Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Aerobic granulation with low pH, low alkalinity municipal wastewater Leong, Jason 2014

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

Item Metadata

Download

Media
24-ubc_2014_november_leong_jason.pdf [ 2.24MB ]
Metadata
JSON: 24-1.0166061.json
JSON-LD: 24-1.0166061-ld.json
RDF/XML (Pretty): 24-1.0166061-rdf.xml
RDF/JSON: 24-1.0166061-rdf.json
Turtle: 24-1.0166061-turtle.txt
N-Triples: 24-1.0166061-rdf-ntriples.txt
Original Record: 24-1.0166061-source.json
Full Text
24-1.0166061-fulltext.txt
Citation
24-1.0166061.ris

Full Text

  Aerobic Granulation with Low pH, Low Alkalinity Municipal Wastewater  by Jason Leong B.A.Sc., The University of British Columbia, 2012  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIRMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2014 © Jason Leong, 2014    ii  Abstract Aerobic granulation was successfully achieved using municipal wastewater with supplemental carbon addition, after a 166 day start-up period. During the start-up period, an increase in biomass particle size was noticed that correlated to a stepwise reduction in settling and decant time. During this time, the granular type formation had a fluffy structure with noticeable filamentous outgrowths. On the 82nd day of start-up, supplemental carbon, in the form of sugar, was added to the feed tank to increase the Organic Loading Rate (OLR) of the system. In addition to an increase in COD, a pH and alkalinity decrease in the feed occurred due to acetogenesis. Soon after the supplemental carbon addition, the biomass developed a denser morphology, with gelatinous, white-coloured, granules appearing on the 166th day. Although it was shown that aerobic granulation with low pH, low alkalinity municipal wastewater with supplemental carbon was possible, some issues were discovered. These issues include reduced nutrient removal, and lower granule density. The reduction in nutrient removal performance is potentially due to the lack of nitrification/denitrification, and shifting of the microbial ecology to favour Glycogen Accumulating Organisms (GAOs) or heterotrophs over Polyphosphate Accumulating Organisms (PAOs). It is hypothesized that the reduced pH of the influent was the main factor that caused the reduction in nutrient removal. The lower granule density is hypothesized to be a result of the absence of denitrification induced chemical phosphorus precipitation within the granule. As such, the lower granule density results in reduced granule settling velocities and increased Sludge Volume Index (SVI).   iii  Preface       This dissertation is original, unpublished, independent work by the author, Jason Leong    iv  Table of Contents  Abstract ............................................................................................................................. ii Preface .............................................................................................................................. iii Table of Contents ............................................................................................................. iv List of Tables .................................................................................................................... vi List of Figures ................................................................................................................. vii Nomenclature .................................................................................................................... x Acknowledgements ........................................................................................................ xiv Dedication ........................................................................................................................ xv 1 Introduction .............................................................................................................. 1 1.1 The Nitrogen Cycle............................................................................................. 1 1.2 The Phosphorus Cycle ........................................................................................ 2 1.3 Eutrophication..................................................................................................... 3 2 Literature Review ..................................................................................................... 4 2.1 Nitrogen Removal ............................................................................................... 4 2.1.1 Physical Treatment ......................................................................................... 4 2.1.2 Chemical Treatment........................................................................................ 4 2.1.3 Biological Treatment ...................................................................................... 5 2.2 Phosphorus Removal ........................................................................................ 10 2.2.1 Chemical Treatment...................................................................................... 10 2.2.2 Biological Treatment .................................................................................... 12 2.2.2.1 Enhanced Biological Phosphorus Removal (EPBR) ............................ 12 2.2.2.2 Microorganism-Mediated Chemical Precipitation ............................... 16 2.3 Aerobic Granular Sludge (AGS) ...................................................................... 17 2.3.1 Granule Formation Mechanism .................................................................... 20 2.3.1.1 Quorum Sensing ................................................................................... 21 2.3.1.2 Extracellular Polymeric Substances (EPS) ........................................... 22 2.3.2 Parameters Influencing Granulation ............................................................. 23 2.3.2.1 Selective Pressure ................................................................................. 23 2.3.2.2 Hydrodynamic Shear Force .................................................................. 23 2.3.2.3 Growth Rate .......................................................................................... 25 2.3.2.4 Feast-Famine......................................................................................... 26 2.3.2.5 Organic Loading Rate (OLR) ............................................................... 27 2.3.3 Nutrient Removal in Aerobic Granular Sludge ............................................ 29 2.3.3.1 Nitrogen Removal ................................................................................. 29 2.3.3.2 Phosphorus Removal ............................................................................ 29   v  2.3.4 Performance Issues ....................................................................................... 31 2.3.4.1 Filamentous Bacteria ............................................................................ 31 2.3.4.2 PAO/GAO Competition ....................................................................... 32 2.3.5 Full Scale Aerobic Granulation .................................................................... 34 3 Thesis Objectives .................................................................................................... 36 4 Materials and Methods .......................................................................................... 37 4.1 Sequencing Batch Reactor (SBR) Setup .......................................................... 37 4.2 Sequencing Batch Reactor (SBR) Operation .................................................... 40 4.3 Apparatus Feed Tank ........................................................................................ 41 4.4 Reactor Maintenance ........................................................................................ 42 4.5 Analytical Methods ........................................................................................... 42 4.5.1 Performance Measurements.......................................................................... 42 4.5.2 Cycle Measurements ..................................................................................... 45 4.6 Quality Assurance/Quality Control .................................................................. 45 4.7 Statistical Analysis............................................................................................ 46 5 Results and Discussion ........................................................................................... 47 5.1 Reactor Start-up ................................................................................................ 47 5.1.1 Start-up Performance .................................................................................... 52 5.2 Reactor Performance with Varying Influent pH ............................................... 58 5.3 Reactor Cycle Profiles with Varying Influent pH ............................................ 65 5.4 Granule Characteristics and Morphology with Varying Influent pH ............... 70 6 Summary and Conclusions .................................................................................... 73 6.1 Recommendations for Future Work ................................................................. 75 References ........................................................................................................................ 77 Appendices ...................................................................................................................... 81 Appendix A: Additional Photographs .......................................................................... 82 Appendix B: MDL Calculations .................................................................................... 85     vi  List of Tables  Table 1. Cycle time distribution for start-up and operation of SBR ................................. 41 Table 2. Parameter measurement locations ...................................................................... 42 Table 3. Parameter analytical methods and preservation techniques ............................... 43 Table 4. Reactor cycle distribution during start-up phase ................................................ 47 Table 5. Feed characteristics during reactor start-up ........................................................ 48 Table 6. Phase 7 influent characteristics........................................................................... 58 Table 7. Phase 7 treated effluent characteristics ............................................................... 58      vii  List of Figures  Figure 1. Flow diagram of nitrogen cycle in aquatic systems (Water Environment Federation, 2010) ................................................................................................................ 2 Figure 2. Common biological nitrogen removal processes (a) MLE (b) Step Feed (c) 4-stage Bardenpho (Metcalf & Eddy et al., 2002) ................................................................. 8 Figure 3. Polyphosphate accumulating organisms under (a) anaerobic conditions and (b) aerobic conditions (Bitton, 2011) ..................................................................................... 13 Figure 4. Common BNR processes for biological phosphorus and nitrogen removal (a) 5-stage Bardenpho (b) Johannesburg (c) Modified UCT (Metcalf & Eddy et al., 2002) .... 14 Figure 5. Some examples of aerobic sludge granules (Tay et al., 2004a) (Adav et al., 2008) ................................................................................................................................. 17 Figure 6. Granulated sequencing batch reactor process in comparison to a standard activated sludge process (de Bruin et al., 2004) ............................................................... 19 Figure 7. Conceptual model for aerobic granulation (Verawaty et al., 2013) .................. 21 Figure 8. Effect of hydrodynamic shear forces induced by superficial air flow on granular SVI (●) and specific gravity (○) (Liu and Tay, 2002) ...................................................... 