et al, 1996). According to Beaubien et al (1996), the permeate flux is also impacted by low mixed liquor suspended solids (MLSS) concentration (< 2500 mg/L) under pressure limited conditions. The permeate flux is independent of pressure at a relatively high TMP and is governed by the mass transfer of materials away from the membrane surface. Under these conditions, the permeate flux is dependent on the cross flow velocity and the MLSS concentration (Beaubien et al, 1996). Beaubien et al (1996) observed a linear increase in the permeate flux with an increase in the cross flow velocity. However, the increase of permeate flux was lower at higher MLSS concentrations. This discrepancy can be attributed to the higher rate of mass transfer towards the membrane and the increase in viscosity of the mixed liquor at higher MLSS concentration. The flux to TMP relationship is usually complicated by fouling. According to Beaubien et al. (1996), at a higher TMP, the fouling rate increases even below the critical flux, and therefore flux will decrease with time. Therefore, they suggested that for relatively high pressure systems, it is possible to maximize the permeate flux while minimizing the membrane fouling. The optimal operating pressure in an AnMBR can be calculated using the Equation 2.1. 12 BACKGROUND AND LITERATURE REVIEW Where, the APOPT is the optimal trans-membrane pressure, R'm is the resistance due to membrane-solute interactions (i.e. resistance due to pore plugging and adsorption), and fi is a mass transfer parameter (Beaubien et al, 1996). According to Equation 2.1, when fouling is governed by the accumulation of cake layers on the membrane surface, the TMP is relatively low. However, when the fouling is governed by internal fouling (i.e. resistance due to membrane-solute interactions is high), the optimal TMP is relatively high (Beaubien et al, 1996). 2.5.3 Hydrodynamics The manipulation of cross-flow velocity and gas sparging techniques are used extensively in external and submerged membranes, respectively, to provide high shear conditions at the membrane surface. Choo and Lee (1998) observed a significant decrease in the resistance due to concentration polarization and the resistance due to cake layer formation by increasing the cross-flow velocity. However, according to Choo et al (2000), a plateau was reached at a Reynold's number of 2000, beyond which there was no reduction of resistance with increasing in cross-flow velocity. Elmaleh and Abdelmoumni (1997) also reported that the total fouling resistance could be reduced to zero when the cross-flow velocity in a tubular membrane system exceeded 3 m/s. Many studies have observed that the increase in the cross-flow velocity lead to an increase in permeate flux (Beaubien et al, 1996; Imasaka et al, 1989; Strohwald and Ross, 1992). In particular, Beaubien et al. (1996) reported a significant increase in flux from 15 L/m2hr to 35 L/m hr, when the cross-flow velocity was increased from 1.1 m/s to 2.2 m/s. However, Bourgeous et al. (2001) noted that the increase of permeate flux as a result of a high cross-flow velocity comes at a cost. They reported that although an increase in the cross-flow velocity from 1 to 2 m/s increased the permeate flux by 20%, the power cost of the system increased by 58%. Also, the high cross-flow velocity required to generate high shear conditions can generate large axial pressure gradients, resulting in a non-uniform TMP in tubular membranes (Lee et al, 1995). As a result, some part of the membrane can be under non-optimal TMP conditions. 13 BACKGROUND AND LITERATURE REVIEW A high cross-flow velocity can negatively impact the performance of an AnMBR. Brockmann and Seyfried (1996) and Ghyoot and Verstraete (1997) reported that the biomass activity could be significantly affected by the shear induced due to high cross-flow velocities. Also, Choo and Lee (1996a) suggested that the high cross-flow velocities can significantly increase the cell lysis and thus result in a decrease in the overall activity of the biomass in an AnMBR. Kim et al. (2001) suggested that in an aerobic MBR, high shear conditions could result in the release of high concentrations of extracellular polymeric substances (EPS) into the bioreactor. A high concentration of soluble EPS (also known as soluble microbial products (SMP)) has been reported to be a major contributor towards fouling in membranes (Rosenberger et ai, 2006; Jarusutthirak and Amy, 2006). McMahon et al. (2001) also suggested that higher mixing intensities could break up the microbial aggregates and thus inhibit the interspecies substrate transfer which is essential for the stable operation of anaerobic systems. However, according to Elmaleh and Abdelmoumni (1997), anaerobic biosolids are not impacted greatly by the high cross-flow velocity when compared to aerobic biomass. Gas sparging techniques are used to disrupt the formation of a cake layer on the surface of submerged membranes. Gas sparging produces high shear on the membrane surface much like the shear produced by the cross-flow velocity in the external membranes. Air is used for sparging in aerobic MBRs. However, in anaerobic MBRs, air sparging for a longer duration can result in non-anaerobic conditions that can significantly reduce the activity of the acid-forming microorganisms in the system (Lee et al, 2001b). Stuckey and Hu (2003) effectively used the headspace gas as a source of relatively inert gas for continuous sparging of the submerged AnMBR system. They observed a decrease in TMP when the gas sparging was increased at a constant permeate flux. However, a plateau was reached beyond which no additional significant reduction in the TMP could be achieved by increasing the extent of gas sparging (Stuckey and Hu, 2003). Kayawake et al. (1991) reported that the permeate flux in an anaerobic MBR with a ceramic membrane system could be doubled by adapting a head space gas sparging system. Nitrogen was also used for sparging in the anaerobic MBR system (Imasaka et al., 1989). 14 BACKGROUND AND LITERATURE REVIEW 2.5.4 Mixed liquor suspended solids The negative impact of an increase in MLSS on permeate flux has been reported in some studies (Beaubien et al, 1996; Harada et al, 1994; Oh et al, 2004; Pillay et a/., 1994). The major reason behind the decrease in permeate flux is the increased opportunity for cake formation and fouling due to higher MLSS concentration (Liao et al, 2006). However, some authors have suggested that a constant permeate flux can be maintained at a "critical" MLSS concentration in an AnMBR (Liao et al, 2006). Ross et al (1990) reported that in an anaerobic digestion system, a flux of 1000 L/m2hr was maintained up to an MLSS concentration of 40 g/L, and the flux was reduced to 400 L/m2hr when the MLSS concentration was 60 g/L. Similar results were reported by Strohwald and Ross (1992), whereby they observed a rapid decline in permeate flux at MLSS concentration > 20 mg/L. Nonetheless, the relation between MLSS and permeate flux is dependent on the design of the reactor. Reactors with suspended growth biomass, such as the continuous stirred-tank reactor (CSTR), expose the bulk biomass to the membrane modules (Liao et al, 2006). The increase of SRT in a completely mixed AnMBR and the corresponding increase in MLSS concentration will improve the biological performance of the reactor, but will decrease the permeate flux of the system. Therefore, in such systems there should be a balance between the MLSS concentration and the critical flux. 2.5.5 Membrane cleaning Membrane modules from AnMBRs are chemically cleaned by caustic and acidic solutions after being irreversibly fouled in the bioreactor. Caustic solutions are considered to be effective at removing organic/biological foulants from a membrane surface, whereas the acidic solutions are considered to be effective at removing the inorganic foulants from a membrane surface (Lee et al, 2001b). However, a number of studies have reported that it is possible to consistently recover the permeate flux in an organic membrane by back-flushing the membrane with acidic solution (Choo et al, 15 BACKGROUND AND LITERATURE REVIEW 2000; Kang et al, 2002), or with a caustic solution followed by a cleaning with acidic solution (Lee et al, 2001b). Also, according to some studies, it is not possible to recover the permeate flux consistently in an inorganic membrane by back-flushing an acidic solution (Choo et al, 2000; Kang et al, 2002; Yoon et al, 1999). Relaxation is another successful technique that has been used extensively to increase the permeate flux in membrane systems. In this approach, the filtration process is interrupted periodically by reducing the trans membrane pressure to zero and the permeate flux recovers after every relaxation cycle. Wen et al. (1999) investigated a number of relaxation scenarios with permeation and relaxation times ranging from 2 to 8 minutes and 0.5 to 2 minutes, respectively. Their results suggested that the permeate flux was highest at intermediate permeate times (i.e. four minute) and the intermediate relax times (i.e. one minute). 2.6 Membrane Fouling in Anaerobic M B R Systems Fouling in membrane bioreactors is due to the physiochemical interactions between the biofluid and the membrane modules (Chang et al, 2002). The substances that cause fouling of the membranes are collectively referred as the foulants. Membrane fouling is a time-controlled process which can be either reversible or irreversible (Wiesner and Aptel, 1996). In a reversible fouling process, the foulants are weakly attached to the membrane substratum and as a result, they can be removed by following an appropriate physical washing protocol. Irreversible fouling results from stronger bonds between the foulants and the membrane materials (i.e. chemical bond) and generally, irreversible fouling can only be removed by chemical cleaning. However, restricting the categorization of membrane fouling to reversible or irreversible is somewhat simplistic (Chang et al, 2002). For example, gel layer formation over a membrane surface is often irreversible although it is notionally reversible since it forms a cake layer (Chang et al, 2002). Also, some kinds of fouling by pore blocking and adsorption may be partially reversible depending on the strength of adhesion and the vigor of the physical wash (Chang et al, 2002). The fouling in an anaerobic MBR can be broadly divided into three categories, i.e. biofouling, organic fouling and inorganic fouling. 16 BACKGROUND AND LITERATURE REVIEW 2.6.1 Biofouling Biofouling in membranes is caused by the interactions between the membrane surface and the components of the biological treatment broth (Liao et al, 2006). Pore clogging, sludge cake formation and adsorption of extra-cellular polymeric substances (EPS) are the three types of biofouling encountered in membranes (Liao et al, 2006). Pore clogging/ blocking can be defined as the clogging of the membrane pores by cell debris and colloidal particles that are of similar size to the membrane pore size. In this mechanism, the particles with a size comparable to the pore size accumulate in pores during permeation and thus, reduce the surface area for filtration. Choo and Lee (1996b) observed that the colloids, not the dissolved and the cellular fractions, were the main foulants in microfiltration and ultrafiltration membranes. In addition, pore clogging has been observed to be greater in pressure driven external membranes because pump-induced shear stress decreases the average particle size and liberates colloids that can lead to pore clogging (Choo and Lee, 1998; Kim et al, 2001). The observed improvement in membrane flux after backwashing provided further proof that pore clogging was involved in membrane fouling (Liao et al, 2006). Sludge cake formation in membranes usually occurs when the shear stress at the membrane surface is not adequate to remove the accumulated solids (Liao et al, 2006). Choo and Lee (1996a) and Kang et al. (2002) reported the presence of a thick cake layer composed of biomass and struvite on polymeric membrane surfaces causing major hydraulic resistance. The study by Cui et al. (2003), in which they observed an increase in the permeate flux due to an increase in shear at the membrane surfaces by gas bubbling, provides further evidence that sludge cake formation is involved in membrane fouling. The extent of biofouling due to sludge cake deposition will depend in part on the concentration of suspended material that is brought to the membrane (Liao et al, 2006). Membranes used with a continuously-stirred tank reactor (CSTR) configuration have a greater chance of being fouled due to cake deposition because the membrane modules are 17 BACKGROUND AND LITERATURE REVIEW exposed to high suspended solids concentrations. Configurations that do not present the full MLSS concentration to the membranes should be less challenged by cake deposition (Liao et al, 2006). The third mechanism of biofouling is caused by the accumulation and adsorption of EPS on the membrane and pore surfaces. In recent years, a lot of research has been done on the effect of EPS on membrane performance (Cho and Fane, 2003; Jarusutthirak and Amy, 2006; Nagaoka et al, 1998; Rosenberger et al, 2006). It was found that the higher the bound EPS content in the mixed liquor, the more rapidly membrane fouling proceeded. Therefore, the bound EPS content of the mixed liquor was suggested as a probable