24 Figure 9. Influence of increased shear forces on EPS composition (PS = polysaccharides, PN = protein, ● = PS/PN, ○ = SOUR) (Liu and Tay, 2002) ............................................ 25 Figure 10. Effect of OLR on MLVSS (○) and F/M ratio ()in a synthetic fed aerobic granular system (Tay et al., 2004b) .................................................................................. 28 Figure 11. Mechanism for nitrification/denitrification in aerobic granules (Bassin et al., 2012) ................................................................................................................................. 29   viii  Figure 12. Progression of So/Xo (F/M) ratio during GSBR operation (Liu and Liu, 2006) .......................................................................................................................................... 32 Figure 13. Nereda® process cycles (Giesen et al., 2013) ................................................. 34 Figure 14. Schematic of sequencing batch reactor setup .................................................. 39 Figure 15. Sequencing batch reactor cycling .................................................................... 40 Figure 16. Progression of sludge morphology during start-up period .............................. 50 Figure 17. Particle size progression during the start-up period ........................................ 51 Figure 18. PO4 removal during start-up period ................................................................ 53 Figure 19. NH3 removal during start-up period ................................................................ 54 Figure 20. COD removal and OLR during start-up period ............................................... 55 Figure 21. Reactor and treated effluent VSS concentrations with SRT during start-up period ................................................................................................................................ 56 Figure 22. NO2 and NO3 concentrations in treated effluent during start-up period ......... 57 Figure 23. PO4 removal under varying influent pH .......................................................... 60 Figure 24. NH3 under varying influent pH ....................................................................... 61 Figure 25. COD removal and OLR under varying influent pH ........................................ 62 Figure 26. Reactor and treated effluent VSS concentration and SRT under varying influent pH ........................................................................................................................ 63 Figure 27. NO2 and NO3 concentration in treated effluent under varying influent pH .... 64 Figure 28. Normalized PO4 and NH3 cycle profiles under varying influent pH of 6.23 (n=8) and 5.62 (n=4) ......................................................................................................... 67 Figure 29. Normalized acetic and propionic acid concentrations at an influent pH of (a) 6.23 (n=4) and (b) 5.62 (n=4) ........................................................................................... 68   ix  Figure 30. Reactor pH cycle profiles under varying influent pH of 6.23 (n=6) and 5.62 (n=3) ................................................................................................................................. 69 Figure 31. Oxygen saturation profiles under varying influent pH of 6.23 (n=6) and 5.62 (n=5) ................................................................................................................................. 69 Figure 32. Particle size under different influent pH conditions........................................ 72 Figure 33. Granule appearance after approximately 30 days of operation with (a) 6.23 pH (phase 7-1) (b) 5.62 pH (phase 7-2) influent .................................................................... 72 Figure 34. Aerobic granular sequencing batch reactor setup with feed tank .................... 83 Figure 35. Examples of granules grown during study ...................................................... 84   x  Nomenclature °C Degrees Celsius AGS Aerobic Granular Sludge AHL N-acyle-homoserine-lactone Al Aluminum AlPO4 Aluminum Phosphate AOB Ammonia Oxidizing Bacteria ATP Adenosine Triphosphate BNR Biological Nutrient Removal C Carbon Ca Calcium Ca(OH)2 Calcium Hydroxide CaCO3 Calcium Carbonate CH3OH Methanol cm Centimetre COD Chemical Oxygen Demand (mg/L) d Day D10 10th Percentile of Particle Size Distribution (μm) D50 50th Percentile of Particle Size Distribution (μm) D90 90th Percentile of Particle Size Distribution (μm) DNA Deoxyribonucleic Acid DO Dissolved Oxygen (mg/L)   xi  EBPR Enhanced Biological Phosphorus Removal EPS Extracellular Polymers F/M Food to Microorganism Ratio (dimensionless) Fe Iron FePO4 Iron (III) Phosphate FNPT Female-National Pipe Thread GAO Glycogen Accumulating Organism GC Gas Chromatograph GSBR Granulated Sequencing Batch Reactor h Hour H2SO4 Sulfuric Acid H3PO4 Phosphoric Acid HDPE High Density Polyethylene HOCl Hypochlorus Acid HRT Hydraulic Residence Time (h) IWA International Water Association K Potassium kg Kilogram L Litre m Metre MDL Method Detection Limit Mg Magnesium mg Milligram   xii  mL Millilitre MLD Million Litres per Day MLE Modified Ludzack-Ettinger MLSS Mixed Liquor Suspended Solids MLVSS Mixed Liquor Volatile Suspended Solids N Nitrogen N2 Nitrogen Gas NH2Cl Chloramine NH3 Ammonia NH4+ Ammonium NO2- Nitrite NO3- Nitrate NOB Nitrite Oxidizing Bacteria O2 Oxygen Gas OLR Organic Loading Rate (kg/m3/day) P Phosphorus PAO Polyphosphate Accumulating Organism pH Power of Hydrogen PHB Poly-3-hydroxybutyrate PN Protein PP Polypropylene PS Polysaccharide PS Polystyrene   xiii  PSD Particle Size Distribution Q Flow rate (L/day) RAS Return Activated Sludge RNA Ribonucleic Acid RPM Revolutions per Minute s Second SBR Sequencing Batch Reactor sCOD Soluble Chemical Oxygen Demand (mg/L) SND Simultaneous Nitrification-Denitrification SRT Sludge Retention Time (days) SVI Sludge Volume Index (mL/g) TP Total Phosphorus (mg/L) TSS Total Suspended Solids (mg/L) UCT University of Cape Town VER Volume Exchange Ratio (%) VFA Volatile Fatty Acid Vr Reactor Volume (L) μm Micrometre     xiv  Acknowledgements   I would like to thank the following people for their assistance,  For their supervision throughout my thesis project: Dr. Don S. Mavinic, The University of British Columbia Dr. Babak Rezania, The University of British Columbia/ Prongineer R&D Dr. Dean Shiskowski, Associated Engineering For their assistance and companionship at the Annacis Wastewater Center: Sifat Kalam, The University of British Columbia Patrick Tsao, Prongineer R&D  Ron Howell, Metro Vancouver  For providing material, advice, and analytical assistance: Timothy Ma, The University of British Columbia Paula Parkinson, The University of British Columbia  For providing sludge from the UBC Pilot Plant: Fred Koch, The University of British Columbia Mike Harvard, The University of British Columbia Rony Das, The University of British Columbia Chaoyang Yue, The University of British Columbia  For raising me and putting up with my shenanigans: Mom and Dad  I would also like to thank Metro Vancouver for allowing me to work at the Annacis Wastewater Center, and Mantech Inc. for graciously donating analytical equipment used in this project   xv  Dedication       Dedicated to all those people who kept me out of jail          1  1 Introduction Concrete treatment objectives for municipal wastewater started to appear in the 1900s and concerned the removal of suspended solids, biodegradable organics, and pathogenic organisms. By the 1970s, an understanding of the relationship between the environmental impacts of eutrophication and wastewater nutrients, such as nitrogen and phosphorus, forced improvements in wastewater treatment objectives. Current best practices for municipal wastewater treatment incorporate methods to remove nitrogen and phosphorus. 1.1 The Nitrogen Cycle Nitrogen is one of two biologically important macronutrients utilized by biological systems. In aquatic systems, nitrogen can exist as ammonia (NH3), ammonium (NH4+), nitrate (NO3-), nitrite (NO2-), and nitrogen gas (N2). Ammonium, nitrate, and nitrite are highly soluble forms of nitrogen, while nitrogen gas and ammonia are insoluble gases that are readily released to the atmosphere. In the nitrogen cycle, as displayed in Figure 1, ammonium is produced from the decay of organic nitrogen. Ammonium is then converted into nitrites and nitrates through the nitrification process. These are then converted into nitrogen gas through denitrification, and subsequently released to the atmosphere. A potential shortcut of the nitrification/denitrification process is anammox which utilizes ammonium and nitrites to form nitrogen gas. Nitrogen release into aquatic systems can potentially have negative impacts due to eutrophication or increase in water toxicity (Water Environment Federation, 2010)(Galloway, 2003).    2   Figure 1. Flow diagram of nitrogen cycle in aquatic systems (Water Environment Federation, 2010) 1.2 The Phosphorus Cycle Phosphorus is a biologically important element that is utilized in biochemical reactions, DNA and RNA synthesis, biological respiration via ATP, and utilized as structural support in cellular membranes. Inorganic forms of phosphorus are the most common forms of phosphorus found on the planet. However, most inorganic forms of phosphorus are not available for direct assimilation for biochemical reactions and must be converted to orthophosphates to become biologically available. Terrestrial phosphorus resides in bedrock, soil, and biomass, with the vast majority residing in bedrock as the mineral apatite (Ca10(PO4)6(OH,F,Cl)2). During weathering of the bedrock, the phosphorus minerals are solubilized by exposure to naturally occurring acids, assimilated by plants, and released to the soil as decaying biomass. The terrestrial process of phosphorus migration has been modified by humans through the mining of phosphate rock for use as fertilizer, deforestation, and waste disposal. This has led to an increase in migration of   3  phosphorus from the terrestrial cycle into the aquatic cycle, thereby leading to increased eutrophication within the receiving body of water (Ruttenberg, 2003).  1.3  Eutrophication Eutrophication is caused by an increase in nutrients, typically nitrogen and phosphorus, in surface waters. The increased concentration of nutrients consequently results in an increase in growth of phytoplankton and aquatic plants, and thereby causes a decrease in water quality due to an increase in odors, reduction in visual aesthetics, increased potential for toxic phytoplankton blooms, and depletion of dissolved oxygen. Eutrophication can be mitigated by reducing the nutrient load to receiving bodies of water through controlling point and nonpoint sources. A major point source of nutrients is sewage discharge from urban areas. Advanced wastewater treatment techniques that utilize physical, chemical, or biological methods to remove nitrogen and phosphorus from wastewater are currently in use to control the nutrient load to receiving bodies. One major nonpoint source of nutrients is from agricultural runoff. Methods to reduce this include soil stabilization, reduction of overland water flow, use of slow release fertilizers, and reduction in excess fertilizer application (Prepas and Charette, 2003).    4  2 Literature Review Physical, chemical, and biological treatment options are available to accomplish nutrient removal of nitrogen and phosphorus from municipal wastewater. Outlined below are technical descriptions of physical, chemical, and biological removal methods and descriptions of best available treatment processes for nutrient removal.  