index for membrane fouling (Chang and Lee, 1998). However, with the improved understanding of fouling mechanisms, the soluble EPS or soluble microbial products (SMPs) are now most often associated with increased fouling in membranes (Jarusutthirak and Amy, 2006; Rosenberger et al, 2006). Soluble microbial products (SMPs) are defined as "the pool of organic compounds that are released into the solution from substrate metabolism (usually with biomass growth) and biomass decay" (Barker and Stuckey, 1999). The SMPs are classified into two groups based on the bacterial phase from which they are derived: growth-associated products and non-growth-associated products (Barker et al, 2000). The growth-associated SMP, also called utilization-associated products (UAP), are directly produced from biomass growth and substrate metabolism (Jarusutthirak and Amy, 2006). Non-growth-associated SMP, also known as biomass-associated products (BAP), occur as a result of biomass decay and cell lysis during endogenous decay. Typically, cells consume electron-donor substrate to build active biomass and also at the same time, they produce bound extracellular polymeric substances and UAP (Jarusutthirak and Amy, 2006). Bound EPS are hydrolyzed to biomass-associated product (BAP), while the active biomass undergoes endogenous decay to form residual dead cells (Jarusutthirak and Amy, 2006). Therefore, the SMP is also known as soluble EPS (Laspidou and Rittmann, 2002). The production rate of SMP is proportional to the concentration of biomass due to the release of organic materials from cell lysis (Jarusutthirak and Amy, 2006). The major 18 BACKGROUND AND LITERATURE REVIEW constituents of the SMP are proteins, polysaccharides and some humic-like materials. The SMPs have been found to comprise the majority of the soluble organic matter in effluents from biological wastewater treatment processes (Jarusutthirak and Amy, 2006). Also, the presence of SMP in wastewater affects discharge levels of chemical oxygen demand (COD), biochemical oxygen demand (BOD), total organic carbon (TOC) and is a big constraint to wastewater reuse. In addition, biological systems with longer SRTs tend to produce more SMPs due to higher accumulation of biomass in the system (Jarusutthirak and Amy, 2006). 2.6.2 Organic fouling and inorganic fouling Organic fouling in membrane bioreactors is caused by the accumulation and adsorption of organic constituents on the membrane surface. The relatively higher effluent COD concentrations in AnMBR systems as compared to the aerobic MBR systems may increase the contribution of organic fouling in AnMBRs (Liao et al, 2006). Also, higher OLRs in the MBRs will lead to higher residual CODs and lower membrane fluxes (Hernandez, et al, 2002). Furthermore, it is the absolute residual COD and not the COD removal efficiency that affects fouling in AnMBRs (Liao et al, 2006). Powdered activated carbon and zeolites have been added into AnMBRs to absorb soluble organic compounds and hence reduce organic fouling (Choo et al, 2000; Park et al, 1999). Inorganic fouling occurs due to the accumulation of inorganic colloids and crystals on the membranes and pore surfaces. Choo and Lee (1996a) reported that struvite contributed significantly to the fouling of AnMBRs. The amount of struvite that precipitated could be estimated based on a mass balance analysis of the concentrations of magnesium in the influent and effluent of an AnMBR and also the concentration of ammonia and phosphate in the mixed liquor (Choo and Lee, 1996a). AnMBRs can be more susceptible to inorganic fouling than aerobic MBRs because of greater opportunity for pH shifts due to carbon dioxide partial pressure changes and the production of high ammonia and phosphate concentrations (Liao et al, 2006). However, the extent of the effect of struvite as a foulant depends on the type of membrane used in an AnMBR. As 19 BACKGROUND AND LITERATURE REVIEW discussed in Section 2.4.1, the fouling of organic membranes is predominately governed by biological/organic interactions with the membrane. On the other hand, struvite is the major contributor towards internal pore fouling in inorganic membranes. 2.7 Conclusions Based on the extensive literature review presented above, it can be said that both external membrane and submerged membrane can be successfully used in anaerobic processes. However, the vacuum-driven submerged membrane alternatives show a lot of promise as compared to external membrane processes. This is because high energy consumption is one of the biggest limiting factors in external membranes. Also, the use of head space gas in submerged anaerobic membranes can be a very successful technology. However, application of anaerobic membrane technology has been severely limited by the fouling problem encountered in the process. A lot of research has been done in recent years to understand the parameters governing fouling of aerobic membranes. Although the microbiological conditions of an anaerobic MBR is different than those of an aerobic MBR, research on fouling mechanisms in aerobic MBRs can be a starting point for future research on fouling of anaerobic membranes. In addition, more work should be done to develop and optimize hydraulic and operating conditions to minimize fouling in an anaerobic MBR. 20 BACKGROUND AND LITERATURE REVIEW 2.8 Research Objectives The overall objective of the present research was to address the anaerobic membrane bioreactor technology implementation issues and the reason behind the rapid fouling of membranes in an anaerobic MBR. The specific objectives were 1. to assess the treatment performance of a submerged membrane AnMBR treating low strength municipal wastewater at an ambient temperature, 2. to assess and compare the membrane filtration characteristics of the mixed liquors generated in AnMBRs to the mixed liquor generated in an aerobic MBR operated with the same influent wastewater, and 3. To identify the mechanisms governing the fouling of AnMBR as compared to the aerobic MBR. 21 MATERIALS AND METHODS CHAPTER THREE MATERIALS AND METHODS 3.1 Bench Scale Anaerobic Submerged Membrane at the U B C Pilot Plant The design of the submerged membrane AnMBR system was based on information gained from previous operation of aerobic MBRs at the UBC pilot plant and with suggestions provided by membrane manufacturers. A schematic of the submerged membrane AnMBR system is shown in Figure 3.1. Screen Feed pump Off-gas Temp, controller Gas counter - 0 — Gas sampling j » ^ Vacuum pump for gas LJ sparging Suction pump Solenoid valve Permeate tank L Figure 3.1. Schematic of the submerged membrane anaerobic bioreactor system. The submerged membrane AnMBR system (Figure 3.1) consisted of a completely mixed anaerobic bioreactor, equipped with a timer-controlled feed pump and one 22 MATERIALS AND METHODS continuously operated permeate pump connected to two identical hollow fiber submerged membrane modules. The dimension of the reactor was 20 cm (Length) x 17 cm (Width) x 120 cm (Depth) and the reactor was made from plexiglass material. The feed pump of the submerged AnMBR operated in a semi-continuous mode in 5 minutes ON and 10 minutes OFF cycles. The feed rate for this project was 60 L/day and the liquid volume inside the reactor was 24 L. A liquid level sensor was installed outside the bioreactor to control the permeate discharge and maintain the bioreactor liquid level within high and low alarm set-points. A chemical feed pump was installed for continuous, slow rate addition of acetate into the system. The submerged membrane bioreactor temperature and pH (influent and effluent) were measured continuously during the whole project period. Gas production rates were measured by allowing the off-gas to pass through a wet-tip gas meter. Air/gas sparging systems were installed with the membrane modules to control fouling and biomass attachment to the membrane surfaces. Head space gas was withdrawn from above the liquid level in the reactor and was pumped to the gas sparging systems located beneath each membrane unit. The sparging intensity for each membrane module was set to approximately 9 L/minute. A collapsible gas buffer bag was installed in the reactor off-gas system to prevent the generation of a vacuum in the off-gas system. The submerged membrane AnMBR was operated under constant permeate flux and variable TMP conditions. A relatively constant flux (3-5 L/m hr) was maintained using a variable speed Masterflex peristaltic pump. However, an absolutely constant permeate flux could not be consistently maintained due to extensive fouling of the membranes. The permeate pump operated with a 5 minute ON and a 1 minute OFF sequence to introduce relaxation cycles to enhance the permeate flux. The membranes were chemically cleaned when the TMP increased to a value greater than approximately 6 to 8 psi (41-55 kPa) as recommended by the membrane manufacturer. Initially, two used hollow fiber US Filter membrane modules (Memcor) were installed in the reactor. New US Filter membrane modules replaced the old modules on day 67 because the used modules were not able to sustain an adequate flux. The properties of the US Filter Membrane module are summarized in Table 3.1. 23 MATERIALS AND METHODS Table 3.1. Description of membrane modules used in anaerobic submerged MBR. Parameter Membrane Description Polymeric Surface area (m2) Nominal Permeate Flux (L/m hr) Nominal Pore size (um) Dimensions Pressure Mixing Memcor (US Filter) Polyvinylidene fluoride (PVDF) submersible module, outside/in hollow fiber 0.5 25 0.05 75 mm dia x 550 mm long 3 to 4 psi (21-28 kPa) vacuum to maximum 10 psi (69 kPa) Gas sparging Source: Memcor Australia PTY Limited, South Windsor, NSW, Australia. 3.2 Initial Testing and Acclimatization of the Anaerobic Reactor Initial testing of the system was completed with tap water to detect any unforeseen problems in the design of the reactor. After sorting out all the design problems, it was decided that the system would be seeded with sludge from an external membrane anaerobic reactor operating in the pilot plant. However, during the transfer of the inoculum, the anaerobic biomass was exposed to air and the subsequent sludge activity in the submerged reactor was very low. As a result, the submerged reactor was seeded again on 7th October, 2005 (first day of operation), with anaerobic digester sludge from a local full scale municipal wastewater treatment plant. The MLSS concentration of the digester seed sludge was 13,900 mg/L. A total of 10 L of digester sludge was added to 40 L of wastewater in the period between day 1 and day 6 of the operation. During the acclimatization period, 2 L of sludge was added to 16 L of wastewater on the first day of operation. On day 2 and day 4, 2 L of sludge was added to 3 L of the wastewater. No sludge was added on day 3 of the operation. Finally, a total of 4 L of the digester sludge was added to 18 L of wastewater on day 5 and day 6 of the operation. The temperature was between 30 °C to 36 °C and the pH was approximately 7.5 during the seeding period. Following the second seeding, the reactor was operated in a fill and draw mode for four days during the acclimatization process. The target HRT for this study was about 8 to 10 24 MATERIALS AND METHODS hrs. On day 4 of the operation, semi-continuous feeding was initiated with a flow rate of 6 L/day. Subsequently, the flow rate was increased linearly from 6 L/day, reaching the target 60 L/day on day 15 of the operation. The influent was continuously supplemented with approximately 300 mg/L of acetate during the acclimatization process 3.3 Membrane Clean ing The submerged membrane modules were chemically cleaned periodically to recover the permeability of the membranes. The cleaning procedures employed for the submerged membrane AnMBR are presented below. During the cleaning period, the reactor was shut down temporarily and the membrane modules were taken off-line for chemical cleaning. The membranes were transferred into a submerged membrane chemical cleaning module which consisted of a chemical cleaner reservoir and a pump which drew permeate through the membrane and back to the reservoir. The chemical cleaning sequence consisted of the following steps. 1. The membrane modules were soaked in tap water for 20 minutes. 2. After cleaning with tap water, the membranes were soaked in a 1000 mg/L bleach solution at pH 10.5 while sparging with air for 10 minutes. The modules were subsequently permeated with the bleach solution for another 10 minutes while sparging. 3. The membranes were soaked in tap water for 5 minutes and then filtered in tap water for approximately 5 minutes while sparging (i.e. until pH was approximately 7). 4. After cleaning with tap water, the membranes were soaked in a 2000 mg/L citric acid solution at pH 2 while sparging with air for 10 mins. The modules were subsequently permeated with citric acid solution for another 10 minutes while sparging. 5. Steps 1 to 4 were repeated two additional times to complete the chemical cleaning of the anaerobic membranes. During the cleaning period, oxygen entered into the AnMBR system. Once the membrane modules were reinstalled in the reactor after cleaning, the reactor was purged with nitrogen gas to remove the remaining oxygen from the system. 