2.1 Nitrogen Removal 2.1.1 Physical Treatment A physical treatment method for removal of volatile forms of nitrogen, such as ammonia, is to use gas stripping. Gas stripping occurs by entraining a contaminant-free gas, such as air, into the wastewater to induce an interfacial transfer of the volatile material into the gas phase. This unit operation most commonly occurs in aerated grit chambers, biological treatment reactors, or transfer channels. However, stripping towers, in which the gas phase and liquid phase are contacted in a packed bed, can also be utilized. The advantage of the stripping tower is that the packing material in the tower allows for an increased interfacial contact area between the gas and liquid, and thereby increases the mass transfer coefficient of the process. Disadvantages of this process are that it is only capable of removing volatile forms of nitrogen, requires extensive aeration capability, and may require treatment of the contaminated gas (Metcalf & Eddy et al., 2002). 2.1.2 Chemical Treatment One method to remove ammonia from wastewater is to chemically treat the wastewater through chlorination. At low doses, chlorine, in the form of hypochlorus acid (HOCl), readily reacts with ammonia to form chloramines (NH2Cl). At sufficiently higher doses   5  of chlorine, the chloramines are oxidized and released as N2 into the atmosphere. This is known as breakpoint chlorination. The overall reaction of breakpoint chlorination can be seen in Equation (1)  NH4+ + 1.5 HOCl → 0.5 N2 + 1.5 H2O + 2.5 H+ + 1.5 Cl- (1) Disadvantages of this process include increased chemical costs, toxicity of potential residual chlorine and chloramines to aquatic fauna, and potential decrease in pH of the wastewater following breakpoint chlorination (Metcalf & Eddy et al., 2002).  2.1.3 Biological Treatment Biological nitrogen removal in wastewater treatment plants commonly occurs through two processes, nitrification and denitrification. Nitrification is a process in which chemoautotrophs convert ammonia into nitrates in a two-step process. The first reaction occurs occur when ammonia-oxidizing bacteria (AOB), such as Nitrosomonas, Nitrosospira, Nitrosococcus, Nitrosoblobus, and Nitrosovibrio convert ammonia into nitrites, as seen in Equation (2).  NH3 + 1.5 O2 → NO2- + H+ + H2O (2) The second step occurs when nitrite-oxidizing bacteria (NOB), such as Nitrobacter, Nitrospina, Nitrospira, and Nitrococcus convert nitrite into nitrate, as seen in Equation (3).  NO2- + 0.5 O2 → NO3- (3) To ensure proper nitrification, multiple factors need to be monitored and controlled. Such parameters include oxygen level, temperature, and pH. The oxygen requirement to   6  oxidize 1 mg of ammonia is stoichiometrically calculated to be 4.6 mg O2. In practice, the dissolved oxygen level in the biological reactor is maintained above 2 mg/L. Controlled temperature is important in the process due to its influence on biological kinetics. The temperature ranges of nitrifiers are within 8-30°C, with an optimal temperature range of 25-30°C. Lastly, nitrification reactions will cease below a pH of 6. This is complicated by the fact that AOB produce hydronium as a by-product of the reaction, thereby decreasing the pH of the system. To maintain the optimum pH range of 7.5 to 8.5 for the bacteria, alkalinity must be maintained in the system. The amount of alkalinity, in the form of CaCO3, is stoichiometrically calculated to be 7.14 mg per 1 mg of nitrogen removed (Bitton, 2011)(Metcalf & Eddy et al., 2002).  After the nitrification stage, denitrification can be used to remove nitrates from the wastewater. Denitrification is an anaerobic biological process in which nitrates are reduced by in a multistep reaction into N2, as seen in Equation (4).   NO3- → NO2- → NO → N2O → N2 (4) Microorganisms that are capable of denitrification include Psuedomonas, Bacillus, Spirilum, Hyphomicrobium, Agrobacterium, Acinetobacter, Propionobacterium, Rhizobium, Corynebacterium, Cytophaga, Thiobacills, and Alcaligenes. For denitrification to occur, the denitrifying bacteria require an electron donor to complete the reduction process. In practice, the electron donors come in the form of organic compounds such as acetic acid, citric acid, methanol, ethanol, raw domestic wastewater, or endogenous respiratory by-products. An example of denitrification using methanol as a carbon source is illustrated in Equation (5).   7   6 NO3- + 5 CH3OH → 3 N2 + 5 CO2 + 7 H2O + 6OH- (5)  When using methanol, the ratio of CH3OH to NO3- should be greater than 3 to ensure complete denitrification. Another parameter that must be monitored during denitrification is pH. The pH range for denitrification is between 7 to 8.5, with the optimum pH around 7. Since alkalinity is produced during the denitrification process, the alkalinity can be used to support the nitrification process. The amount of alkalinity produced per mg of nitrate reduced is 3.6 mg, with a more conservative estimate being 3 mg. This is approximately half of the alkalinity consumed during nitrification. Lastly, the optimal temperature range for denitrification is between 35-50°C (Bitton, 2011).    There are a number of wastewater treatment plant processes that have been designed to biologically remove nitrogen from wastewater. Common biological nitrogen removal processes include the Modified Ludzack-Ettinger (MLE) process, the 4-stage Bardenpho process, and the Step Feed process, as illustrated in Figure 2. Typical design SRT for these systems range from 10 to 20 days at 10 °C to 4 to 7 days at 20°C (Metcalf & Eddy et al., 2002)   8   Figure 2. Common biological nitrogen removal processes (a) MLE (b) Step Feed (c) 4-stage Bardenpho (Metcalf & Eddy et al., 2002) The MLE process is commonly used to retrofit existing activated sludge facilities for biological nitrogen removal. In the MLE process, the reactor is configured to have an anoxic zone prior to the aerobic zone. The reasoning for this configuration is to allow the alkalinity produced by denitrification in the anoxic zone to be available for the nitrification step in the aerobic zone. Due to the internal recycle, the nitrate removal of the process is limited by the internal recycle rate. This can potentially lead to low nitrate removal efficiencies. An additional limitation of the MLE process is that it can generally only achieve a total nitrogen effluent concentration of 5 to 8 mg/L, which is comparatively high compared to other systems (Metcalf & Eddy et al., 2002). (a)  (b)  (c)     9  The Step Feed process is another process that is commonly used to retrofit existing activated sludge facilities to accommodate nitrogen removal. The Step Feed process is novel because it distributes influent between multiple anoxic/aerobic stages. Nitrogen removal in the system is highly dependent on the flow distributions between the stages. With proper configuration, total nitrogen effluent concentrations between 3 to 5 mg/L can theoretically be achieved. The low theoretical nitrogen effluent concentration is a result of the process’s ability for lower return activated sludge (RAS) dilution. This maintains a higher MLSS concentration at the head of the process, and therefore, greater treatment capacity. A disadvantage of the process is that it is a complex operation that usually requires a flow splitting control system for optimal performance (Metcalf & Eddy et al., 2002).  The 4-stage Bardenpho process is another commonly utilized biological nitrogen removal process. The 4-stage Bardenpho process is similar to an MLE process with the addition of an anoxic and aerobic stage. The additional anoxic and aerobic stage enables the 4-stage Bardenpho process to achieve less than 3 mg/L of total nitrogen. Also, the additional denitrification step after the initial nitrification zone allows for a lowered demand on the internal recycle to control effluent nitrates. After the initial anoxic and aerobic zones, the amount of readily biodegradable carbon is significantly reduced. This lack of readily biodegradable carbon reduces the denitification rate of the denitrifying bacteria by a factor of 3 to 8. To meet a low total nitrogen effluent concentration, an external carbon source, such as methanol, is required. This chemical requirement increases the operating costs of the process (Metcalf & Eddy et al., 2002).     10  2.2 Phosphorus Removal 2.2.1 Chemical Treatment Historically, phosphorus removal has been accomplished the using chemical precipitation method. Chemical precipitation occurs when multivalent metal ion salts are added to facilitate the precipitation of phosphates found in wastewater. The metal ions that are most commonly used to form phosphate precipitates are Ca2+, Al3+, and Fe3+.  Phosphate precipitation with Calcium is usually conduced with Ca(OH)2 (Calcium hydroxide) to form Ca10((PO4)6)OH2 (Hydroxylapatite) at a pH greater than 10, as seen in Equation (6).  10 Ca+ + 6 PO43- + 2 OH- ↔ Ca10(PO4)6(OH)2 (6) In most cases for this reaction, the amount of available alkalinity in the wastewater serves as the limiting factor in the precipitation of phosphorus. Due to the pH requirement of 10 and the reaction’s consumption of alkalinity, the amount of lime required for the precipitation reaction is generally 1.4 to 1.5 times the total alkalinity in the wastewater. Some disadvantages of this process include increased sludge production, pH adjustment of wastewater prior to discharge, and increased operation costs due to chemical addition (Metcalf & Eddy et al., 2002).  Phosphate precipitation using Aluminum or Iron can be accomplished with a number of salt compounds such as ferric chloride, alum, ferric sulfate, or ferrous sulfate. The reactions of these salts with phosphates can be seen in Equation (7) and (8)   11   Fe3+ + HnPO43-n ↔ FePO4 + nH+ (7)  Al3+ + HnPO43-n ↔ AlPO4 + nH+ (8) For the precipitation reaction to properly occur, pH must be maintained above the minimum solubility levels of 6.3 for AlPO4 and 5.3 for FePO4. In practical applications, it has been found that good phosphorus removal can occur within a pH range of 6.5 to 7 and chemical dosages within a 1 to 3 metal ion to phosphorus molar ratio. Disadvantages of the process include the possibility of low pH effluent, increased cost of chemical addition, and increased sludge production (Metcalf & Eddy et al., 2002).    12  2.2.2 Biological Treatment 2.2.2.1 Enhanced Biological Phosphorus Removal (EPBR) Biological phosphorus removal utilizes Polyphosphate Accumulating Organisms (PAOs) to accumulate phosphorus in excess of their metabolic requirements. In order to force the PAOs to uptake excess phosphorus, the bacteria is subjected to alternating aerobic and anaerobic conditions. The bacterium suspected of being responsible for phosphorus removal during biological treatment is Candidatus Accumulibacter phosphatis (Bitton, 2011).  During the anaerobic stage, the PAOs utilize energy from the metabolism of poly-3-hydroxybutyrate (PHBs) carbon reserves, to provide the energy for inorganic phosphorus uptake and conversion to polyphosphates. The polyphosphates are stored intracellularly as polyphosphate granules. Under anaerobic conditions, volatile fatty acids (VFAs) are taken up by the PAOs and stored as PHBs. The energy required to convert the VFAs to PHBs is supplied by the conversion of polyphosphate to ATP by polyphosphate hydrolysis and phosphorylation. The conversion of polyphosphate results in the release of inorganic phosphorus and protons as a by-product. It should be noted that the presence of electron acceptors, such as O2, and NO3-, will prevent the assimilation of VFAs and subsequent release of inorganic phosphorus (Bitton, 2011). An illustration of the mechanism of phosphorus uptake by PAOs is demonstrated in Figure 3.   13   Figure 3. Polyphosphate accumulating organisms under (a) anaerobic conditions and (b) aerobic conditions (Bitton, 2011) To maintain a successful EPBR process, operational conditions must be controlled to ensure favourable conditions for PAO growth. Operational conditions that favour the growth of PAOs include a DO concentration greater than 1 mg/L, pH greater than 6.5, SRT greater than 2.5 days, and sufficient concentrations of Mg, K, and Ca cations. The recommended molar cation to phosphorus ratio for EBPR are 0.71 Mg/P, 0.50 K/P, and 0.25 Ca/P (Metcalf & Eddy et al., 2002). A number of processes have been developed that utilize the biological phosphorus removal mechanisms as mentioned above. Many processes currently in use to remove phosphorus have the added benefit of nitrogen removal through nitrification and denitrification. Some commonly used biological nutrient removal (BNR) processes include the 5-stage Bardenpho process, the Modified UCT process, and the Johannesburg process, as illustrated in Figure 4 (Metcalf & Eddy et al., 2002).    14   Figure 4. Common BNR processes for biological phosphorus and nitrogen removal (a) 5-stage Bardenpho (b) Johannesburg (c) Modified UCT (Metcalf & Eddy et al., 2002) The 5-stage Bardenpho process utilizes alternating anaerobic, anoxic, and aerobic zones for combined phosphorus, nitrogen, and carbon removal. The initial anaerobic stage stresses PAOs to uptake VFAs and release inorganic phosphorus. After the anaerobic stage, an anoxic stage is utilized for denitrification. An internal recycle from the downstream aerobic zone is utilized to supply nitrates to the anoxic zone. The subsequent aerobic zone serves to provide nitrification, carbon removal, and force phosphorus (a)  (b)  (c)     15  uptake by the PAOs. The last anoxic and aerobic zone serves a final denitrification step and gas stripping stage for residual N2. The 5-stage Bardenpho is typically capable of producing an effluent with 3 to 5 mg/L of total nitrogen. However, it is not typically capable of efficient phosphorus removal due partially to the introduction of nitrates from the RAS return into the anaerobic zone (Metcalf & Eddy et al., 2002). The modified UCT process utilizes 4 zones in the process; an anaerobic zone, 2 anoxic zones, and finally an aerobic zone. Like the 5-stage Bardenpho, the process begins with an anaerobic zone that forces PAOs to uptake VFAs and release inorganic phosphorus. After treatment in the anaerobic zone, effluent passes into the first anoxic zone where denitrification occurs. This zone primarily serves to denitrify the return activated sludge of any nitrates prior to being recycled into the anaerobic zone. The return activated sludge must be denitrified to ensure that nitrates do not interfere with the VFA uptake process in the anaerobic zone. The second anoxic stage provides the majority of the denitrification in the process by treating the nitrate recycle from the final aerobic zone. In the last zone, phosphorus uptake, nitrification, and the removal of the remaining carbon occurs prior to discharge. Due to the process improvements, the modified UCT is able to achieve better phosphorus removal due to less nitrate interference (Metcalf & Eddy et al., 2002). The Johannesburg process is another process that tries to minimize nitrates fed into the anaerobic zone. The Johannesburg process is a 4 zone process that alternates between anoxic, anaerobic, anoxic, and aerobic zones. The first stage in the process is an anoxic zone where return activated sludge is denitrified prior to being fed into the anaerobic zone. In the subsequent anaerobic zone, PAOs uptake VFAs, and influent is fed into the   16  process. The second anoxic stage receives the effluent from the anaerobic zone in addition to an internal recycle from the aerobic zone. This is where the majority of denitrification occurs in the Johannesburg process. The final aerobic zone is used for nitrification, phosphorus uptake, and carbon removal. Compared to the modified UCT process, the Johannesburg process is able to maintain higher MLSS concentrations inside the anaerobic reactor (Metcalf & Eddy et al., 2002).  2.2.2.2  Microorganism-Mediated Chemical Precipitation  Phosphate solubility in aqueous solution is controlled by pH and the presence of metal ions such as Ca2+, Mg2+, Fe3+, and Al3+. When precipitation occurs, formation of insoluble phosphate compound such as hydroxyapatite (Ca10(PO4)6(OH)2), vivianite (Fe3(PO4)2∙8H2O), and variscite (AlPO4∙2H2O) occurs (Bitton, 2011). In systems undergoing denitrification, the increase in pH in flocs and biofilms, in conjunction with a suitable concentration of suitable metal ions, can induce phosphate precipitation around the bacteria in the biofilms or flocs (Arvin and Kristensen, 1982).    17  2.3 Aerobic Granular Sludge (AGS) Aerobic granules were first cultivated in 1997 in a laboratory setting by a joint research project between the Technical University of Munich and the Delft University of Technology (Winkler, 2012). As the technology matured, aerobic granules were defined by the International Water Association (IWA) as “aggregates of microbial origin, which do not coagulate under reduced hydrodynamic shear, and which settle significantly faster than activated sludge flocs” (de Kreuk et al., 2005b). For granular sludge to settle faster than activated sludge flocs, their diamater must generally be greater than 0.2 mm (de Kreuk et al., 2007). Some examples of sludge granules can be seen in Figure 5.   Figure 5. Some examples of aerobic sludge granules (Tay et al., 2004a) (Adav et al., 2008)   18  Current operations utilizing Aerobic Granular Sludge (AGS) are only able to be accomplished using sequencing batch reactors (SBRs). This process is commonly referred to as granulated sequencing batch reactors (GSBRs). The operation schedule for the GSBRs are split into an anaerobic feeding period, an aeration phase, a sedimentation phase, and a decantation phase (de Bruin et al., 2004). During initial full scale operations of the GSBR systems, a number of advantages were discovered. One advantage that was discovered was that granular sludge could simultaneously remove nitrogen, phosphorus, and COD while producing minimal waste sludge. This translates into a 75% smaller footprint, when compared to conventional activated sludge processes (de Bruin et al., 2004). A process flow diagram comparison of GSBR and conventional activated sludge BNR processes can be seen in Figure 6.   19   Figure 6. Granulated sequencing batch reactor process in comparison to a standard activated sludge process (de Bruin et al., 2004)     20  2.3.1 Granule Formation Mechanism The mechanism for granule formation is a multi–step process, in which bacterium initiates physical contact, adhere, develop permanent attachment structures, and then mature into a three-dimensional, aggregate structure. In the first step, physical contact of bacterium can be accomplished through hydrodynamic forces, diffusion, gravity, Brownian motion, or cell mobility. Once the bacterium come into contact, initial physical, chemical, or biochemical attractive forces allow the bacterium to maintain contact. Subsequently, the bacterium develops permanent attachment structures through production of extra-cellular polymers (EPS), modification of cell metabolism, or alteration of gene expression. Granule size increases through microbial outgrowths and aggregation with other granules. The immature granule develops into a spherical pellet through the influence of hydrodynamic shear forces and collisions. Fragmentation of large granules is seen as substrate limitations to the interior of the granule cause degradation of the granule core  (Beun et al., 1999; Liu and Tay, 2002; Verawaty et al., 2013). Substrate limitations are usually a result of granules reaching a critical size of 600 μm to 800 μm, which limits the diffusion of substrate into the interior of the granule (Verawaty et al., 2013). An illustration of the granule maturation process is described in Figure 7. The resultant granule structure appears to be a two phase structure, where the outer layer consists mainly of polysaccharides and bacterium, while the inner layer is mainly a non-cellular protein core (McSwain et al., 2005)(Chen et al., 2007).   21   Figure 7. Conceptual model for aerobic granulation (Verawaty et al., 2013) Much research has focused on determining the optimal conditions to form and maintain viable granular sludge cultures. The factors that have been determined to be important to the formation and stability of AGS include selective pressure through settling velocity, hydrodynamic shear forces, organism growth rate, and formation of EPS (de Kreuk et al., 2007) 2.3.1.1 Quorum Sensing Quorum sensing is a bacterial, cell-to-cell communication process in which bacteria secrete autoinducer signaling molecules. The concentration of autoinducer signaling molecules are a function of cell density, and at threshold concentrations, the autoinducer molecules causes an alteration in gene expression that can regulate the physiological activities of a grouping of cells (Miller and Bassler, 2001). In a study of a granular system, autoinducer molecule N-acyl-homoserine-lactone (AHL) concentrations increased during biomass transition from floccular to granular. In addition, production of AHLs correlated with an increase in PS/PN ratio of the EPS (Tan et al., 2014). It is postulated that quorum sensing induces bacteria to shift to an attached growth state that   22  allows for granulation to occur (Ren et al., 2010). Increased selective pressure from decreasing settling times were also found to correlate to an increase in autoinducer molecules, potentially indicating the importance of selective pressure to induce cell-to-cell communication and EPS regulation (Zhou et al., 2013).  2.3.1.2 Extracellular Polymeric Substances (EPS) EPS are polymeric substances excreted by microbial organisms that play a crucial role in the formation of flocs and granule. EPS is generally composed of carbohydrates (polysaccharides), amino sugars, uronic acids, proteins, lipids, and nucleic acids (Bitton, 2011). EPS enables the formation of granules by providing a matrix for the immobilization of cells and maintaining structural integrity of the granules (Liu and Tay, 2002). EPS can also neutralize the negative surface charge of dispersed bacteria, thereby facilitating bacteriaal cohesion through a bridging mechanism (Tay et al., 2001). EPS produced in granules have been found to be rheologically different from EPS produced in flocs. Granular EPS has been described as having gel-forming characteristics, while floc EPS being described as more paste-like. As such, the gel-like EPS in granular systems are able to better function as a structural support compared to paste-like EPS produced by floccular systems (Seviour et al., 2009).      23  2.3.2 Parameters Influencing Granulation 2.3.2.1 Selective Pressure In GSBR processes, settling time is utilized as a control variable in order to cultivate a healthier granular culture. This is easily controlled in the SBR process by altering the amount of time allotted for the sedimentation phase in the operational cycle. By specifying a particular amount of time for sedimentation, granules of a particular size will have enough time to settle, while flocs and smaller granules will remain in suspension, and subsequently washed out during the decantation phase. This, in turn, forces the remaining microbes to modify their metabolisms to increase their surface hydrophobicity. In order to maintain this selective pressure, a maximum settling time of 5 minutes is required for successful aerobic granulation (Qin et al., 2004).  2.3.2.2 Hydrodynamic Shear Force Typically in GSBR processes, the aeration system is used to impart hydrodynamic shear force onto the sludge granules, by maintaining a high superficial air velocity. One benefit of increased hydrodynamic shear forces is a decrease in Sludge Volume Index (SVI) and an increase in specific gravity of the aerobic granules, as seen in Figure 8. The decrease in SVI and increase in specific gravity is due to the development of more compact and denser granules, which consequently has the effect of reducing the settling time of the granules (Liu and Tay, 2002).     