25 MATERIALS AND METHODS 3.4 Monitoring 3.4.1 Routine monitoring The sampling and analysis schedule of the submerged AnMBR is summarized in Table 3.2. The schedule was developed to obtain a consistent and comprehensive evaluation of the performance of the AnMBR. The routine monitoring program began on day 14 of the operation. All the onsite readings (i.e. temp, pH, permeate flux) were usually taken in the mornings and only once per day. pH was measured using a portable WP pH Testr BNC (r), which was calibrated a minimum of three times a week. Samples were usually transported to the laboratory within 20 minutes for chemical analysis. The results from the monitoring program are presented in Chapter Four. Table 3.2. Monitoring program to track ongoing process performance Monday Tuesday Wednesday Thursday Friday Saturday Sunday Temp X X X X X X X pH X X X X X X X Gas Count X X X X X X X Flux X X X X X X X COD (influent and effluent) X X X VFA (influent and effluent) X X X Gas Composition X X X TSS/VSS X 3.5 Analytical Methods The samples collected from the submerged AnMBR were analyzed for parameters chemical oxygen demand (COD), volatile fatty acid (VFA), mixed liquor suspended solids (MLSS)/ mixed liquor volatile suspended solids (MLVSS), gas composition and 5-day biochemical oxygen demand (BOD5). Grab samples of the influent and the effluent were collected from the submerged AnMBR on every Monday, Wednesday and Friday. Samples taken to the laboratory for analysis were transported in a timely fashion (usually 26 MATERIALS AND METHODS within 20 minutes) and preserved in accordance with Standard Methods (APHA et al, 1999). Normally, 25 mL of influent and 25 mL effluent samples were collected in plastic bottles for COD and VFA analysis. During the last period of the project, 1 L of influent and 1 L of effluent were also collected for BOD5 analysis. MLSS/ MLVSS samples (50 mL) were collected from the reactor on every Monday by using the sampling port in the reactor. All the solids analyses were conducted immediately after sample collection. Since 50 mL of mixed liquor was collected every week for MLSS/ MLVSS analysis, the wasting did not greatly affect the solids retention time (SRT) of the submerged AnMBR system. 3.5.1 Chemical oxygen demand (COD) The COD samples were prepared and analyzed in general accordance with 5220 D. Closed Reflux, Colorimetric Method in Standard Methods (APHA et al, 1999). The samples were analyzed in Hach DR/2000 Direct Reading Spectrophotometer. 3.5.2 Volatile fatty acids (VFA) The influent VFA samples were filtered through 0.45 um filter papers before being transferred to 2 mL glass vials. The effluent samples were directly transferred to the glass vials because the AnMBR membrane pore size was smaller than the pore size of the filter paper normally used for sample filtration. Both the influent and effluent VFA samples were preserved by adding 2% phosphoric acid solution (H3PO4). The samples were analyzed by gas chromatography (GC), using an HP 5890 Series II Gas Chromatograph FID (Flow Injection Detector).The description of the GC used for the analysis was as follows. Column- HPFFAP 25 m x 0.32 mm x 0.25 um film Detector- Helium (He) Carrier, 7 psi (48 kPa) Head Pressure Injection Temperature- 175° C Detector Temperature- 250° C Oven Initial Temperature- 130° C 27 MATERIALS AND METHODS Oven Final Temperature- 150 C 3.5.3 Gas composition The head space gas was collected in a Septum-Port Gas Sampling Tube for gas composition analysis. The gas samples were analyzed in a Fisher-Hamilton Gas Partitioner to determine the percentage of methane, carbon dioxide, oxygen and nitrogen present in the sample. The Fisher-Hamilton Gas Partitioner was calibrated by injecting samples of pure gas into the partitioner. The description of the Fisher-Hamilton Gas Partitioner used for the analysis was as follows. Detectors- Thermal conductivity cell with four tungsten filaments Standard Columns-Column No. 1- 6 ft x 1/ 4 in aluminum packed with 30% Di-2-ethylhexylsebacate (DEHS) on 60-80 mesh Chromosorb B Column No.2 - 6 1/2 ft x 3/16 in aluminum packed with 60-80 mesh molecular sieve 5A Cell Temperature- Approximately 70 °C Column Temperature- Ambient 3.5.4 M L S S / M L V S S The suspended solids concentration of the anaerobic MBR was not estimated using the methods described in Standard Methods (APHA et al, 1999), because the anaerobic solids were very fine in nature and it was difficult to filter the suspended solids with 0.45 um filter paper. Therefore, the anaerobic mixed liquor suspended solids concentration was estimated using a centrifuge method. In this method, a fixed volume (V) of mixed liquor was dried at 105°C for 24 hours. The weight of the mixed liquor residue after drying was recoded as SSiotai. An 8 mL aliquot of the anaerobic mixed liquor was centrifuged at 20,000 x g for 30 minutes in a Thermo IEC Multi High Speed Centrifuge. Then, volume (V) of the centrate was dried at 105°C for 24 hours. The weight 28 MATERIALS AND METHODS of the mixed liquor after drying was recorded as SScentrifuge. The MLSS was calculated by using the following formula: MLSS - (SSiot.1- SScentrifuge)/ V (3.1) The original mixed liquor (SSiotai) and the centrifuged mixed liquor (SScentrifuge) residues were then ignited at 550°C for 24 hours. The weight of the samples was recorded as NVSrotai (non volatile solids) and NVScemrifuge respectively after heating. The MLVSS of the mixed liquor was calculated by using the following formula: MLVSS = MLSS - (NVSrotai- NVScemrifuge)/ V (3.2) 3.5.5 BOD 5 test BOD5 tests were conducted in the last stages of the project by following 5210 B. 5-Day BOD Test method in Standard Methods (APHA et al, 1999). The BOD samples were analyzed using a YS1 52 Dissolved Oxygen Meter. 3.6 Periodic Monitoring In addition to routine monitoring of the parameters, additional parameters were also monitored on a periodic basis during on-line fouling/filtration tests. A description of on-line filtration tests and the parameters is presented below. 3.6.1 On-line filtration tests The on-line filtration tests were conducted to assess the filtration characteristics of anaerobic mixed liquor under a relatively constant permeate flux. These tests were also conducted on aerobic mixed liquor from a membrane enhanced biological phosphorus removal (MEBPR) process operating at the pilot plant. 29 MATERIALS AND METHODS The pilot scale membrane enhanced biological phosphorus removal (MEBPR) reactor consisted of three reaction zones as shown in Figure 3.2. The three zones of the MEBPR were operated under anaerobic, anoxic and aerobic conditions with liquid volumes of 0.23, 0.59 and 1.31 m3 respectively. The submerged membrane Ml was the main membrane unit of the MEBPR process and it consisted of two coupled membranes. The surface area of each module was 12 m and the membranes had a nominal pore size of 0.04 um. The two hollow fiber modules were supplied by Zenon Environmental Inc, Oakville, Ontario. The membranes operated in the permeation mode for 9.5 minutes and the back flush mode for 0.5 minute in each cycle of 10 minutes. The flux of the membrane system (Ml) was approximately 10 L/m2hr throughout the period of study. The aerobic zone was aerated with a flow rate of 0.34 m3/ min and the air was supplied in a mode of 20 seconds ON and 10 seconds OFF. Two Memcor submerged membrane modules (M2) (exactly the same modules as those of the AnMBR) were also installed in the aerobic compartment on day 135 of the present study. In addition, the membrane modules in the AnMBR and the M2 membrane systems were operated with the same permeate flux (15 L/m2hr), relaxation cycles (5 minutes ON and 1 minute OFF) and sparging intensities (9 L/minute per module). The membrane system M2 will be referred to as the aerobic MBR in the following discussion. More detailed description of the MEBPR process can be found in the doctoral thesis of Alessandro Monti titled "A Comparative Study of Biological Nutrient Removal Processes with Gravity and Membrane Solids-Liquid Separation". 30 MATERIALS AND METHODS Air sparging (A,) P, Permeation Permeate I tank Influent U- Sewer Figure 3.2. Schematic of the MEBPR system Ai, A2: Air sparging system of membrane modules Mi and M 2 respectively Pi_7: Pumps in the MEBPR system The Memcor membrane modules of the anaerobic MBR were cleaned and installed on day 133 for on-line studies. The influents to the AnMBR and the aerobic MBR were both drawn from the same wastewater source. Also, the operating permeate flux selected for the on-line filtration tests was substantially higher than that used in AnMBRs prior to the tests (3-5 L/m2'hr - see Section 3.1). A higher flux was selected for the on-line tests so that both the AnMBR and aerobic MBR on-line systems would be subjected to commercially relevant operating conditions. Discussions with the membrane supplier (US Filter) indicated that an operating permeate flux in the order of 15 to 30 L/m hr is typical for full-scale aerobic MBR systems treating municipal wastewaters. A recycling tube was installed to maintain high permeate flux in the anaerobic MBR. Initially, a low permeate flux was employed in the AnMBR since a previous study in this program indicated that operating at a permeate flux greater than approximately 5 L/m hr resulted in rapid fouling under anaerobic treatment conditions. To assess the fouling propensities of each mixed liquor, the TMPs of the anaerobic membranes and the aerobic membranes were monitored regularly over time. 31 MATERIALS AND METHODS The membranes used in on-line fouling studies were ultimately characterized by scanning electron microscopy (SEM)-Energy Dispersive X-ray (EDX) microanalysis, and by extraction of membrane foulants and chemical analysis of the extracted fouling materials. For these studies, test fibers were cut randomly from the Memcor membrane modules and the ends of the cut fibers that remained on the modules were sealed with epoxy. 3.6.2 Membrane sampling The fouled test fibers were rinsed with deionized distilled water (DDW) to remove the scum. The scum was removed from the fibers because it was believed that since the scum was loosely attached to the fibers, and it did not contribute towards irreversible fouling of the membranes. Virgin membrane fibers were used as controls and these were first soaked in a 200 mg/L chlorine (OCT) solution overnight and then cleaned with DDW to remove the preservatives prior to analysis. The virgin membranes were supplied by Zenon and these had a similar composition to as that of the fouled Memcor membranes. 3.6.3 Scanning electron microscopy (SEM)-Energy dispersive x-ray (EDX) analysis The samples for Scanning Electron Microscopy (SEM)-Energy Dispersive X-ray (EDX) were prepared by adopting the SEM microwave processing protocol. (1) The membrane fibers were fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer at 28° C at power level 1 (about 100W) in a Pelco laboratory microwave operating on a 2 mins ON, 2 mins OFF and 2 mins OFF vacuum mode. This step was repeated one more time without changing anything. (2) The fibers were rinsed with cacodylate buffer inside the Pelco microwave at 28° C at power level 2 (about 200W) for 40 seconds. This process was repeated three times for proper rinsing of the membranes. (3) The buffer rinsed samples were post fixed with 1% osmium textroxide in 0.1 M cacodylate buffer at 28° C at power level 1 (about 100W) in a Pelco laboratory microwave operating on a 2 mins ON, 2 mins OFF and 2 mins OFF vacuum mode. This step was repeated one more time 32 MATERIALS AND METHODS without changing anything. (4) The samples were rinsed with distilled water. (5) The samples were dehydrated with 50% ethanol at 28° C at power level 2 (about 200W) for 40 seconds. The samples were further dehydrated with 70% and 90% ethanol at 28° C at power level 2 (about 200W) for 40 seconds. Finally, the samples were dehydrated three times with 100% ethanol without changing anything. (6) The dehydrated samples were then dried with hexamethyldisilazane (HMDS) and coated with gold (Nanotech sputter coater) for better imaging in SEM analysis. The processed samples were then placed under Hitachi S4700 and Hitachi S-2600M for SEM imaging and EDX microanalysis respectively. 3.6.4 Extraction of membrane foulants The fouled membranes were subjected to basic extraction and acidic extraction. For basic extraction, the membrane fibers from the anaerobic MBR and the aerobic MBR were soaked in 9.0 mL of 0.1 M NaOH solution and sonicated (Aquasonic Model 550HT, VWR) at 40°C for 1 hour. The fibers were then removed and the solution (i.e. extract) was filtered through a 0.45 um filter paper and then neutralized with 3 mL of 0.15 M H2SO4. All the extraction tests were done in triplicate (i.e. using three fibers). For acidic extraction, the procedure was similar to basic extraction except that the membrane fibers were soaked in 9 mL of 0.5 M H2SO4 and subsequently neutralized with 3 mL of 3 M NaOH solution. 