24   Figure 8. Effect of hydrodynamic shear forces induced by superficial air flow on granular SVI (●) and specific gravity (○) (Liu and Tay, 2002) Hydrodynamic shear forces also have an important influence on the EPS matrix of the granules. It has been discovered that an increased hydrodynamic shear force influences the biological community within the granule structures to excrete more EPS, which consequently increases the structural stability of the granules. Increased shear forces also have a marked influence on the composition of the EPS produced during granulation (Tay et al., 2001). The shear forces cause the bacterial cultures to produce EPS that has higher polysaccharides content, as seen in Figure 9.    25   Figure 9. Influence of increased shear forces on EPS composition (PS = polysaccharides, PN = protein, ● = PS/PN, ○ = SOUR) (Liu and Tay, 2002) The presence of cellular polysaccharides function by promoting cell interaction by bridging cells to form three-dimensional structures and promoting surface charge neutralization on the surface of bacteria (Tay et al., 2001). 2.3.2.3 Growth Rate Slow growing organisms were found to increase the stability of granules during extended operations, even at low oxygen concentrations. When slow growing organisms are selected, the outward growth rates of the granules are balanced with detachment forces, thereby achieving a smoother spherical granule surface (de Kreuk and van Loosdrecht, 2004). Another effect of slower growth rates is an increase in density of the granule (de Kreuk and van Loosdrecht, 2004). The combined effect of greater sphericity and density leads to faster settling times of the granules, which can aid in increasing selective pressure on the granules. Additionally, slower growth rates decreases substrate gradients within the granule, due to the lower outward growth velocity (de Kreuk and van Loosdrecht, 2004). In order to select for slow growing organisms, an anaerobic   26  feast/aerobic famine regime is instituted in order to force selection of organisms that convert VFAs, such as acetate, in slowly degradable storage polymers (PHA/PHB). As such, this leads to a dominance of PAOs and GAOs in the granular sludge (de Kreuk and van Loosdrecht, 2004).  2.3.2.4 Feast-Famine Feast-famine is a period of cell starvation occurring typically during the end of the aerobic portion of the reactor cycle. The effects of periodic starvation of aerobic granules during a cycle have been controversial. Conflicting reports have been issued on the cell starvation’s effects on EPS degradation and cell hydrophobicity (Wang et al., 2006).  A study with varying cycle times, to simulate different starvations times, showed that short starvation periods decreased time to granulation; however, it resulted in long term instability of the granules (Liu and Tay, 2008). One issue with this study was that by varying the cycle times to simulate different starvation periods, the OLR and selection pressure experienced by each reactor was not kept consistent. This could explain the difference in time to granulation and long-term, granule stability. In another study looking at the effects of C, N, P, and K starvation, showed a reduction in microbial activity, granule structural integrity, EPS content, and settleability (Wang et al., 2006). Another study was conducted where no feast-famine regime occurring during the cyclic operation of the reactor. It was shown that granule formation still occurred despite the lack of a famine regime (Liu et al., 2007). These findings provide additional questions regarding the notion that feast-famine regimes are required for granule formation and stability.    27  2.3.2.5 Organic Loading Rate (OLR) Organic loading rate (OLR) is an indicator of the amount of substrate supplied to the reactor. OLR is calculated using Equation (9)                           (9) where OLR is the organic loading rate in kg/m3/d, CODfeed is the COD concentration of the influent feed in mg/L, Q is the total daily flowrate to the reactor in L/d, and Vr is the volume of the reactor in L. In a study conducted by Tay et al. (2004b) granule formation, and morphology was demonstarted to be greatly influenced by OLR. Tay observed that at an OLR below 2 kg/m3/d, granule formation could not be achieved. At moderate OLR, granule formation was achieved with the most stable appearing at 4 kg/m3/d. At a high OLR, granule formation was achieved; however, granule destabilization and biomass washout occurred shortly there after. Figure 10 demonstrates this by showing the highest MLVSS concentration within the reactor occured a moderate OLR of about 4 kg/m3/d, while at lower and higher OLR, the MLVSS in the reactor remained low. This is due to to a fast settling granule biomass forming at a moderate OLR, thereby preventing biomass washout. Note that the study was conducted with synthetic feed with an acetate carbon source; optimum OLR for a system will likely vary depending on the carbon source (Pijuan, 2013)    28   Figure 10. Effect of OLR on MLVSS (○) and F/M ratio ()in a synthetic fed aerobic granular system (Tay et al., 2004b) In addition, performance was also found to be most efficient at a moderate OLR with an average sCOD removal of 99% (Tay et al., 2004b). Granules grown at an OLR of 4 kg COD/m3/d were found to have a high surface hydrophobicity, and EPS with a higher PS/PN ratio (Tay et al., 2005).     29  2.3.3 Nutrient Removal in Aerobic Granular Sludge  2.3.3.1 Nitrogen Removal AGS is capable of simultaneous nitrification and denitrification (SND) due to the presence of aerobic and anoxic zones within the granule. The aerobic zone is located in the surface layer of the granule and is maintained through aeration of the bulk liquid. In this aerobic zone, ammonia is adsorbed onto the surface of the granule, and is converted by AOB into nitrates, and nitrites by NOB. The anoxic zone within the granule is a result of the large size of granules limiting the mass transfer of oxygen into the interior of the granule, where denitrification can occur. Denitrification can occur through the traditional path using denitrifying bacteria. However, if a lack of VFA occurs within the interior of the granule, PHB storing bacteria such as denitrifying PAOs and GAOs, can be utilized for denitrification, as demonstrated in Figure 11 (Bassin et al., 2012).    Figure 11. Mechanism for nitrification/denitrification in aerobic granules (Bassin et al., 2012) 2.3.3.2 Phosphorus Removal Phosphorus removal with AGS is accomplished via two mechanisms: EBPR, and biologically mediated chemical precipitation. EBPR was accomplished by utilizing alternating anaerobic and aerobic feeding periods to facilitate the growth of PAO cultures   30  within the interior of the granule (de Kreuk et al., 2005a). In addition to EBPR, phosphorus is can be removed as inorganic phosphate precipitate through biologically mediated chemical precipitation. This is possible only if pH, and metal ion concentrations are suitable for formation of precipitate. The internal pH of the granule can be potentially higher than the bulk fluid if denitrification occurs in the system. This would result in a pH increase within the granule, and thereby, create favourable conditions for phosphorus precipitation. Due to biologically mediated chemical precipitation of phosphorus, the majority of the inorganic phosphorus found in granules is found in the interior of the granule, as hydroxyapatite (Angela et al., 2011).     31  2.3.4 Performance Issues 2.3.4.1 Filamentous Bacteria During operations, overgrowth of filamentous bacteria leads to issues with long term stability of the aerobic granular sludge process. As overgrowth of filamentous bacteria occurs within the reactor, a decrease in the settleability of aerobic granules occurs, which results in a washout of aerobic granules. Operational conditions that are thought to favour filamentous bacterial growth include availability of readily biodegradable organics, low substrate availability, low dissolved oxygen concentrations, long solids retention time, nutrient deficiency, and high wastewater temperature (Liu and Liu, 2006).  For GSBRs operating at steady state conditions, a long Solid Retention Time (SRT) of approximately 25 days is common. This long SRT in GSBR systems tend to allow the growth of slow growing filamentous bacteria (Liu and Liu, 2006). Controlling SRT within a GSBR is difficult because it requires controlling settling time within the reactor. Since settling time is dependent on a variety of factors, settling time of the granules is not constant. Thus, it becomes difficult to maintain a constant SRT within a GSBR system with a fixed settling time.  Another issue for GSBRs operating at steady state conditions is that the high biomass concentration leads to low F/M ratios within the reactor, as seen in Figure 12. Low F/M conditions favour the growth of filamentous bacteria over floc forming bacteria, since filamentous bacteria are able to survive more efficiently at low substrate concentrations.    32   Figure 12. Progression of So/Xo (F/M) ratio during GSBR operation (Liu and Liu, 2006) Low substrate concentrations within the bulk liquid also present issues with substrate diffusion into the interior of the granule. Under substrate diffusion limited conditions, granules develop a porous and irregular shape. These factors can assist in the promotion of filamentous growth (Liu and Liu, 2006). 2.3.4.2 PAO/GAO Competition An important parameter to increase the performance of the granular system is to shift the microbial ecology of the reactor to favour PAOs. Granules with a high PAO fraction form denser granules with higher settling velocities than granules with higher GAO fractions (Winkler et al., 2013). The difference in settling velocities from PAOs and GAOs allows for a segmented granular sludge bed to form, thereby allowing selective wasting of GAO dominant granules (Winkler et al., 2011). By shifting the microbial ecology to PAO dominant granules, an increase in phosphorus removal, decrease in settling time, and an increase in granule density, can be witnessed.    33  One parameter that affects the competition between PAOs and GAOs is temperature. Temperatures greater than 20°C tend to favour GAOs over PAOs. At high temperatures, selective sludge wasting of GAO granules can be utilized to favour PAO granule formation (Bassin et al., 2012). As such, lower temperatures are seen to enhance EBPR processes (Bitton, 2011). pH is also found to affect microbial ecology of EBPR systems. At high pH values of 7.5 to 8 , PAOs uptake VFAs and phosphorus more readily, and therefore have an advantage over GAOs (Oehmen et al., 2005). Optimal pH range for PAOs is between 7 to 8, with GAO predominating at pH lower than 7 (Bitton, 2011)    34  2.3.5  Full Scale Aerobic Granulation Royal HaskoningDHV has been the first company develop a full scale aerobic granular system in 2005, through their proprietary process called Nereda®. The Nereda® process, as seen in Figure 13, occurs in 3 steps rather than the conventional 4 step process of sequencing batch reactors. The proprietary process eliminates the decant phase by simultaneous drawing effluent during the feeding phase.  Figure 13. Nereda® process cycles (Giesen et al., 2013) Currently there are 10 full-scale Nereda® reactors operating internationally, ranging between 0.25 to 5 MLD in design flows (Giesen et al., 2013). One such example is at the Gansbaai wastewater treatment plant in South Africa. There the process was designed for a peak flow of 5000 m3/d. The influent concentrations experienced at the Gansbaai treatment plant in 2011 were 1265 mg/L COD, 75 mg/L NH4-N, 19 mg/L TP, and 450 mg/L of suspended solids. Effluent concentrations were lower than the design   35  requirements, leading to 97% COD  removal, >98% NH4-N removal, 82% TP removal, and 99% suspended solids removal (Giesen et al., 2013). At another treatment plant in Frielas, Portugal, a comparative analysis of a granular and activated sludge system demonstrated that the granular system was capable of 30% reduction in operational aeration costs for the removal of an equivalent amount of COD (Giesen et al., 2013).     36  3 Thesis Objectives BNR is increasingly becoming a desirable treatment option in many municipalities faced with meeting strict effluent guidelines. Current best practices for BNR can be unattractive due to their requirements for large capital investment and large operational footprints. Aerobic granular systems provide an alternative solution to conventional treatment systems while having lower cost and lower operational footprint; at the same time high removal efficiencies are maintained. These benefits make aerobic granular sludge technology an attractive alternative for municipalities.   Much of the established literature has demonstrated the cultivation of aerobic granular sludge, using synthetic feeds with neutral to alkaline conditions that favour the growth of PAOs, nitrifiers, and denitrifiers. These bacteria have been shown to promote the growth of dense granules, which lends itself to a more stable granular system in a sequencing batch reactor. In contrast, wastewater generated in the Metro Vancouver area is typically low in COD, low in alkalinity, and slightly acidic. These could potentially present issues with the formation of dense, stable granulation.         37  4 Materials and Methods 4.1 Sequencing Batch Reactor (SBR) Setup The sequencing batch reactor utilized in this study was a cylindrical Plexiglas reactor that was 38.1 cm in height and 11.43 cm in diameter (see Figure 14). On the left side of the reactor, five 3/8” (0.95 cm) FNPT threaded sample ports were located from the bottom of the reactor at 6.99 cm (port 1), 14.61 cm (port 2), 16.51 cm (port 3), 26.04 cm (port 4), and 35.56 cm (port 5) respectively. Ports 1, 3, and 4 were fitted with 3/8” (0.95 cm) ball valves to facilitate sampling and draining of the reactor. Port 2 was utilized for decanting of the reactor and either fitted with a 1/2” (1.27 cm) solenoid valve or connected to a peristaltic pump. Port 5 was not utilized and fitted with a threaded plug. On the right side of the reactor (port 6), a 3/8” (0.95 cm) FNPT threaded port was located 6.35 cm from the bottom of the reactor. This port was fitted with a 1/2” (1.27 cm) solenoid valve, and connected to a submersible centrifugal pump (Little Brother 2E-38NT) to facilitate feeding of the reactor. The centrifugal pump was configured to operate in line with the feed tank rather than submersed. Feeding for the reactor was characterized as fast feed as the flow rate into the reactor was left uncontrolled.  The reactor was operated 24 hours per day and automatically cycled between the un-aerated period, aerated period, settling period, and decant period using a ChronTrol CD-03 programmable timer. During the un-aerated feeding period, the reactor was fed to a liquid level of 36.8 cm using a custom float level controller. The total liquid level of the reactor was therefore set to 3.65 L. A mixer with a straight blade impeller operating at 47 RPM was utilized to keep the biomass in suspension during the un-aerated period. During the aerobic phase of the cycle, air was provided by a Hi-Blow HP-20 air pump.   38  Airflow was fed through a cylindrical air stone at 14.8 L/min to provide a superficial airflow velocity of 2.4 cm/s. During the decant phase, the liquid level of the reactor decreased to the level of port 2, leaving approximately 1.5 L of liquid remaining in the reactor. This resulted in a Volume Exchange Ratio (VER) of 59%. The initial seed sludge was taken from a Membrane Enhanced UCT Process operated at the UBC Pilot Plant. At the time of seed sludge acquisition, the process was operated at an SRT of 25 days.    39   Figure 14. Schematic of sequencing batch reactor setup   40  4.2 Sequencing Batch Reactor (SBR) Operation In the study, the SBR had four modes of operation in which it underwent an un-aerated phase, aerated phase, a settling phase, and a decant phase. A fast bottom feeding was initiated at the beginning of the un-aerated phase and ceased when a low level was sensed by the level sensor. During the un-aerated phase, a mixer was active to ensure suspension of the biomass. At the end of the un-aerated phase, the mixer was deactivated and the reactor aerated at a superficial airflow velocity of 2.4 cm/s. The DO level was left uncontrolled, and quickly rose to saturation. After the settling phase, the treated effluent was decanted through the decant port. An illustration of the cycling is show in Figure 15.   Figure 15. Sequencing batch reactor cycling Total cycle time was set at a constant 2 hours. The total cycle time of 2 hours was chosen to increase the OLR of the system. The cycle time distributions are listed in Table 1.    41  Table 1. Cycle time distribution for start-up and operation of SBR Phase Un-aerated Aerated Settling Decant Start-up  40 minutes 40 – 76 minutes 30 – 2 minutes 10 – 2 minutes Operational 40 minutes 78 minutes 1 minute 1 minute During start-up, the settling and decant times were slowly reduced to ensure retention of the fast settling biomass. As the settling and decant times were reduced, the aerated period was increased proportionally to maintain an overall cycle time of 2 hours. The reduction in settling time resulted in an increase in minimum settling velocity from 0.42 m/h to 12.57 m/h. Due to the constant cycle time of 2 hours, the Hydraulic Retention Time (HRT) was constant throughout the experiment at 3.39 hours. During the start-up and operational phase, no intentional wasting of biomass occurred to prevent excessive wasting of granular biomass. Thus, this left the SRT of the reactor uncontrolled. In all, the reactor was operated continuously for 281 days with the operational phase beginning at approximately the 165th day.  4.3 Apparatus Feed Tank Feed for the experiment was anaerobically stored in a sealed 6000 L holding tank. Primary effluent from the Annacis Wastewater Treatment Plant was transferred into the tank on a bi-weekly to monthly basis. COD of the stored wastewater at times was increased through the addition of 1 to 2 kg of refined white sugar to increase the Organic Loading Rate (OLR) of the SBR. A submersible pump was placed within the reactor to ensure the holding tank remained completely mixed.    42  4.4 Reactor Maintenance  The reactor was taken offline on a weekly to bi-weekly basis to remove biofilm growth on the reactor walls, mixer blade, and aerator. After draining of the reactor, the biofilm was removed with a dilute bleach solution in addition to mechanical scrubbing.  4.5 Analytical Methods 4.5.1 Performance Measurements To quantify the performance of the reactor, a multitude of parameters (see Table 2) were measured for the reactor influent, reactor MLSS, and reactor effluent. Influent samples were collected as grab samples from the sample port connected to the feed line during feeding of the reactor. Reactor samples were taken during the aeration phase of the reactor cycle to ensure a thoroughly mixed sample. Reactor effluent was dispensed into a collection receptacle and thoroughly mixed prior to obtaining a grab sample. Table 2 indicates the location of measurements for each parameter. The analytical methods for each parameter are presented in Table 3, and follow Standard Methods (APHA, 2005). Table 2. Parameter measurement locations Analyte Influent MLSS Effluent Frequency pH Yes No Yes 3-5/week Alkalinity Yes No Yes 3-5/week COD Yes No Yes 3-5/week sCOD Yes No Yes 3-5/week VFA Yes* No No ~3-5/week NH3-N Yes No Yes 3-5/week NO3-N Yes No Yes 3-5/week NO2-N Yes No Yes 3-5/week Ortho-P Yes No Yes 3-5/week TSS/VSS Yes Yes Yes 3-5/week SVI No Yes* No 2-3/month Particle Size No Yes* No 3-5/week * No replicates taken    43  Table 3. Parameter analytical methods and preservation techniques Analyte Analytical Method Instrument Preservation Method Container pH Electrometric Method  Standard Methods 4500 H+ Mantech TitraSip SA Immediate Analysis 300 mL HDPE Alkalinity Titration Method  Standard Methods 2320 B Mantech TitraSip SA Immediate Analysis 300 mL HDPE COD Closed Reflux Calorimetric Method  Standard Methods 5220 D Hach DR2800 Spectrophotometer  Hach Model 45600 COD Reactor  Acidification with H2SO4 to pH<2  Refrigeration to 4°C 10 mL PP Vial sCOD Closed Reflux Calorimetric   Standard Methods 5220 D Hach DR2800 Spectrophotometer  Hach Model 45600 COD Reactor  Filtered with 0.45 μm nitrocellulose filters  Acidification with H2SO4 to pH<2  Refrigeration to 4°C 10 ml PP Vial VFA Gas Chromatographic Method  Standard Methods 5560 D HP5580 Series 2  Filtered with 0.45 μm nitrocellulose filters  Acidification with H3PO4 to pH<2  Refrigeration to 4°C GC Vial NH3-N  Flow Injection Analysis   Standard Methods 4500-NH3 H Lachat Quikchem 8000  Filtered with 0.45 μm nitrocellulose filters  Acidification with H2SO4 to pH<2  Refrigeration to 4°C PS Culture Tube 13x100 mm   44  NO3-N Cadmium Reduction Flow Injection Method  Standard Methods 4500-NO3- I Lachat Quikchem 8000  Filtered with 0.45 μm nitrocellulose filters  Preserved with Phenylmercuric Acetate Solution  Refrigeration to 4°C PS Culture Tube 13x100 mm NO2-N Colorimetric Method  Standard Methods 4500-NO2- B Lachat Quikchem 8000  Filtered with 0.45 μm nitrocellulose filters  Preserved with Phenylmercuric Acetate Solution  Refrigeration to 4°C PS Culture Tube 13x100 mm Ortho-P Flow Injection Analysis for Orthophosphate  Standard Methods 4500 Ortho-P G Lachat Quikchem 8000  Filtered with 0.45 μm nitrocellulose filters  Acidification with H2SO4 to pH<2  Refrigeration to 4°C PS Culture Tube 13x100 mm TSS/VSS Standard Methods 2540 D/E Ohaus Adventurer AR0640 Refrigeration to 4°C  300 mL HDPE Container  50 mL PP Centrifuge Vial SVI Standard Methods 2710 D n/a Immediate Analysis n/a Particle Size Light Scattering Method  Standard Methods 2560 D Mastersizer 2000 with Hydro 2000S auto sampler Refrigeration to 4°C 50 mL PP Centrifuge Vial     45  4.5.2 Cycle Measurements Cycle profiles of Ortho-P, NH3-N, VFA, pH, and DO were constructed through sampling from the reactor at 10 minute intervals throughout a 2 hour reactor cycle. Sample preservation, analysis, and storage methods were identical to the performance measurements, as indicated in Table 3. pH measurements were obtained using a portable Oakton pH 11 meter and Oakton pH probe with ATC. DO measurements were made using an Oakton Waterproof DO 300 Meter and galvanic probe. 4.6 Quality Assurance/Quality Control All samples were either analyzed immediately or immediately preserved and stored in a 4°C refrigerator for, at maximum, a week prior to analysis. Analysis was conducted in triplicate for measured parameters, unless otherwise indicated in Table 2. Sample containment and preservation techniques for all measured parameters are listed in Table 3.  For the Lachat Quikchem 8000, calibrations were performed for each successive run. For the phosphate measurements, a 7 point calibration curve of 0.25, 0.5, 1, 2.5, 5, 10, and 25 mg/L PO4 was utilized. For ammonia measurements, a 7 point calibration curve of 0.5, 1, 2, 5, 10, 20, and 50 mg/L NH3 was utilized. For NO3/NO2 measurements, a 5 point calibration curve of 0.5, 1, 5, 10, and 25 mg/L NO3/NO2 was utilized. MDL measurements for the Lachat Quikchem 8000 were also conducted using 0.05 mg/L spiked solutions of PO4, NH3, and NO3/NO2. 9 to 10 replicates were measured for each spiked parameter and the MDLs were calculated to be 0.04 mg/L for PO4, 0.02 mg/L for NH3, and 0.03 mg/L for NO3/NO2. See Appendix B for supplementary MDL calculations.    46  For each successive batch of prepared COD vials, a calibration curve was prepared using KHP standards. Vials were randomly selected from the batch and screened for optical clarity prior to use in creating the calibration curve. For the high range COD solution, a 6 point calibration curve was created using 0, 100, 200, 300, 400, and 500 mg/L COD equivalent KHP solutions. For the low range COD solutions, a 5 point calibration curve was created using 0, 50, 100, 150, and 200 mg/L COD equivalent KHP solutions.  