3.6.5 Chemical analysis of extracted fouling material The extracted samples were analyzed for total Kjeldahl nitrogen (TKN), total phosphorus (TP), total organic carbon (TOC) and soluble microbial product (SMP) contents. TKN/TP samples were digested and analyzed in general accordance with 4500-Norg D. Block Digestion and Flow Injection Analysis method in Standard Methods (APHA et a l , 1999). A 5 mL aliquot of the extract was digested with 5 mL of digestion 33 MATERIALS AND METHODS reagent (200 mL H2SO4 and 134 g K2SO4 made up to 1 L with distilled water) for 3.5 hr at 140°C and 3.5 hr at 365°C in Technicon BD-40 block digestor. Following digestion, 30 mL of distilled water was added to the digestion vials and mixed with the samples. The samples were analyzed in Lachat Quickchem 800 Flow Injection Analyzer. The TOC samples were analyzed by following 5310 B. High Temperature Combustion Method of Standard Methods (APHA et al, 1999). The samples were placed into COD vials, acidified to pH ~ 3 with 50% H3PO4 and stored at 4°C prior to analysis. The TOC samples were analyzed by using a Shimadzu™ TOC-500 analyzer. The extracted foulant samples were analyzed for soluble microbial product (SMP) content, in terms of carbohydrate, protein and humic-like substance content. The analysis of the carbohydrate content of the extracts was completed following the method suggested by Frolund et al. (1996) and the analysis of protein and humic-like substance content of the extracts was determined following the modified Lowry method (Frolund et al, 1996). The concentration of carbohydrates was determined by the anthrone method, in which glucose was used as the standard (Frolund et al, 1996). Anthrone reagent was prepared by dissolving 0.125% (w/v) anthrone in 94.5% (v/v) H2S04. Once the reagent was prepared, 0.40 mL of acid/basic extract was mixed with 0.80 mL of the anthrone reagent. The mixture was then immediately placed in an oven set at 100 °C for 15 minutes and subsequently cooled in a water bath at 4°C for 5 minutes. The anthrone method is a colorimetric method and therefore, the absorbance of the mixture was measured at a 625 nm wavelength. The absorbance was measured using a Unicam UV 300 UV-Visible Spectrometer. Proteins and humic-like compounds were analyzed using the modified Lowry method, in which bovine serum albumin (BSA) and humic acids were used as the standards respectively (Frolund et al, 1996). In this method, five reagents were prepared prior to the analysis. Reagent 1 contained 143 mM NaOH and 270 mM Na2C03, Reagent 34 M A T E R I A L S A N D M E T H O D S 2 contained 57 mM C11SO4 and Reagent 3 was a 124 mM Na-tartrate solution. Reagent 4 was prepared by mixing of Reagents 1 to 3 in the proportion 100:1:1. Reagent 5 was prepared by diluting Folin reagent with distilled water in the ratio of 5:6. Once the reagents were prepared, 0.50 mL of foulant extract was mixed with 0.70 mL of Reagent 4 using a vortex mixer. The resulting solution was then mixed with 0.1 mL of Reagent 5 using the vortex mixer and was left at room temperature for 45 minutes. The absorbance of the mixture was measured at 750 nm in a Unicam UV 300 UV-Visible Spectrometer and was recorded as Atotai. In the modified Lowry method, proteins and humic compounds interfere with each other during analysis. When the mixture is prepared without the addition of CuS04, the color development is attributed to humic-like compounds and chromogenic amino acids. In that case, the color developed by BSA decreased to 20% but no decrease was observed for humic acids. Therefore, the content of proteins and humic-like compounds was calculated using the following equations (Zuohong Geng, Environmental Engineering Group, Department of Civil Engineering, UBC, Vancouver, B.C., pers. comm.). Atotai = Aproteins + Ahumic-like Ablind = 0.2Aproteins + Ah umic-like Aproteins = 1.25 (Atotai — Ablind) Ahumic-like = Ablind — 0.2 Aproteins Where Atotai is the total absorbance of the mixture with addition of CuS04, Abiindis the total absorbance of the mixture without addition of CUSO4, Ahumic-iike is the absorbance due to humic-like compounds, and Aproteins IS the absorbance due to proteins. The concentrations of proteins and humic-like substances in the extract were calculated by fitting the values of Aproteins and Ahumic-iike into the standard curves of BSA and humic-like acids, respectively. (3.3) (3.4) (3.5) (3.6) 35 MATERIALS AND METHODS 3.7 Mixed Liquor Characterization Extraction of extra-cellular polymeric substances (EPS) from mixed liquor was carried out by following the cation exchange resin method (Frolund et al, 1996). This method works on a principle which assumes that EPS is mainly bound to cell aggregates through the bridging of divalent cations such as calcium and magnesium. These divalent metal ions can be removed by using cation exchange resin (CER), which will weaken the EPS matrix and thus, EPS will be more easily released from sludge floes into the liquid phase (Flemming et al, 2000). The step-by-step procedure of EPS extraction was as follows. (1) Exactly 50.0 mL of activated sludge mixed liquor was centrifuged at 15,000 x g for 30 minutes at room temperature. The resulting supernatant was used for unbound or soluble EPS measurement. (2) The sludge pellets were washed with about 50 mL of EPS extraction buffer (2 mM Na3P04, 4 mM NaH2P04, 9 mM NaCI and 1 mM KCI at pH 7) by re-suspending the pellets in the buffer and centrifuging the suspension at 15,000 xg for 30 minutes. The supernatants of the washed sludge pellets were decanted from the centrifuge tubes and discarded. (3) The sludge pellets were transferred to 125-mL Erlenmeyer conical glass flasks after mixing with 20.0-30.0 mL of EPS extraction buffer. An additional 10.0 mL of the buffer was used to rinse each centrifuge tube and this was also transferred to the flask. (4) A measured amount of cation exchange resin (DOWEX Marathon C, 20-50 mesh, Fluka 91973) was added to the flask in the ratio of 60 g CER/g SS. (5) The flask was placed on a refrigerated incubator shaker (New Brunswick Scientific, Edison, NJ, USA) and agitated at 400 rpm at 4 °C for 2 hours. (6) After agitation, the sludge/CER mixture was immediately centrifuged at 12,000 *g for 20 minutes at room temperature. The final supernatant collected was sampled for bound EPS analysis. All the sludge samples were processed in triplicates. When the soluble and bound EPS samples were not analyzed immediately, they were stored at a temperature of -20°C. During the time of analysis, the frozen samples were thawed at 4 °C and centrifuged at 12,000 x g for 5 minutes to remove any remaining floe components. Finally, the bound and the soluble EPS samples from the anaerobic and aerobic mixed liquors were analyzed for carbohydrates, proteins and humic substances. 36 PvESULTS AND DISCUSSION CHAPTER FOUR RESULTS AND DISCUSSION 4.1 Assessment of the Operational Performance of Submerged AnMBR 4.1.1 UBC pilot plant influent wastewater characteristics The anaerobic MBR and the aerobic MBR were both operated at the UBC pilot plant. The source of the wastewater was the same for both the reactors as discussed in Section 3.6.1. The influent characteristics of the UBC pilot plant wastewater are summarized in Table 4.1. Table 4.1. UBC pilot plant influent wastewater characteristics. Parameter Value TSS (mg/L) 90.1 (±45) COD to t (mg/L) 258 (±59) COD s o l (mg/L) 186 (±64) Acetate (mg/L) 18.2 (±10) Propionate (mg/L) 5.2 (±6) Total VFAs (mg COD/L) 19.0 (±12) TKN (mg N/L) 35.9 (±11) NH4-N (mg N/L) 26.7 (±5) N03-N (mg N/L) Not detected TP (mg/L) 4.2 (±3) PO4-P (mg/L) 3.0 (±2) Temperature (°C) 20.2 (±2) pH 7.2 (±0.5) Notes: The influent characteristics were mean values over the period June 2005-May 2006 ± Corresponds to the standard deviation of the measurements made 37 RESULTS AND DISCUSSION 4.1.2 Description of operating conditions The submerged AnMBR was put into operation on October 7, 2005. Initially, the submerged membrane bioreactor operated in a batch feed mode for 3 days. The reactor started operating at a nominal HRT of 100 days on day 4 of the operation. Following a period of acclimatization of the anaerobic biomass, the AnMBR HRT was finally reduced to 10 hrs by the 15th day of operation (Figure 4.1). The flow rate was also maintained at 60 L/day from day 15 of to the end of the study period. As discussed in Section 3.2, the target HRT of 8-10 hrs was maintained throughout the entire period of the study. 110 100 80 s so 30 10 • 40 60 100 120 Days 140 180 200 Figure 4.1. Hydraulic retention times in the submerged membrane bioreactor. The process organic loading rate for the submerged membrane system is shown in Figure 4.2. The organic loading rate (OLR) was maintained in the range of 1 to 1.5 kg COD/m day for the entire project period. The variation in the OLR of the submerged AnMBR system was largely dependent on the influent wastewater COD concentration. In addition, this organic loading rate was achieved with 200 mg/L of acetate supplementation to the AnMBR system. 38 RESULTS AND DISCUSSION • • • • • 0 20 40 60 80 100 120 140 160 180 200 220 Days Figure 4.2. Organic loading rates in the submerged membrane bioreactor. The pH and the temperature variations in the AnMBR are shown in Figure 4.3 and Figure 4.4, respectively. Since small amounts of sodium bicarbonate were added to the influent wastewater on daily basis (a requirement of another research program), the pH values were consistently higher than 7. The pH values of the influent and effluent were almost same for this period of the study. The target temperature for AnMBR operation was 35°C. However, the heater couldn't heat up the reactor to 35°C and therefore, the reactor was totally insulated to hold the temperature. Even the insulation was unable to maintain a temperature of 35°C inside the reactor. Nonetheless, for most part of the project, the temperature was consistently around 32°C. As discussed in Section 3.1, new membrane modules were installed in the AnMBR on the 67th day of the operation. The gas sparging mechanism in the new modules was slightly different from that of the old modules. The gas sparging in the new modules appeared to result in less heat loss than previously and therefore, it became possible to maintain the temperature at around 35°C as presented in Figure 4.4. However, the temperature dropped slightly in the period between day 130 and day 160 of the operation. This was attributed to the introduction of a recycling tube on day 134 to 39 RESULTS AND DISCUSSION maintain high flux in the AnMBR for the on-line filtration studies. The recycling tube was not heat insulated and therefore, the heat loss from the liquid going through the tube cooled the reactor during this period. 8.0 7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2 5.0 <sxm-m<M-<m—or>aoo-«>—OJ>«DCOO-C« «£•-<*•-•»-<>—mt>oo m—o—o »-co <x> o-oo-o-mo —«» o o o o—o— -<sm-20 40 60 80 100 120 Days 140 160 180 200 220 I O pH effluent • pH influent j Figure 4.3. pH of the submerged membrane bioreactor influent and effluent. 40.0 n 35.0 E 3 e E 30.0 25.0 20.0 • 4 » 4 » • 0 20 40 60 80 100 120 140 160 180 200 220 Days Figure 4.4. Temperature of the submerged membrane bioreactor. 40 RESULTS AND DISCUSSION 4.1.3 Treatment performance The submerged membrane influent and effluent COD results are summarized in Figure 4.5. When the influent wastewater was supplemented with acetate, the average influent COD value was approximately 472 mg/L. However, the raw wastewater (without acetate supplementation) average influent COD value was approximately 258 mg/L. In addition, the raw wastewater average influent soluble COD value was approximately 137 mg /L. As shown in Figure 4.5, the permeate COD value was consistently lower than 80 mg/L during the whole project period. Therefore, the average total COD removal efficiency was approximately 85% for this project. The removal efficiency was calculated based on the permeate COD concentrations and the acetate-supplemented total influent concentrations. Table 4.2 demonstrates that when the COD removal efficiency was calculated using the raw wastewater influent total COD values (i.e. without acetate supplementation), the observed average removal efficiency was reduced to 71 %. 800 700 600 J 5 0 0 400 a o w 300 200 100 A A - A -A A A A « . . , A A A * A A A A A A * A * * H . .4A ™ A A — A — * -" n S O . O O p CO 20 40 60 80 100 120 Days 140 160 180 200 220 A Total Inf w/ A O USF Effl m Total Raw Inf o Sol Raw Inf Figure 4.5. Total and soluble COD concentrations in influent and permeates for the submerged membrane bioreactor. (w/A: influent with acetate supplementation) 41 RESULTS AND DISCUSSION Table 4.2. Measured COD removal efficiencies during the periods of operation with and without acetate supplementation. Parameter Average Standard Deviation With acetate supplementation Total COD removal (%) 85 6 Without acetate supplementation Total COD removal (%) 71 13 As illustrated in Figure 4.6, the permeate VFA concentration values were in the range of 6-8 mg/L throughout the project. The low concentration of permeate VFA in the current project suggests that VFA removal was almost complete in the submerged membrane bioreactor. Also, an average of 0.62 mg/L of propionic acid was measured in the permeate of the AnMBR during the duration the project. However, butyric acid was not detected in the permeate. The low concentration of VFAs in the submerged AnMBR permeate indicates that the anaerobic biological treatment process was functioning adequately in this study. The above observation suggests that the observed average of 71% COD removal is the best possible COD removal that can be achieved under the present study conditions. In addition, the average observed permeate COD concentration of 72 mg/L signifies the fact that AnMBR treatment alone probably cannot achieve a secondary quality effluent. 