For the HP5580 Series 2 Gas Chromatograph, a calibration curve was made for each run for VFA analysis. A 6 point calibration curve was created using 2, 10, 20, 50, 100, and 200 mg/L for acetic, propionic, n-butyric, iso-butyric, n-valeric, and iso-valeric acids. The Oakton pH meter and ManTech TitraSip SA were calibrated before each use using a 3 point calibration with pH 4, 7, and 10 buffers. The Oakton Waterproof DO 300 Meter was calibrated using a 2 point calibration with a sodium sulfite solution and an oxygen saturated solution. 4.7 Statistical Analysis All error bars constructed from the data refer to 95% confidence intervals from 3 to 6 replicates. Any statistical difference was determined to a 95% confidence, if the confidence intervals did not overlap.    47  5 Results and Discussion 5.1 Reactor Start-up The initial seed sludge utilized for the reactor start-up had an initial SVI30 of 133 mL/mg and D50 of 41 μm. The sequencing batch reactor was seeded with 2 g VSS/L of sludge obtained from the UBC Pilot Plant (see Section 4.1 for details regarding the UBC Pilot Plant process). The reactor was started on October 27, 2013, with an initial cycle distribution of a 40 minute un-aerated period, 40 minute aerated period, 30 minute settling period, and a 10 minute decant. The long settling and decant times were utilized to prevent excessive biomass washout and ensure biomass acclimatization. As seen in Table 4, the settling times and decant were decreased while proportionally increasing the aeration time, to maintain an overall cycle time of 2 hours. The time required for a sludge bed to form during the settling period was used to determine the length of the settling period. This approach was taken to maintain selective pressure on the system to gradually increase washout of floccular biomass. As such, the standard rule-of-thumb of 3 SRTs to achieve steady state was not followed. Eventually, the decant time was decreased to accelerate the washout of non-settled biomass. One effect of decreasing settling times and decant times was an increase in biomass particle size, as seen in Figure 17.  Table 4. Reactor cycle distribution during start-up phase Phase 1 2 3 4 5 6 End Date (d) 20 27 48 63 151 164 Un-aerated (mins) 40 40 40 40 40 40 Aerated (mins) 40 60 62 71 72 76 Settling (mins) 30 10 8 5 4 2 Decant (mins) 10 10 10 4 4 2 Time (mins) 120 120 120 120 120 120   48  Initially, the reactor was fed with stored primary effluent from the Annacis Wastewater Treatment Plant. To increase the OLR of the system, sugar was added to the system on the 82st day. The amount of sugar required stoichiometrically calculated from the difference in COD required to achieve a desired OLR of 3 kg COD/m3/d.  The sugar underwent anaerobic fermentation in the storage tank which caused a significant increase of acetic and propionic acids. Characteristics of the feed, pre and post sugar addition can be seen in Table 5.  Table 5. Feed characteristics during reactor start-up   Pre Sugar Post Sugar Average ± 95% CI n Average ± 95% CI n COD (mg COD/L) 202 ± 19 37 357 ± 23 30 OLR (kg COD/m3/d) 1.43 ± 0.14 37 2.53 ± 0.18 30 sCOD (mg COD/L) 69 ± 9 42 243 ± 13 29 NH3 (mg/L) 32.8 ± 1.9 39 33.8 ± 2 30 PO4 (mg/L) 3.41 ± 0.14 39 2.81 ± 0.25 30 TSS (mg/L) 42.2 ± 5.9 40 39.2 ± 7.3 28 VFA (mg COD/L) 40.6 ± 15.57 26 189.4 ± 11.9 19 Alkalinity (mg CaCO3/L) 164 ± 7 41 132 ± 7 30 pH 6.90 ± 0.09 39 6.07 ± 0.05 30 It should be noted that the added supplemental sugar altered the characteristics of the feed to beyond what is typically seen in municipal wastewater. The rationale for the additional carbon was to increase the OLR to a level which would allow for granulation. Due to operational limitations of the apparatus, sugar addition was limited to the feed tank.  With the increase in OLR, the granular-type biomass developed a denser appearance and reduction in filamentous outgrowth. These findings correlate to the findings by Tay et al.   49  (2004) regarding the effects of OLR on granulation. By approximately the 166th of operation, white-coloured, gelatinous-type, granulation was achieved in the reactor. These granules had a D50 of 513 ± 60 μm. This median particle size is close to the critical size range of granules 600 μm to 800 μm, as stated by Verawaty et al. (2013). The progression of sludge morphology and particle size during start-up can be seen in Figure 16 and Figure 17 respectively.    50   Figure 16. Progression of sludge morphology during start-up period Seed Sludge Day 11 Day 99 Day 156 Day 128 Day 46   51   Figure 17. Particle size progression during the start-up period 0200400600800100012001400160018000 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160Particle Size (um) Day D10D50D90Phase 1 Phase 5 Phase 6 Phase 3 Phase 4 Phase 2 External Carbon Addition   52  5.1.1 Start-up Performance Prior to sugar addition, the removal efficiencies of PO4, NH3, and COD were as presented in Figure 18, Figure 19, and Figure 20. The average PO4, NH3, and COD removal efficiencies for that that period were 17.9% ± 7.0%, 7.0% ± 3.1 %, and 79.8% ± 3.2%. The variability was likely due to issues with obtaining consistent feed and feed storage. After sugar addition, average PO4 and NH3 removal efficiencies increased significantly to 74.2% ± 4.9% and 30.2% ± 4.0% while COD removal increased slightly to 87.8% ± 1.8%.Although the COD removal rates remained relatively high throughout the start-up, the system did not develop a feast-famine regime.  The SRT of the system varied due to the lack of SRT control through sludge wasting from the reactor. During the initial acclimation phase, the SRT increased to 40 days; however, it soon dropped as the settling and decant times were reduced, increasing the washout of floccular biomass. As the start-up period progressed, the SRT of the system remained below 10 days, as seen in Figure 21.    During the start-up period, it appeared that nitrification/denitrification was not occurring due to the lack of significant amounts of nitrites and nitrates in the reactor effluent (as seen in Figure 22). The lack of nitrification and denitrification is likely due to less than optimum feed pH of 7.5 to 8 and system SRT of 10 to 20 days at 10°C (Metcalf & Eddy et al., 2002). To achieve a suitable pH for nitrification, the average alkalinity needed would be approximately 235 mg CaCO3/L, which is far higher than the average alkalinity of 164 ± 7 mg CaCO3/L and 132 ± mg CaCO3/L pre and post sugar addition (Metcalf & Eddy et al., 2002).   53   Figure 18. PO4 removal during start-up period 0123450102030405060708090100110120130140150160PO4 Concentration (mg/L) Days Feed Treated Effluent0%10%20%30%40%50%60%70%80%90%100%PO4 % Removal  Phase 1  Phase 1 Phase 3 Phase 4 Phase 5 Phase 6 Phase 2  Phase 1 External Carbon Addition MDL: 0.04 mg/L   54   Figure 19. NH3 removal during start-up period 010203040500102030405060708090100110120130140150160NH3 Concentration (mg/L) Days Feed Treated Effluent0%10%20%30%40%50%60%70%80%90%100%NH3 % Removal  Phase 1  Phase 1 Phase 3 Phase 4 Phase 5 Phase 6 Phase 2  Phase 1 External Carbon Addition MDL: 0.02 mg/L   55   Figure 20. COD removal and OLR during start-up period 01002003004005006000102030405060708090100110120130140150160COD Concentration (mg/L) Days Feed Treated Effluent00.511.522.533.544.550%10%20%30%40%50%60%70%80%90%100%OLR (kg/m3/d) COD % Removal  % Removal OLRPhase 1  Phase 1 Phase 3 Phase 4 Phase 5 Phase 6 Phase 2  Phase 1 External Carbon Addition   56    Figure 21. Reactor and treated effluent VSS concentrations with SRT during start-up period 0250500750100012501500175020002250250027503000325035000102030405060708090100110120130140150160VSS (mg/L) Days MLVSS Reactor Effluent051015202530354045SRT (d) Phase 2 Phase 1  Phase 1 Phase 3 Phase 4 Phase 5 Phase 6 External Carbon Addition   57   Figure 22. NO2 and NO3 concentrations in treated effluent during start-up period   0.00.10.20.30.40.50.60.70.80.91.00102030405060708090100110120130140150160NO2 and NO3 Concentration (mg/L) Days NO2 Treated Effluent NO3 Treated EffluentPhase 1  Phase 1 Phase 4 Phase 6 Phase 2  Phase 1 External Carbon Addition MDL: 0.03 mg/L Phase 3  Phase 1 Phase 5   58  5.2 Reactor Performance with Varying Influent pH Once granulation was achieved, the system entered into phase 7, which consisted of a 40 minute anaerobic period, 78 minute aerobic period, 1 minute settling period, and 1 minute decant period. During the operational part of phase 7, the COD of the influent was changed by adding varying amount of sugar to the feed tank. As a result of the increased sugar addition in phase 7-2, the pH of the feed decreased compared to phase 7-1 due to an increase in acetogenesis (as seen in Table 6).  Table 6. Phase 7 influent characteristics   Phase 7-1 Feed Phase 7-2 Feed Average ± 95% CI n Average ± 95% CI n COD (mg COD/L) 295 ± 27 17 507 ± 45 16 OLR (kg COD/m3/d) 2.08 ± 0.19 17 3.58 ± 0.32 17 sCOD (mg COD/L) 190 ± 27 16 361 ± 33 16 NH3 (mg/L) 33.4 ± 2.4 17 42.5 ± 2.4 16 PO4 (mg/L) 3.19 ± 0.35 17 4.28 ± 0.19 16 TSS (mg/L) 46.5 ± 12.3 17 54.6 ± 7.3 16 VFA (mg COD/L) 129.1 ± 31 8 307.4 ± 30.2 16 Alkalinity (mg CaCO3/L) 138 ± 10 17 122 ± 9 16 pH 6.23 ± 0.06 17 5.62 ± 0.12 16 Table 7. Phase 7 treated effluent characteristics   Phase 7-1 Treated Effluent Phase 7-2 Treated Effluent Average ± 95% CI n Average ± 95% CI n sCOD (mg COD/L) 46 ± 8 16 65 ± 7 15 NH3 (mg/L) 26.4 ± 2.4 17 26.4 ± 2.4 16 PO4 (mg/L) 1.49 ± 0.32 17 1.78 ± 0.25 15 TSS (mg/L) 112.1 ± 37 16 148.9 ± 29.4 16 Alkalinity (mg CaCO3/L) 142 ± 6 17 147 ± 7 16 pH 7.9 ± 0.04 17 7.77 ± 0.08 16     59  The average PO4, NH3, and COD removal efficiencies for phase 7-1 were 54.4 % ± 8.3%, 21.9% ± 4.1 %, and 84.0% ± 3.0%. During phase 7-2, the average PO4 remained constant at 58.9% ± 4.7%, while NH3 and COD removal increased slightly to 37.9% ± 4.7% and 87.1% ± 0.9%. The performance data can be seen in Figure 23, Figure 24, and Figure 25. As with the start-up period, the system did not develop a feast-famine regime.   60   Figure 23. PO4 removal under varying influent pH 012345166176186196206216226236246256266PO4 Concentration (mg/L) Days Feed Treated Effluent0%10%20%30%40%50%60%70%80%90%100%PO4 % Removal  Phase 7-1 Phase 7-2 MDL: 0.04 mg/L   61   Figure 24. NH3 under varying influent pH 01020304050166176186196206216226236246256266NH3 Concentration (mg/L) Days Feed Treated Effluent0%10%20%30%40%50%60%70%80%90%100%NH3 % Removal  Phase 7-1 Phase 7-2 MDL: 0.02 mg/L   62   Figure 25. COD removal and OLR under varying influent pH 0100200300400500600700800166176186196206216226236246256266COD Concentration (mg/L) Days Feed Treated Effluent01234560%10%20%30%40%50%60%70%80%90%100%OLR (kg/m3/d) COD % Removal  % Removal OLRPhase 7-1 Phase 7-2   63   Figure 26. Reactor and treated effluent VSS concentration and SRT under varying influent pH 0200400600800100012001400160018002000166176186196206216226236246256266MLVSS mg/L Days Feed Reactor Effluent012345678910SRT (d) Phase 7-1 Phase 7-2   64   Figure 27. NO2 and NO3 concentration in treated effluent under varying influent pH 0.000.050.100.150.200.250.30166176186196206216226236246256266NO2 and NO3 Concentration (mg/L) Days NO2 Treated Effluent NO3 Treated EffluentPhase 7-1 Phase 7-2 MDL: 0.03 mg/L   65  5.3 Reactor Cycle Profiles with Varying Influent pH Cycle profiles constructed during phases 7-1 and 7-2 showed a significant difference of PO4 release during the anaerobic period. During phase 7-1, a release of PO4 occurred concurrently to an uptake in propionic acid during the anaerobic period (as shown in Figure 28 and Figure 29). This selective decrease of propionate over acetate is likely due to the preference of PAOs to uptake propionate over acetate (Bitton, 2011) (Oehmen et al., 2006). Subsequently, in the aerobic portion of the cycle, an uptake of PO4 was witnessed.  At the lower average influent pH conditions of phase 7-2, no significant PO4 release and VFA uptake occurred during the anaerobic period, as seen in Figure 28 and Figure 29, thus indicating an absence of PAOs in the system. This may be a due to the pH during the anaerobic period (as seen in Figure 30) being significantly lower than the optimal pH of 7 to 8 as required by PAOs (Bitton, 2011).  