42 RESULTS AND DISCUSSION 1000 100 10 20 40 60 80 100 120 Days 140 160 180 200 220 A A Acetic • NA Acetic o USF Acetic Figure 4.6. Acetic acid concentrations in the submerged membrane bioreactor influent and permeate. (A: influent with acetate supplementation; NA: influent without acetate supplementation) The substantially high concentration of average permeate COD (i.e. 72 mg/L) in the submerged AnMBR indicates that a significant amount of the permeate COD was composed of material that is refractory to anaerobic treatment conditions applied in this study. Therefore, the permeate COD can be divided conceptually into two categories-volatile acid COD (or biodegradable COD), and anaerobically non-biodegradable COD (or "refractory" COD) (Dr. Eric Hall, Environmental Engineering Group, Department of Civil Engineering, UBC, Vancouver, B.C., pers. comm.). As shown in Table 4.3, the average refractory COD concentration was 66 mg/L for the present study. Table 4.3. Average influent COD, permeate COD, VFA-COD, "refractory" COD and removal efficiency. Parameter Value Average influent total COD (wastewater origin) (mg/L) 258 Average permeate total COD (mg/L) 72 Wastewater COD removal efficiency (%) 71 Average permeate VFA-COD (mg/L) 6 Estimated average "refractory" COD (mg/L) 66 43 RESULTS AND DISCUSSION The nature of residual material in the treated permeate was further investigated by conducting BOD5 analysis in the final stages of the study. Four sets of raw wastewater and AnMBR permeate samples were taken for analysis of BOD5 between day 155 to day 165 of the operation. For this study, the average raw wastewater (without acetate supplementation) BOD5 concentration was 158 mg/L and the average permeate BOD5 concentration was 56 mg/L (Table 4.4). From Table 4.4, it can be concluded that the estimated average BOD5 removal efficiency was 65% in the anaerobic MBR. This removal efficiency was probably the best BOD5 removal efficiency that could be achieved under the treatment conditions applied. Also, the estimated ratio of BOD5/COD was calculated to be 0.76 for the membrane permeate. This ratio indicates that much of the refractory COD (under anaerobic conditions) was in fact, rapidly biodegradable under aerobic conditions. Therefore, more research is needed to understand the composition of residual material in the permeate and hence, decide whether any improvement can be achieved in the current anaerobic biodegradation process, or whether aerobic post-treatment is compulsory to achieve significantly lower levels of permeate COD and BOD5. Table 4.4. Measurements of BOD5 in selected raw wastewater and permeate samples of the submerged membrane bioreactor. Parameter Value Average influent BOD5 (without acetate supplementation) (mg/L) 158(±18) Average Permeate BOD5 (mg/L) 56(±6) Wastewater BOD5 Removal Efficiency (%) 65 Estimated Ratio of Permeate BOD5/COD 0.76 Notes: ±: Standard deviation of analysis done in triplicate 44 RESULTS AND DISCUSSION 4.1.4 Biogas composition and production The measured biogas data for the anaerobic MBR are summarized in Figure 4.7. As expected, there was no oxygen inside the reactor during the entire project. An average of 5 to 10% CO2 and 70 to 85% of CH4 was measured in the anaerobic biogas. The irregularity in the percentage of nitrogen and methane (Figure 4.7) results from the opening of the reactor for maintenance work. Every time the reactor was opened, the reactor headspace was filled with nitrogen gas to prevent oxygen from contacting the mixed liquor. However, the lowest nitrogen content was around 10% by volume throughout this project. Similar results were also reported by Hu and Stuckey (2006) while treating dilute wastewater in a submerged anaerobic MBR. In normal anaerobic systems, denitrification does not occur unless NO3" or NO2" are present in significant quantities (Hu and Stuckey, 2006). Since the UBC pilot plant influent does not contain high amount of nitrate, it was concluded that nitrate was not the source for nitrogen production in the anaerobic MBR. However, nitrogen can enter into the reactor as a dissolved component of the influent wastewater at low temperature. At a higher temperature, the dissolved nitrogen leaves the reactor as a component of off-gas. The solubility of nitrogen in water is 16 mL/ 1,000 mL at 20° C. As mentioned earlier, the flow rate of the anaerobic MBR was 60 L/d. Therefore, the average dissolved nitrogen entering into the anaerobic MBR was approximately 0.96 L/d. According to Figure 4.8, the average measured biogas production was approximately 4.8 L/d. The ratio of dissolved nitrogen gas to the measured biogas production indicates that the average off-gas should have 20 % nitrogen. This observation might answer the steady presence of nitrogen in the AnMBR biogas (Figure 4.8). 45 RESULTS AND DISCUSSION 100% 90% 80% 70% ~ 60% V ex S 50% c I 40% a. 30% 20% 10% 0% N 2 injection N 2 injection J3i • • CD CD • " . . .. • n N, inicction " ^injection • • Q • • i—cr -cr—m-• NTlnjection • <h CD C P a o *>OCOOCOOC<*> Ao o cc? o o ° o ^ ° o o o ° ° oo ( ^ A A A AAAA**A6 . AA—AA-AAA AAA A A A / W - A A A f A A A A A - d f t A - f AAA AAA A - A A A - A - A - A - A A AAA o ° o 0 o 9 o —AAA AAA AAA-AA-A-A-A— 20 40 60 80 100 120 Days 140 160 180 200 220 o C02 A 02 o N2 • CH4 Figure 4.7. Measured percent composition of biogas from the submerged membrane bioreactor. In an anaerobic biological wastewater treatment system, the majority of the COD removed is converted to biogas containing CH4 and CO2. The anaerobic biogas usually contains 65 to 75% methane (Hall, 1992). However, the observed composition of biogas is dependent on factors such as process operating conditions, the characteristics of the wastewater and the influent COD concentration. The measured and expected gas production rates for this study are summarized in Figure 4.8. The average expected biogas production rate was approximately 12 L/d and the measured biogas production was approximately 4.8 L/d. The expected gas production was estimated by using the following formula: Expected Gas ( T o t a l i n f l u e nt COD - Permeate COD) = Q x Total Biogas Yield x ^ '-Production 1000 (4.1) where the expected gas production was in L/day, Q, the influent wastewater flow rate was in L/day, the COD in mg/L and the total biogas yield was assumed to be 0.50 46 RESULTS AND DISCUSSION L/g COD (Dr. Eric Hall, Environmental Engineering Group, Department of Civil Engineering, UBC, Vancouver, B.C., pers. comm.). 20 18 16 1? 14 •o =ll2 c o '•g 10 3 •o o a -a—a— ~n o—a .an n a a • D o • D O • a a - D - o -o a — - . ^ : v - « • . ^ - ^ o , , , , 1 — , V 0 20 40 60 80 100 120 140 160 180 200 220 Days i Measured Production • Expected production | Figure 4.8. Measured and expected total biogas production rates in submerged membrane bioreactor. The measured biogas production rates were low from day 67 to day 120 as can be seen in Figure 4.8. This is because the reactor was opened numerous times for maintenance and on-line filtration studies. The measured gas production rates were essentially constant after this period, when the reactor remained closed. Figure 4.8 also indicates that the measured gas production rates were always lower than the expected values. This inconsistency can be explained by two important factors. First, when dilute wastewaters are treated anaerobically, a considerable amount of the CO2 and a smaller amount of CH4, leave the process as dissolved components of the effluent. The higher loss of C0 2 is observed in an AnMBR because C0 2 quickly reaches equilibrium in the bulk solution by forming bicarbonate and then leaves the reactor as a dissolved component of the effluent (Hu and Stuckey, 2006). This results in the reduction 47 RESULTS AND DISCUSSION of the measured biogas production rates and enrichment of CH4 in the biogas, relative to CO2. The unusually high percentage of CH4 in relation to CO2 in Figure 4.9 supports the fact that a substantial amount of the CO2 exits in effluent. 100% 90% 80% ^ 70% - 60% u w> 3 50% B £ 40% 30% 20% 10% 0% ODTJtP • OGtE DO •_• n Q o D C p a rj_oa r f Pq3DQ_ r n j 0 <&> -o— c-• o o * y * <#> -+- H 13- -+-• CH4 OC02 0 20 40 60 80 100 120 140 160 180 200 220 Days Figure 4.9. Adjusted percentage composition of CH 4 and CO2 in submerged membrane bioreactor biogas. Note: Gas composition data has been adjusted by eliminating the N2 component. Hu and Stuckey (2006) reported that only 83 % of the theoretical methane yield (0.395 m3 CH 4 /kg COD removed at 35 °C) was observed in a submerged AnMBR treating dilute wastewater. The solubility of methane in water is 15 mL/ 1,000 mL at 1 atm and 35 °C (Hu and Stuckey, 2006). Therefore, the average methane loss in the AnMBR effluent was 0.9 L/d. In Figure 4.10, the measured and the total expected methane production values were calculated by multiplying the fraction of methane in the biogas (Figure 4.7) by the measured biogas and the total expected biogas production values respectively (Figure 4.8). The total measured methane production value was calculated by adding the estimated dissolved methane in the effluent, to the measured biogas methane production values. Based on Figure 4.10, it can be concluded that the total measured methane values were relatively close to, but still less than, the total expected methane values. This phenomenon may not be significant in high strength 48 RESULTS AND DISCUSSION wastewater, because the volumes of biogas produced in such systems would greatly exceed the solubility of CH 4 and C O 2 . 20 -T 18 -16 0 20 40 60 80 100 120 140 160 180 200 220 Days • Measured methane production A Total measured methane production (with dissolved methane in effluent) • Total expected methane production Figure 4.10. Measured, total measured (including dissolved methane in the effluent) and expected total methane production rates in submerged membrane bioreactor. The second important factor in the biogas production inconsistency is most likely associated with the presence of particulate COD in the influent wastewater. In an anaerobic MBR, all of the particulate COD is retained by the membrane modules. It is hard to know how much of particulate COD was converted to CH 4 and C O 2 , and it can be expected that at least some portion of the particulate COD was biodegradable. However, the calculation of expected biogas production assumes that all of the particulate COD is biodegradable and hence, this can lead to an overestimation of the amount of biogas produced in the AnMBR. 49 RESULTS AND DISCUSSION 4.1.5 Bioreactor suspended solids concentration Figure 4.11 shows the measured suspended solids concentrations in the submerged membrane bioreactor over the course of this project. There was no sludge wasting from the reactor during the entire project other than through sample collection. Therefore, the mixed liquor suspended solids concentration (MLSS) reached 14,000 mg/L by day 62 of operation. However, a lot of mixed liquor was unintentionally lost during the installation of the new membrane modules on day 67 of the operation. At the end of the study period, the MLSS concentration recovered to the target range of 8,000-10,000 mg/L. It is believed that most of the suspended solids accumulation was due to filtration and retention of influent particulate material. Only 50 mL of the sludge was removed from the AnMBR for sampling every week. In this project, the mixed liquor volatile suspended solids (MLVSS) concentration was observed to be approximately 80 % of the MLSS concentration. 16000 J 14000 "Si ~12000 B O | 10000 | 8000 o w 6000 a tr if 4000 u S 2000 0 A A A A A A A A a * A A A A A A A A A A A , A A A * A A A A A * A A A A A A A A A A A A A A A 20 40 60 j A MLVSS centrifuge A MLSS centrifuge | Figure 4.11. Submerged membrane bioreactor sludge concentration. 80 100 120 140 160 180 200 220 Days 50 RESULTS AND DISCUSSION 4.1.6 Membrane flux and trans-membrane pressure (TMP) The flux and TMP profiles of the submerged AnMBR are shown in Figure 4.12 and Figure 4.13 respectively. As mentioned earlier, two sets of Memcor hollow fiber membranes were used in this project. The two old Memcor modules were replaced on day 67 because they were unable to sustain a constant flux of 5 L/m2 hr in the reactor. The flux was essentially steady in the new modules until the flux was increased to 15 L/m hr on day 134 for the on-line filtration studies. The submerged membrane bioreactor was operated in a constant flux and variable pressure mode for this study. However, (as shown in Figure 4.12) it was not possible to maintain a steady flux of 15 L/m2hr in the AnMBR due to membrane fouling. The trans-membrane pressure increased linearly and the membranes were cleaned after the TMP reached the range of 6-9 psi (41-62 kPa). 20.0 18.0 16.0 14.0 ^•12.0 s _10.0 = 8.0 u. 6.0 4.0 2.0 0.0 Flux increased for on-line study 1 -New-membranes-installed o * Chemical cleaning of membranes — • • « • — • -La* iChemical cleaning of membranes — I 1 — 100 120 Days 20 40 60 80 140 160 180 200 220 Figure 4.12. Flux of the submerged membrane bioreactor. 51 RESULTS AND DISCUSSION Figure 4.13. Trans-membrane pressure (TMP) of the submerged membrane bioreactor. The permeate flux and TMP profiles indicate that the permeate flux that could be maintained in the submerged AnMBR system was 5 L/m2 hr. However, as discussed in Section 3.6.1, an operating permeate flux in the order of 15 to 30 L/m2hr is typical for full-scale aerobic MBR systems treating municipal wastewaters. As shown in Figure 4.12, it was not possible to sustain such high permeate flux in the submerged AnMBR. Therefore, membrane flux became the limiting factor for submerged AnMBR operation. The on-line membrane filtration tests were conducted to determine the factors behind the low sustainable flux of the anaerobic membrane as compared to those reported for the aerobic membrane bioreactors. 4.2 Assessment of Membrane Filtration Characteristics in Submerged AnMBR The on-line filtration tests were conducted in the submerged AnMBR and in the aerobic compartment of the membrane enhanced biological phosphorus removal (MEBPR) process at the pilot plant. Two Memcor (M2) membrane modules with the same characteristics as that of the modules in the submerged AnMBR were installed in 52 RESULTS AND DISCUSSION the MEBPR aerobic zone to create similar conditions for on-line filtration studies (Section 3.6.1). The permeate flux was set at 15 L/m2'hr and the HRT was 10 hrs for both the systems. However, the MLSS concentration was different in the processes, i.e. approximately 5000 mg/L in the MEBPR process and 8,000-10,000 mg/L in the AnMBR process. The results from the on-line filtration tests are presented below. 4.2.1 Flux and permeability profile of anaerobic MBR and aerobic MBR The flux and TMP were measured during the on-line tests to estimate the rate of fouling in the anaerobic MBR and the aerobic MBR. The operating permeate flux of 15 L/m2hr was selected for both the AnMBR and aerobic MBR systems so that they would be subjected to conditions similar to those applied in full scale aerobic applications. However, it was relatively difficult to maintain the target flux in the anaerobic MBR as shown in Figure 4.14. 20 4 2 0 -I , , , • 1 , 1 , 0 10 20 30 40 50 60 70 80 Days - • - Aerobic MBR flux - • - Anaerobic MBR flux Figure 4.14. Flux of the aerobic MBR and the anaerobic MBR during the on-line filtration tests 53 RESULTS AND DISCUSSION The permeability profiles of the aerobic MBR and the anaerobic MBR are shown in Figure 4.15. Figure 4.15 indicates that the anaerobic MBR fouled more rapidly than the aerobic MBR. The TMP profile in Figure 4.13 also indicates that the submerged AnMBR fouls rapidly when the flux is approximately 8-10 L/m hr. This result was expected since sustainable operating flux for an air-sparged submerged hollow fiber aerobic MBR typically ranges from 15 to 30 L/m hr. 5 i o -i , , , , , r , —: 1 0 10 20 30 40 50 60 70 80 Days Aerobic MBR Permeability —•— Anaerobic MBR Permeability Figure 4.15. Permeability profile of the aerobic MBR and anaerobic MBR during the on-line filtration tests. The comparative study of permeability for a selected operating permeate flux indicated that the anaerobic membrane TMP increased rapidly as compared to the aerobic membrane TMP. The rapid increase in the TMP of the anaerobic MBR was thought to be caused by a higher accumulation rate of foulants on the membranes. Therefore, more experiments were conducted to identify and characterize the parameters responsible for rapid increase of trans-membrane pressure of the AnMBR. 54 RESULTS AND DISCUSSION 4.2.2 Scanning electron microscopy -Energy dispersive X-ray analysis The difference in the rate of decrease of permeability of the anaerobic MBR and the aerobic MBR was further investigated by SEM-EDX microanalysis. The SEM-EDX microanalysis is the most direct way of analyzing fouling in membrane fibers. To understand the fouling phenomenon better, virgin membranes were also analyzed by the SEM-EDX method. SEM images of a typical virgin membrane at different magnifications are presented in Figure 4.16. From Figure 4.16 (c) and Figure 4.16 (d), it can be observed that the virgin membrane surface had a homogeneous, cross linked and porous structure. The cracks in the membrane surface in Figure 4-16 (c) were thought to be artifacts of the sample preparation procedure. 55 RESULTS AND DISCUSSION (c) (d) Figure 4.16. SEM images of virgin membrane surfaces at different magnifications (45, 2.5K, 10K and 25K respectively) 56 RESULTS AND DISCUSSION Figure 4.17 and Figure 4.18 present SEM images (different magnifications) of fibers from the anaerobic MBR (operating day 205) and aerobic MBR respectively. The fibers were cut when the membrane modules were thought to be completely fouled in the MBRs. The images indicate that the aerobic membranes supported microbial colonies (rod shaped structures), whereas hardly any microbial growth was found in the case of the anaerobic membranes. Liao et al. (2004) reported a significant presence of microbial colonies on membranes in aerobic bioreactors. Nonetheless, high shear caused by the gas/air sparging made it difficult for colonies to attach and grow on gas-sparged membrane fibers. However, in the present study, the presence of microbial colonies on aerobic MBR membranes was much more significant than for the anaerobic counterpart. The anaerobic membranes were covered by relatively dense material as can be seen in Figure 4.17 (b) and Figure 4.17 (c). On the other hand, the aerobic membranes were covered with porous structures. The anaerobic membrane images demonstrate that there were very few pores that could be seen in a completely fouled membrane. The aerobic membrane images in Figure 4.18 show that the foulant materials did not cover the pores completely. The porous structures in aerobic membranes were probably due to the accumulation of foulants in the membrane surface under continuous suction. This accumulation was more severe under anaerobic conditions, leading to total blocking of the pores. The dense accumulations might explain the rapid fouling in the anaerobic membranes as compared to the aerobic membranes. 57 RESULTS AND DISCUSSION (c) (d) Figure 4.17. SEM images of anaerobic membrane surfaces at different magnifications (50,2.5K, 10K and 25K respectively) 58 RESULTS AND DISCUSSION (c) (d) Figure 4.18. SEM images of aerobic membrane surfaces at different magnifications (50, 2.5K, 10K and 25K respectively) 59 RESULTS AND DISCUSSION The Energy Dispersive X-ray (EDX) technique was used to analyze the type of foulants accumulated on the membrane fibers. The results of the EDX analysis are summarized in Table 4.5. The analysis indicates that the membranes themselves were mostly comprised of carbon and fluorine. Since the membrane fibers were made from polyvinylidene fluoride (PVDF) material, high percentages of carbon, fluorine and oxygen were expected in the EDX analysis. However, the percentage of carbon in a completely fouled anaerobic membrane and aerobic membrane was not much higher than the percentage of carbon in the clean membrane. Also, the percentage of fluorine was almost constant for the membranes. So, it can be suggested that the fouled membrane surfaces were not completely covered by a surface foulant layer. Since the online tests were performed by high gas/air sparging intensity, the high shear stress on the surface may be the reason for less surface fouling in the membranes. However, this is a contradictory observation to Figure 4.17 in which it can be clearly seen that the anaerobic membranes were densely covered with foulants. Table 4.5. Typical EDX analysis of clean membrane, anaerobic MBR membrane and aerobic MBR membrane Material Clean membrane concentration (wt %) Fouled AnMBR membrane concentration (wt %) Fouled aerobic MBR membrane concentration (wt %) Test 1 (8.3 kPa) Test 2 (23.5 kPa) Test3 (FF) Fully fouled samples Carbon 45.33 (± 0.65) 46.65 (± 1.16) 47.23 (1±.68) 47.84 (±5.5) 45.68 (± 2.90) Oxygen 8.12 (± 0.40) 5.94 (± 1.55) 5.30 (±0.20) 5.80 (±4.8) 3.42 (± 1.35) Fluorine 43.75 (± 1.06) 45.80 (±2.51) 47.00 (±1.56) 44.35 (±15) 49.37 (± 1.83) Nitrogen Not detected Not detected Not detected Not detected 1.81 (±3.14) Notes: Tests 1, 2 and 3 (for anaerobic MBR) were performed at 8.3 kPa, 23.5 kPa and 57.2 kPa (fully fouled) trans-membrane pressure respectively Tests for aerobic membranes were performed when the fibers were completely fouled ± Corresponds to the standard deviation of the measurements made Only compounds that were detected using EDX microanalysis are presented 60 RESULTS AND DISCUSSION A number of studies have reported that struvite (NH4MgP04,6H20) is a significant component of the foulant layer on a membrane surface in an anaerobic membrane bioreactor (Choo and Lee, 1996; Choo et al. 2000). However, there was no detectable presence of phosphorus (P) or magnesium (Mg) in the EDX analysis of the anaerobic and aerobic membrane fibers. As presented in Figure 4.18, the surface foulant layer of the aerobic MBR appeared to contain a substantial amount of the microbial colonies. However, as presented in Table 4.5, nitrogen accounted for less than 2% of the surface foulant material, suggesting that bacterial colonies accounted for a maximum of 15% of the surface foulant layer. The EDX technique did not clearly indicate the parameters causing fouling of the membranes. Though the results may suggest that the major foulants are of organic origin, it did not give much idea about the type of organic material causing fouling of the membranes. Therefore, EDX analysis was not found to be a very effective technique for the fouling study of membranes. 4.2.3 Chemical analysis of extracted foulant material The role of foulants in flux decline of the anaerobic membrane and aerobic membrane was further analyzed by chemical extraction and the results are summarized in Table 4.6 and Table 4.7. The chemical extraction tests were conducted on day 139, day 156 and day 205 of the operation. As shown in Table 4.6 and Table 4.7, the concentrations of TOC, Total Phosphorus (TP), TKN and soluble microbial products (SMPs) was higher in basic extraction as compared to the acidic extraction. These results indicate that the basic extraction was more effective than acidic extraction in TOC, TP, TKN and SMP extraction studies. The amount of TOC and TP present on the anaerobic membrane was higher than on the aerobic membrane as shown in Table 4-6. However, the amount of TOC was 61 RESULTS AND DISCUSSION much higher than TP in one meter length of fouled fiber for both the reactors. Higher concentration of TP in Test 1 (at 18.3 kPa TMP) of the anaerobic membranes might indicate the presence of struvite as a foulant in the fibers. However, the TP concentration was low in Test 3 (fully fouled membranes) of the anaerobic membranes. Also, the absence of phosphorus and magnesium in EDX analysis suggests that struvite was not the major foulant accumulating on the anaerobic membranes. The TKN concentrations of the acidic and basic extracted foulants for both the anaerobic MBR and the aerobic MBR were similar. Also, the TKN concentrations of the acidic and basic foulants were relatively low. Table 4.6. TOC, TP and TKN analysis of extracted foulant material TOC (ug/m) TP (ug/m) TKN (ug/m) Test 1 Test 2 Test 3 Test 1 Test 2 Test 3 Test 1 Test 2 Test 3 Acidic extraction Anaerobic MBR 3249 (±244) 1872 (±244) 3576 (±50) 267 (±89) 121 (±37) 122 (±91) ND ND ND Aerobic 1922 2425 2275 21 26 17 10 12 14 MBR (±353) (±617) (±341) (±5) (±10) (±4) (±8) (±7) (±3) Basic extraction Anaerobic 5580 2720 4339 1881 898 77 22 6 217 MBR (±612) (±681) (±156) (±846) (±405) (±38) (±2) (±2) (±82) Aerobic 4690 3438 2704 5 12 14 16 24 25 MBR (±853) (±216) (±226) (±4) (±6) (±3) (±19) (±29) (±3) Notes: Tests 1, 2 and 3 (for anaerobic MBR) were performed during experimental runs at 8.3 kPa, 23.5 kPa and 57.2 kPa (fully fouled) trans-membrane pressure respectively ±: Standard deviation of analysis done in triplicates ND: Not detected Tests for aerobic membranes were performed when the fibers were completely fouled The soluble microbial products (SMPs) extracted from the anaerobic and aerobic membrane systems are summarized in Table 4.7. The amount of carbohydrates was 62 RESULTS AND DISCUSSION slightly higher on the aerobic membrane than on the anaerobic membrane. However, the protein and humic-like substances contents were higher on the anaerobic membrane. It can be concluded from the chemical extraction tests that the major extractable foulants were organic compounds like carbohydrates, humic-like substances and proteins. These high molecular weight compounds are the major components of SMPs. Also, it can be concluded from Table 4.6 and Table 4.7 that the nitrogen-containing compounds and phosphorus-containing compounds were present in small amounts when compared to the organic compounds. So, soluble microbial products might be the reason behind the additional fouling of the anaerobic membranes. Table 4.7. SMP analysis of extracted foulant material Carbohydrates (|xg/m) Proteins (ug/m) Humic substances (ug/m) Test 1 Test 2 Test 3 Testl Test 2 Test 3 Testl Test 2 Test 3 Acidic extraction Anaerobic MBR 439 (±161) 1040 (±485) 1297 (±137) ND ND 387 (±189) ND ND 260 (±15) Aerobic MBR 631 (±102) 1417 (±583) 1566 (±101) ND ND 121 (±29) ND ND 70 (±6) Basic extraction Anaerobic MBR 1438 (±330) 1364 (±882) 1570 (±84) NA 1338 (±387) 314 (±62) NA 94 (±35) 786 (±142) Aerobic MBR 1550 (±825) 1579 (±829) 1748 (±86) NA 223 (±45) 302 (±18) NA 559 (±27) 749 (±26) Notes: Tests 1, 2 and 3 (for anaerobic MBR) were performed during experimental runs at 8.3 kPa, 23.5 kPa and 57.2 kPa (fully fouled) trans-membrane pressure respectively ±: Standard deviation of analysis done in triplicates ND: Not detected NA: Results not available due to contamination of samples during analysis Tests for aerobic membranes were performed when the fibers were completely fouled 63 RESULTS AND DISCUSSION 4.3 EPS Analysis of Anaerobic MBR and Aerobic MBR Mixed Liquors The role of extra-cellular polymeric substances (EPS) in fouling of the membranes has attracted a lot of attention from researchers in recent years (Cho and Fane, 2003; Jarusutthirak and Amy, 2006; Nagaoka et al, 1998; Rosenberger et al, 2006). Therefore, in the present study, the bound and the soluble EPS of the anaerobic and aerobic MBR mixed liquors were investigated to find out the possible reason for the rapid fouling of the anaerobic membranes. The particulate-bound carbohydrates, proteins and humic-like substances analysis results for both the mixed liquors are summarized in Figure 4.19, Figure 4.20 and Figure 4.21 respectively. 100 ••— Aerobic mixed liquor •«— Anaerobic mixed liquor 80 H 0 145 160 180 201 230 Operating days Figure 4.19. Bound carbohydrates concentration in mixed liquor 64 RESULTS AND DISCUSSION 100 80 Aerobic mixed liquor Anaerobic mixed liquor c '53 oo Z £« ™ o c oa 60 40 20 A 0 Operating days Figure 4.20. Bound proteins concentration in mixed liquor o 3 •= > O 00 -a c 3 o CQ 100 80 60 40 A 20 Aerobic mixed liquor Anaerobic mixed liquor 180 201 230 Operating days Figure 4.21. Bound humic-like substances concentration in mixed liquor 65 RESULTS AND DISCUSSION 200 150 -•— Aerobic mixed liquor -•— Anaerobic mixed liquor C/3 0. > 00 e 3 o o H 100 50 0 H T 1 1 < 1 1 1 145 160 180 201 230 Operating days Figure 4.22. Total bound EPS concentration in mixed liquor The specific concentration of carbohydrate and protein bound to suspended material was higher in the aerobic mixed liquor as compared to the anaerobic mixed liquor. However, humic-like substances content was higher for the anaerobic mixed liquor. Figure 4.22 illustrates the total amount of bound EPS present in the mixed liquors of both MBRs. The humic-like substances samples were not preserved properly (i.e. refrigerated) during the first two sampling days (day 145 and day 160) and subsequent contamination restricted the humic-like substances analysis to the last three sampling days. Therefore, the total specific bound EPS levels were higher for the aerobic membrane on day 145 and day 160 of the operation. The subsequent sampling days show that there was not much difference in the total bound EPS of the MBRs. For aerobic MBRs, the extent of fouling has been extensively documented to be related to the concentration of the soluble EPS or SMP in mixed liquor (Chang and Lee, 1998; Lee et al, 2001a; Wisniewski and Grasmick, 1998). The role of soluble EPS in the fouling of anaerobic membranes has not been widely reported in the literature and 66 RESULTS AND DISCUSSION therefore, a comparison of the constituents of soluble EPS may explain the role of soluble EPS better in anaerobic MBRs. In the present study, the soluble carbohydrate contents were slightly higher for the anaerobic mixed liquor (Figure 4.23). However, it can be suggested from Figure 4.24 and Figure 4.25 that the anaerobic mixed liquor contained significantly higher concentrations of soluble protein and humic-like substances than did the aerobic mixed liquor. Moreover, the total measured soluble EPS values were higher for the anaerobic mixed liquor as demonstrated in Figure 4.26. These results suggest that the soluble EPS plays a very important role in the fouling of anaerobic membranes. The high concentrations of soluble EPS in the first two samples coincided with the low temperature of the anaerobic reactor as shown in Figure 4.4. This is in agreement with some previous studies which reported higher soluble EPS production at lower temperatures in an anaerobic membrane bioreactor (Barker et al, 2000; Schiener et al, 1998). Similar results also have been reported recently by Rosenberger et al. (2006) for the treatment of municipal wastewater using an aerobic MBR. 100 *— Aerobic mixed liquor ••— Anaerobic mixed liquor 80 A 145 160 180 201 230 Operating days Figure 4.23. Soluble carbohydrates concentration in mixed liquor 67 RESULTS AND DISCUSSION 68 RESULTS AND DISCUSSION o -I 1 1 1 1 1 1 145 160 180 201 230 Operating days Figure 4.26. Total soluble EPS concentration in mixed liquor The total EPS concentration of the aerobic mixed liquor and the anaerobic mixed liquor is shown in Figure 4.27 and Figure 4.28 respectively. The total EPS concentration was almost same for both the mixed liquors, although the total EPS concentration was higher for the aerobic mixed liquor on operating day 160. Also, it is evident that the bound EPS contributes more in case of the aerobic mixed liquor when compared to the anaerobic mixed liquor. 69 RESULTS AND DISCUSSION 500 400 A — • - - Soluble EPS — • - Bound EPS A - Total EPS c o C3 § 200 -c o U 100 A 0 -I 1 1 1 . r 145 160 180 201 230 Operating day Figure 4.28. Total EPS concentration in the anaerobic mixed liquor 70 RESULTS AND DISCUSSION The specific bound EPS concentration study (Figure 4.22) did not clearly suggest that bound EPS was responsible for the higher fouling rate of the anaerobic membranes. On the other hand, the soluble EPS concentration study (Figure 4.26) suggested a role for soluble EPS in the higher fouling rate of the anaerobic membranes. However, as presented in Figure 4.13, the fluxes of the anaerobic system and the aerobic system were not similar during the on-line filtration studies. Therefore, it was assumed that the theoretical EPS flux, (i.e. the amount of EPS transported to the membranes per unit time) was of greater importance to understand the role of EPS in the fouling of the anaerobic membranes with respect to the aerobic membranes. The theoretical EPS flux towards the membrane surfaces was calculated by multiplying the EPS concentration with the membrane flux on the day of sampling. The theoretical total bound EPS flux and the theoretical total soluble EPS flux are shown in Figure 4.29 and Figure 4.30 respectively. The results in Figure 4.29 and Figure 4.30 do not count the effect of gas sparging on accumulation of SMP foulants on the membranes. X 3 q= p~ -a c 3 O 1e+5 8e+4 6e+4 Q o H g 4e+4 2e+4 Aerobic mixed liquor Anaerobic mixed liquor 145 160 180 201 230 Operating days Figure 4.29. Total bound EPS flux in membranes 71 RESULTS AND DISCUSSION X 3 0 0 ™ f i o p o E-50000 40000 30000 A 20000 H 10000 H Aerobic mixed liquor Anaerobic mixed liquor 145 160 180 201 230 Operating days Figure 4.30. Total soluble EPS flux in membranes The theoretical total bound EPS flux results in Figure 4.30 suggest that bound EPS flux to the membrane was higher in case of the aerobic membranes. Therefore, it can be suggested that bound EPS was not the cause behind the higher fouling of the anaerobic membranes. However, the theoretical flux of the soluble EPS was consistently higher in the anaerobic membranes as summarized in Figure 4.28. The higher concentrations of soluble EPS in the anaerobic mixed liquor as compared to the aerobic mixed liquor might be due to the difference in the solids retention time (SRT) of the two reactors. The SRT of the MEBPR process was 20 days, whereas there was no wasting of the mixed liquor (except for weekly sampling or accidental losses) in the anaerobic mixed liquor. The SRT of the AnMBR was approximately 3360 days. In addition, the role of SRT in high production of soluble EPS has been reported in some recent studies (Rosenberger et al, 2006; Jarusutthirak and Amy, 2006). Therefore, the higher SRT in the anaerobic MBR might be a contributor to the high production of soluble EPS and thus, a high rate of fouling of the membranes. 72 CONCLUSIONS CHAPTER FIVE CONCLUSIONS Based on the results obtained from the study of the submerged anaerobic MBR at the UBC Pilot Plant, the following conclusions can be drawn. 1. The anaerobic MBR process was effective in removing COD and VFA. Under acetate-supplemented condition, 80 % COD removal was achieved consistently in the system. Also, VFA removal was essentially complete in the AnMBR. 2. The permeate COD concentration averaged approximately 72 mg/L during the entire project period. This high concentration of permeate COD in the AnMBR suggests that aerobic post treatment is needed to achieve secondary quality effluent. Also, the estimated ratio of BOD5/COD was calculated to be 0.76 for the permeate which suggests that much of the refractory COD was potentially biodegradable under aerobic conditions.. 3. The on-line filtration studies suggest that organic foulants were the major contributor to the fouling of the aerobic MBR and the anaerobic MBR. 4. Scanning electron microcopy studies showed that some microbial colonies were attached to aerobic membranes. However, very little microbial colonization was found on anaerobic membranes. High shear stress caused by the air/gas sparging made it difficult for colonies to attach and grow on membrane surfaces. 5. According to mixed liquor characterization analysis, the total bound EPS concentration was higher in the aerobic membranes. On the other hand, total soluble EPS concentration was higher in the anaerobic membranes. 73 CONCLUSIONS 6. The high soluble EPS concentration in the anaerobic membranes might be the main reason for rapid trans-membrane pressure increase in the anaerobic MBR as compared to the aerobic MBR. 74 RECOMMMENDATIONS CHAPTER SIX RECOMMENDATIONS Based on the experience gained from this study on application of anaerobic MBR in municipal wastewater treatment, the following recommendations were made. 1. The treatment performance achieved in the submerged AnMBR did not materially exceed that of few similar reports available in the literature, and thus, it can be suggested that secondary quality effluent can not be achieved economically using AnMBR technology. More research can be conducted to see the effectiveness of anaerobic systems, such as moving bed biofilm reactor, to remove the bulk of the organic material from wastewater. The moving bed biofilm reactor can be coupled to an aerobic MBR to polish the effluent and produce secondary quality effluent. 2. The overall performance of an AnMBR will be determined by the microbial community. The microbiological conditions of an AnMBR can be different from that in a conventional anaerobic process depending on reactor design. Advanced phylogenetic analytical techniques, such as fluorescent in-situ hybridization and confocal scanning laser microscopy can be used to compare and evaluate the microbial community of an AnMBR and a conventional anaerobic process. 3. The mixed liquor suspended solids and colloidal solids have been reported to be contributors to the fouling of membranes and therefore, their role can be further investigated for anaerobic membrane bioreactors. 4. In this study, it was speculated that the higher SRT in the AnMBR was the reason for the higher concentration of soluble EPS in the anaerobic membrane system. The role of SRT on fouling of anaerobic membranes should be studied more 75 RECOMMMENDATIONS comprehensively to find out a balance between SRT and fouling of the membranes due to soluble EPS. 5. Low temperatures in anaerobic membrane bioreactors have been associated with high production of soluble EPS in the system. More research could be done to find out an ideal temperature range in which the production of soluble EPS can be minimized in the membrane bioreactors and thus, fouling can be reduced. 6. The surface characteristics of activated sludge floes, such as zeta potential and hydrophobicity could play an important role in membrane fouling. Their contribution towards fouling of anaerobic membranes should also be investigated. 7. Sludge floe size distribution also affects membrane permeability. Smaller sludge floes were reported in literature reviews to be associated with membrane fouling. Therefore, study could be done to verify their effect on membrane performance. 8. The concentration of mixed liquor suspended solids and sludge viscosity might have a role in membrane fouling. Their contribution should be examined by conducting comprehensive research work on anaerobic mixed liquor. 76 REFERENCES CHAPTER SEVEN REFERENCES Ahn, K., Cha, H. and Song, K. (1999). Retrofitting municipal sewage treatment plants using an innovative membrane-bioreactor system. Desalination, 124(1-3), 279-286. APHA, AWWA, WEF (1999). Standard Methods for the Examination of Water and Wastewater. (20). American Public Health Association (APHA), American Water Works Association (AWWA), Water Environment Federation (WEF), Washington, D. C. Barker, D. J., Salvi, S. M., Langenhoff, A. A. M. and Stuckey, D. C. (2000). Soluble microbial products in ABR treating low-strength wastewater. Journal of Environmental Engineering, 126(3), 239-249. Barker, D. J. and Stuckey, D. C. (1999). A review of soluble microbial products (SMP) in wastewater treatment systems. Water Research, 33(14), 3063-3082. Beaubien, A., Baty, M., Jeannot, F., Francoeur, E. and Manem, J. (1996). Design and operation of anaerobic membrane bioreactors: development of a filtration testing strategy. Journal of Membrane Science, 109(2), 173-184. Berube, P. R., Hall, E. R. and Sutton, P. M. (2006). Parameters governing the permeate flux in an anaerobic membrane bioreactor treating low-strength/municipal wastewaters: a literature review. In Press. Water Environment Research. Bourgeous, K. N., Darby, J. L. and Tchobanoglous, G. (2001). Ultrafiltration of wastewater: effects of particles, mode of operation, and backwash effectiveness. Water Research, 35(1), 77-90. Brockmann, M. and Seyfried, C. F. (1996). Sludge activity and cross-flow microfiltration a non-beneficial relationship. Water Science & Technology, 34(9), 205-213. Chang, I-.S. and Lee, C-.H. (1998). Membrane filtration characteristics in membrane-coupled activated sludge system - the effect of physiological states of activated sludge on membrane fouling. Desalination, 120(3), 221-233. Chang, I-. S., Clech, P. L., Jefferson, B. and Judd, S. (2002). Membrane fouling in membrane bioreactors for wastewater treatment. Journal of Environmental Engineering, 128(11), 1018-1029. 77 REFERENCES Chen, J., Donghong, D., Hall, E. R. and Berube, P. R. (2005). Membrane bioreactors for anaerobic treatment of wastewaters, Presented at the British Columbia Water and Waste Association Annual Conference, Whistler, Canada. Cho, B. D. and Fane, A. G. (2002). Fouling transients in nominally sub-critical flux operation of a membrane bioreactor. Journal of Membrane Science, 209(2), 391-403. Choo, K.-H., Kang, I.-J., Yoon, S.-H., Park, H., Kim, J.-H., Adiya, S. and Lee, C.-H. (2000). Approaches to membrane fouling control in anaerobic membrane bioreactors. Water Science & Technology, 41(10-11), 363-371. Choo, K.-H. and Lee, C.-H. (1996a). Membrane fouling mechanisms in the membrane-coupled anaerobic bioreactor. Water Research, 30(8), 1771-1780. Choo, K. -H. and Lee, C. -H. (1996b). Effect of anaerobic digestion broth composition on membrane permeability. Water Science & Technology, 34(9), 173-179. Choo, K.-H. and Lee, C.-H. (1998). Hydrodynamic behavior of anaerobic biosolids during crossflow filtration in the membrane anaerobic bioreactor. Water Research, 32(11), 3387-3397. Cui, Z. F., Chang, S. and Fane, A. G. (2003).The use of gas bubbling to enhance membrane processes. Journal of Membrane Science, 221(1-2), 1-35. Elmaleh, S. and Abdelmoumni, L. (1997). Cross-flow filtration of an anaerobic methanogenic suspension. Journal of Membrane Science, 131(1-2), 261-274. Elmaleh, S. and Abdelmoumni, L. (1998). Experimental test to evaluate performance of an anaerobic reactor provided with an external membrane unit. Water Science & Technology, 38(8-9), 385-392. Fane, A. G., Fell, C. J. D. and Suzuki, A. (1983). The effect of pH and ionic environment on the ultrafiltration of protein solutions with retentive membranes. Journal of Membrane Science, 16, 195-210. Flemming, H. -C , Szewzyk, U. and Griebe, T. (2000). Biofilms: investigative methods and applications. Technomic Publishing Company, Inc., Pennsylvania, U. S. A. Frolund, B., Palmgren, R., Keiding, K. and Nielsen, P. H. (1996). Extraction of extracellular polymers from activated sludge using a cation exchange resin. Water Research, 30(8), 1749-1758. Ghyoot, W. R. and Verstraete, W. H. (1997). Coupling membrane filtration to anaerobic primary sludge digestion. Environmental Technology, 18(6), 569-580. 78 REFERENCES Ghyoot, W., Vandaele, S. and Verstraete, W. (1999). Nitrogen removal from sludge reject water with a membrane-assisted bioreactor. Water Research, 33(1), 23-32. Hall, E. R. (1992). Anaerobic treatment of wastewaters in suspended growth and fixed film processes. Design of anaerobic processes for the treatment of industrial and municipal wastes, edited by J. F. Malina Jr. and F. G. Pohland. Technomic Publishing Co. Ltd., Lancaster, PA, p. 41. Harada, H., Momonoi, K., Yamazaki, S. and Takizawa, S. (1994). Application of anaerobic-UF membrane reactor for treatment of a wastewater containing high strength particulate organics. Water Science & Technology, 30(12), 307-319. Hernandez, A. E., Belalcazar, L. C , Rodriguez, M. S. and Giraldo, E. (2002). Retention of granular sludge at high hydraulic loading rates in an anaerobic membrane bioreactor with immersed filtration. Water Science & Technology, 45(10), 169-174. Hogetsu, A., Ishikawa, M., Yoshikawa, T., Tanabe, T., Yudate, S. and Sawada, J. (1992). High rate anaerobic digestion of wool scouring wastewater in a digester combined with membrane filter. Water Science & Technology, 25(7), 341-350. Hong, S. P., Bae, T. H., Tak, T. M., Hong, S. and Randall, A. (2002). Fouling control in activated sludge submerged hollow fiber membrane bioreactors. Desalination, 143(3), 219-228. -Hu, A. Y. and Stuckey, D. C. (2006). Treatment of dilute wastewaters using a novel submerged anaerobic membrane bioreactor. Journal of Environmental Engineering, 132(2), 190-198. Imasaka, T., Kanekuni, N., So, H. and Yoshino, S. (1989). Cross-flow filtration of methane fermentation broth by ceramic membranes. Journal of Fermentation and Bioengineering, 68, 200-206. Jarusutthirak, C. and Amy, G. (2006). Role of soluble microbial products (SMP) in membrane fouling and flux decline. Environmental Science & Technology, 40, 969-974. Kang, I., Yoon, S. and Lee, C. (2002). Comparison of the filtration characteristics of organic and inorganic membranes in a membrane-coupled anaerobic bioreactor. Water Research, 36(7), 1803-1813. Kayawake, E., Narukami, Y. and Yamagata, M. (1991). Anaerobic digestion by a ceramic membrane enclosed reactor. Journal of Fermentation and Bioengineering, 71(2), 122-125. 79 REFERENCES Kim, J., Lee, C. and Chang, I. (2001). Effect of pump shear on the performance of a crossflow membrane bioreactor. Water Research, 35(9), 2137-2144. Kiriyama, K., Tanaka, Y. and Mori, I. (1992). Field test of a composite methane gas production system incorporating a membrane module for municipal sewage. Water Science & Technology, 25(7), 135-141. Laspidou, C. S. and Rittmann, B. E. (2002). A unified theory for extracellular polymeric substances, soluble microbial products, and active and inert biomass. Water Research, 36(11), 2711-2720. Lee, S., Burt, A., Rusoti, G. and Buckland, B. M. (1995). Microfiltration of yeast cells using a rotary disk dynamic filtration system. Biotechnology & Bioengineering, 48, 386-400. Lee, J., Ann, W. and Lee, C. (2001a). Comparison of the filtration characteristics between attached and suspended growth microorganisms in submerged membrane bioreactor. Water Research, 35(10), 2435-2445. Lee, S., Jung, J. and Chung, Y. (2001b). Novel method for enhancing permeate flux of submerged membrane system in two-phase anaerobic reactor. Water Research, 35(2), 471-477. Liao, B.Q., Bagley, D.M., Kraemer, H.E., Leppard, G.G. and Liss, S.N. (2004). A review of biofouling and its control in membrane separation bioreactors. Water Environment Research, 76(5), 425-436. Liao, B.Q., Kraemer, J.T. and Bagley, D. M. (2006). Anaerobic membrane bioreactors: Applications and research directions. Critical Review Environmental Science & Technology, 36, 489-530. McMahon, K. D., Stroot, P. G., Mackie, R. I. and Raskin, L. (2001). Anaerobic codigestion of municipal solid waste and biosolids under various mixing conditions~II: microbial population dynamics. Water Research, 35(7), 1817-1827. Nagaoka, PL, Yamanishi, S. and Miya, A. (1998). Modeling of biofouling by extracellular polymers in a membrane separation activated sludge system. Water Science & Technology, 38(4-5), 497-504. Oh, S. E., Iyer, P., Bruns, M. A. and Logan, B. E. (2004). Biological hydrogen production using a membrane bioreactor. Biotechnology & Bioengineering, 87(1), 119-127. 80 REFERENCES Park, H., Choo, K. H., Lee, C. H. (1999). Flux enhancement with PAC addition in the membrane anaerobic bioreactor. Separation. Science & Technology, 34(14), 2781-2792. Pillay, V. L., Townsend, B. and Buckley, C. A. (1994). Improving the performance of anaerobic digesters at wastewater treatment works: The coupled crossflow microfiltration/digester process. Water Science & Technology, 30(12), 329-337. Rosenberger, S., Laabs, C , Lesjean, B., Gnirss, R., Amy, G., Jekel, M. and Schrotter, J. -C. (2006). Impact of colloidal and soluble organic material on membrane performance in membrane bioreactors for municipal wastewater treatment. Water Research, 40(4), 710-720. Ross, W. R., Barnard, J. P., Le Roux, J. and de Villiers, H. A. (1990). Application of ultrafiltration membranes for solids-liquid separation in anaerobic digestion systems: The ADUF process. Water SA, 16(2), 85-91. Sainbayar, A., Kim, J. S., Jung, W. J., Lee, Y. S. and Lee, C. H. (2001). Application of surface modified polypropylene membranes to an anaerobic membrane bioreactor. Environmental Technology, 22(9), 1035-1042. Schiener, P., Nachaiyasit, S. and Stuckey, D. C. (1998). Production of soluble microbial products (SMP) in an anaerobic baffled reactor: composition, biodegradability, and the effect of process parameters. Environmental Technology, 19(4), 391-399. Shimizu, U., Rokudai, M., Tohya, S., Kauawake, E., Yazawa, T., Tanaka, H. and Eguchi, K. (1989). Effect of pore size on the filtration characteristics of ceramic membrane for membrane bioreactors. Kagaku Kogaku Ronbunshu, 15, 322. Stephenson, T., Judd, S., Jefferson, B., and Brindle, K (2000). Membrane Bioreactors for Wastewater Treatment. IWA Publication, London, UK. Strohwald, N. K. H. and Ross, W. R. (1992). Application of the ADUF process to brewery effluent on a laboratory scale. Water Science & Technology, 25(10), 95-105. Stuckey, D. C. and Hu, A. (2003). The submerged anaerobic membrane bioreactor (SAMBR): An intensification of anaerobic wastewater treatment, Presented at the IWA Leading Edge Conference on Drinking Water and Wastewater Treatment Technologies, Noordwijk/Amsterdam, The Netherlands. Tchobanoglous, G., Burton, F. L. and Stensel, H. D. (2003). Wastewater Engineering Treatment and Reuse. Tata McGraw-Hill Company Inc., New York. 81 REFERENCES Wen, C , Huang, X. and Qian, Y. (1999). Domestic wastewater treatment using an anaerobic bioreactor coupled with membrane filtration. Process Biochemistry, 35(3-4), 335-340. Wiesner, M. R. and Aptel, P. (1996). Mass transport and permeate flux and fouling in pressure-driven processes. Water Treatment Membrane Processes, Wiesner, M. R. (ed). McGraw-Hill Companies, Inc. Wisniewski, C. and Grasmick, A. (1998). Floe size distribution in a membrane bioreactor and consequences for membrane fouling. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 138(2-3), 403-411. Yoon, S. H., Kang, I. J. and Lee, C. H. (1999). Fouling of inorganic membrane and flux enhancement in membrane-coupled anaerobic bioreactor. Separation Science Technology, 34(5), 709-724. Zoh, K. and Stenstrom, M. K. (2002). Application of a membrane bioreactor for treating explosives process wastewater. Water Research, 36(4), 1018-1024. 82 APPENDIX A APPENDIX A: UBC WASTEWATER TREATMENT PILOT PLANT 83 APPENDIX A APPENDIX B APPENDIX B: CLEAN AND FOULED US FILTER MEMBRANE MODULES A P P E N D I X B APPENDIX C APPENDIX C: BOD5 DATA OF THE ANMBR INFLUENT AND EFFLUENT 87 APPENDIX C Table C.l. BOD5 concentration of the influent and effluent Date (Operating day) BOD5 (mg/1) Influent BOD5 (mg/1) Effluent 10-Mar-06 (155) 179 59 13-Mar-06 (158) 162 60 15-Mar-06 (160) 135 47 20-Mar-06 (165) 155 56 88 APPENDIX D A P P E N D I X D : D A T A S U M M A R Y O F T R A N S - M E M B R A N E P R E S S U R E A N D F L U X IN A E R O B I C M E M B R A N E B I O R E A C T O R D U R I N G O N - L I N E F I L T R A T I O N S T U D I E S 89 APPENDIX D Table D.l. TMP and flux data of the aerobic MBR during on-line filtration studies Date (Operating day) TMP (kPa) Flux (LW hr) 17-Feb-06 (134) -4.00 15.00 18-Feb-06 (135) -5.00 13.80 19-Feb-06 (136) -6.00 16.80 20-Feb-06 (137) -6.00 16.20 21-Feb-06 (138) -6.00 15.60 22-Feb-06 (139) -7.00 15.30 23-Feb-06 (140) -7.99 18.00 24-Feb-06 (141) -8.99 16.50 25-Feb-06 (142) -7.99 16.50 26-Feb-06 (143) -8.99 16.50 27-Feb-06 (144) -8.99 16.50 28-Feb-06 (145) -8.99 16.50 1-Mar-06 (146) -9.49 16.20 2-Mar-06 (147) -8.99 16.20 3-Mar-06 (148) -8.49 16.08 4-Mar-06 (149) -8.49 16.50 5-Mar-06 (150) -8.99 15.96 6-Mar-06 (151) -9.49 16.20 7-Mar-06 (152) -9.49 15.90 8-Mar-06 (153) -9.49 16.08 9-Mar-06 (154) -9.49 16.14 10-Mar-06 (155) -9.49 16.20 11-Mar-06 (156) -9.99 15.90 12-Mar-06 (157) -9.99 15.84 13-Mar-06 (158) -10.49 15.78 14-Mar-06 (159) -10.49 15.66 15-Mar-06 (160) -10.99 15.60 16-Mar-06 (161) -10.99 15.60 17-Mar-06 (162) -12.49 15.60 18-Mar-06 (163) -12.49 15.12 19-Mar-06 (164) -12.49 15.12 20-Mar-06 (165) -12.99 15.90 21-Mar-06 (166) -12.99 15.60 22-Mar-06 (167) -12.99 15.48 23-Mar-06 (168) -13.49 15.48 24-Mar-06 (169) -13.49 15.36 25-Mar-06 (170) -13.99 15.30 26-Mar-06 (171) -13.99 15.72 27-Mar-06 (172) -14.99 15.72 90 APPENDIX D 28-Mar-06 (173) -13.99 14.88 29-Mar-06 (174) -16.99 16.14 30-Mar-06 (175) -16.99 16.14 31-Mar-06 (176) -17.99 16.08 1-Apr-06 (177) -17.99 16.02 2-Apr-06 (178) -19.99 16.02 3-Apr-06 (179) -19.99 15.96 4-Apr-06 (180) -20.99 15.90 5-Apr-06 (181) -21.98 15.90 6-Apr-06 (182) -21.98 15.96 7-Apr-06 (183) -21.98 15.84 8-Apr-06 (184) -22.98 15.84 9-Apr-06 (185) -22.98 15.78 10-Apr-06(186) -23.98 15.60 11-Apr-06 (187) -24.98 15.30 12-Apr-06 (188) -24.98 15.24 13-Apr-06 (189) -26.98 15.12 14-Apr-06 (190) -26.98 15.12 15-Apr-06 (191) -27.98 15.18 16-Apr-06 (192) -28.98 15.24 17-Apr-06 (193) -28.98 15.12 18-Apr-06 (194) -28.98 15.12 19-Apr-06 (195) -28.98 15.00 20-Apr-06 (196) -29.98 14.88 21-Apr-06 (197) -30.98 14.94 22-Apr-06 (198) -30.98 14.88 23-Apr-06 (199) -31.98 14.70 24-Apr-06 (200) -31.98 14.64 25-Apr-06 (201) -32.98 14.58 26-Apr-06 (202) -32.98 14.58 27-Apr-06 (203) -33.98 14.52 28-Apr-06 (204) -33.98 14.40 29-Apr-06 (205) -34.98 14.34 30-Apr-06 (206) -34.98 14.40 91 APPEND LX E APPENDIX E: DAILY MONITORING RECORDS SUBMERGED ANAEROBIC MEMBRANE BIOREACTOR PROCESS (In CD-ROM) 92 

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}]}"
                            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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0063221/manifest

Comment

Related Items