The ambient pH affects the cells through the formation of an increased electrical potential across the cell membrane. This increased electrical potential requires more energy GAOs and PAOs to overcome to accomplish VFA uptake during the anaerobic period. PAOs have a selective advantage in overcoming this energy barrier by utilizing the internal energy reserve provided through polyphosphate degradation. The critical point at which  PAOs are able to uptake VFAs faster than GAOs is estimated to be above a pH of 7.25 (Oehmen et al., 2007).  As the systems entered into the aerobic phase the pH rose rapidly during the first 20 minutes of aeration, and eventually leveled off at an average pH of 7.6 for both phases 7-  66  1 and 7-2 (Figure 30). The rise in pH can be attributed to CO2 stripping induced by the aeration system. This is evident during the initial 20 minutes of the aerobic periods where the rapid rise in pH coincides with the rapid rise in oxygen saturation in the reactor (as seen in Figure 31).    NH3 removal, in the granular system, occurred in a linear fashion during the aerobic period, with a higher removal occurring with lower influent pH conditions, as seen in Figure 28. One possible explanation for nitrogen removal in a granular system is simultaneous nitrification-denitrification (SND) occurring within the aerobic/anoxic zones of the granule. For phase 7-1, conditions for SND existed during the first 10 minutes of the aerobic period, with a sufficient DO concentration (~4 mg/L as seen in Figure 31) and readily biodegradable substrate concentration (as seen in Figure 29) available. For phase 7-2, readily biodegradable substrate and DO levels were sufficient until approximately 50 minutes into the aerobic period. However, a lack nitrates and nitrites in the reactor effluent (as shown in Figure 27), low pH conditions during critical portions of the reaction period (as shown in Figure 30), short overall reaction times, low system SRT, and absence of denitrification induced phosphorus precipitation within the granule, indicated that the loss of ammonia was not a result of SND.  Additionally, the pH conditions within the reactor were such that ammonia remained in its ionized form, thereby negating the possibility of volatilization being a primary source of nitrogen removal. Likely, the majority of ammonia removal is attributed to biological uptake for biological synthesis, with higher removal occurring at lower pH conditions due to a greater amount of available carbon substrate.   67    Figure 28. Normalized PO4 and NH3 cycle profiles under varying influent pH of 6.23 (n=8) and 5.62 (n=4)00.511.522.50 10 20 30 40 50 60 70 80 90 100 110 120Normalized PO4 Concentration Time (mins) pH 6.23pH 5.62Anaerobic Aerobic 00.20.40.60.810 10 20 30 40 50 60 70 80 90 100 110 120Normalized NH3 Concentration Time (mins) Anaerobic Aerobic   68    Figure 29. Normalized acetic and propionic acid concentrations at an influent pH of (a) 6.23 (n=4) and (b) 5.62 (n=4)  00.20.40.60.811.21.40 10 20 30 40 50 60 70 80 90 100 110 120Normalized VFA Concentration Time (mins) AceticPropionicAnaerobic Aerobic (a) 00.20.40.60.811.21.40 10 20 30 40 50 60 70 80 90 100 110 120Normalized VFA Concentration Time (mins) AceticPropionicAnaerobic Aerobic (b)   69    Figure 30. Reactor pH cycle profiles under varying influent pH of 6.23 (n=6) and 5.62 (n=3)   Figure 31. Oxygen saturation profiles under varying influent pH of 6.23 (n=6) and 5.62 (n=5) 55.566.577.588.590 10 20 30 40 50 60 70 80 90 100 110 120pH Time (mins) pH 6.23pH 5.62Anaerobic Aerobic 0%20%40%60%80%100%120%0 10 20 30 40 50 60 70 80 90 100 110 120Oxygen Saturation  Time (mins) pH 6.23pH 5.62Anaerobic Aerobic   70  5.4 Granule Characteristics and Morphology with Varying Influent pH The granules formed during phase 7 were largely spherical in shape and possessed a soft, gelatinous texture. The particle sizes for the granules between the two different influent conditions were relatively similar (as shown in Figure 32), with a D50 of 591.83 ± 93.63 μm and 659.43 ± 110.49 μm. The median size of the granules generally stayed within the critical granule size of 600 to 800 μm (Verawaty et al., 2013) throughout the experimental phase. After approximately 30 days of operation, filamentous growth was observed on granules fed with a lower pH influent, while granules with no evidence of filamentous growth were observed with granules fed with the higher pH influent. A visual representation of this can be seen in Figure 33. The increase amount of filamentous outgrowths with low pH influent may result in long term instability for the granular system. This may be due to the microbial ecology of the granules shifting towards GAOs or other heterotrophs, which in turn, leads to instability. The SVI10 and SVI30 of the granules were found to be equivalent at148.8 ± 28.9 mL/g and 157.5 ± 40.6 mL/g for phases 7-1 and 7-2. The lack of difference between SVI10 and SVI30 is indicative of the fast settling characteristics of granular biomass (Schwarzenbeck et al., 2004) (de Kreuk et al., 2005b). In addition to the SVI, the measured settling velocity of the granules was found to be 29.9 ± 4.5 m/h. However, in comparison to other studies conducted, the granules cultivated in this study possessed lower settling velocities and a higher SVI. The lower settling velocity and high SVI might be explained by low granule density, which is, in part, due to the lack of phosphorus precipitation in the granules. Due to the absence of a denitrification induced pH increase in the granules, favourable conditions for phosphorus precipitation within the granule likely did not   71  occur. This lack of dense phosphorus precipitate would result in low density granules, with lower settling characteristics. Fixed suspended solids in the granules were also found to be 8.5 ± 1.0% and 8.5 ± 1.0% of the total suspended solids during phases 7-1 and 7-2. The low amount of fixed suspended solids further indicates low amounts of phosphorus precipitation in the granules.     72    Figure 32. Particle size under different influent pH conditions  Figure 33. Granule appearance after approximately 30 days of operation with (a) 6.23 pH (phase 7-1) (b) 5.62 pH (phase 7-2) influent 020040060080010001200140016001800166 176 186 196 206 216 226 236 246 256Particle Size (μm) Day D10D50D90(a) (b) Phase 7-1 Phase 7-2   73  6 Summary and Conclusions Successful granulation, using municipal wastewater from the Annacis Wastewater Treatment Plant with supplemental carbon addition, was achieved after 166 days of start-up. To achieve granulation, a slow reduction in settling and decant times occurred, preventing excessive washout of the acclimatized biomass. As settling and decant times were reduced, an increase in biomass particle size was noticed. The granular type structure had a fluffy structure with noticeable filamentous outgrowths. By the 82nd day of start-up, sugar was added to the feed tank to increase the OLR of the system from an average of 1.43 ± 0.14 kg COD/m3/d to 2.53 ± 0.18 kg COD/m3/d. In addition to increasing the COD of the feed, sugar addition resulted in acetogenesis occurring within the storage tank, thereby causing a decrease in pH and alkalinity to 6.07 ± 0.05 and 132 ± 7 mg CaCO3/L, as well as an increase in VFA. After the OLR increase, the biomass developed a denser morphology with a noticeable reduction in filamentous outgrowths. By the 166th day, gelatinous, white-coloured granules appeared in the reactor. Throughout the start-up and operational phase of the experiment, minimal amounts of nitrites and nitrates were detected within the treated effluent from the reactor. This indicated that nitrification and denitrification were not occurring significantly within the granular system. The lack of nitrification and denitrification was possible due to compounding factors of low pH during a portion of the reaction period, lack of readily available biodegradable substrate, low system SRT, and short overall reaction time. Due to the lack of denitrification within the granules, the occurrence of chemical precipitation of phosphorus within the granule did not occur, which resulted in low density, slow settling granules and a high SVI.    74  During the experimental phase, differing amounts of sugar were added to the feed tank. As more sugar was added to the feed tank, the pH decreased proportionally due to increasing acetogenesis. From comparison of the cycle profiles of the two different conditions, it was noticed that at an average influent of pH 6.23 ± 0.06, a noticeable anaerobic release of PO4 occurred in conjunction with an uptake of propionate. Subsequently, during the aerobic portion, a noticeable uptake of PO4 occurred. At the conditions at the lower average influent pH of 5.62 ± 0.12, no anaerobic PO4 release or VFA uptake occurred. This indicated that at a lower influent pH, biological phosphorus removal did not occur, likely due to a shift in the microbial ecology to favour GAOs or heterotrophs over PAOs. In all, the cultivation of granules with low pH, low alkalinity municipal wastewater is possible with the addition of supplemental carbon. However, some issues with utilizing low pH, low alkalinity wastewaters include reduced nutrient removal, lower granule density, and potentially long term instability. The reduction in nutrient removal performance is due to the lack of nitrification/denitrification and shifting of the microbial ecology to GAOs or heterotrophs over PAOs. The lower granule density is a result of the absence of denitrification induced chemical phosphorus precipitation within the granule. As such, the lower granule density results in slower granule settling velocities and increased SVI.  Some recommendations to improve the nutrient removal performance and granule characteristics include the addition of calcium hydroxide to the feed to increase pH to approximately pH 7 to 8. The increase in pH would allow for nitrification and denitrification to occur, thereby causing an increase in NH3 removal and an increase in   75  granule density, settling velocity, and reduction in SVI. Additionally, an increase in PO4 removal would occur due to shifting the microbial ecology of the system to favour PAOs. Another potential effect could be an increase in long term stability, due to the prevalence of PAOs in the system.  6.1 Recommendations for Future Work A recommendation for potential future work includes the investigation on the efficacy of aerobic granular systems on the removal of particulate matter. Due to the hydrophobic nature of the EPS matrix, it is postulated that the granule would provide a better adsorption surface for particulate matter and transport medium for exoenzymes, thereby allowing for increased hydrolysis of the particulate matter. Likely, the system will require a longer SRT to allow for the enzymatic systems to develop, which can be facilitated through the formation of a dense granular system. At lower SRT is postulated that protozoa on the granule surface would facilitate particulate removal. Another recommendation for future work would be to develop a stable, continuous aerobic granular system to overcome the inherent disadvantages of SBR systems. A potential process configuration could be in the form of two CSTRs in series, alternating between anaerobic and aerobic conditions, followed by a specially designed clarifier to facilitate adequate selection pressure. Listed below are special considerations that should be accounted for during the design phase. 1. Ensure adequate shear forces are encountered during the aerobic phase. This could be accomplished through superficial airflow velocity and/or a mixer.   76  2. The clarifier should be able to apply adjustable selection pressure to the system to prevent excessive washout of biomass during the transition from floccular to granular biomass. Additionally, a metagenomic analysis could be conducted on two granular systems; with one operating with a low pH influent, and another operating with a standard pH influent. This would further prove the effect of influent pH on the microbial ecology in a granular system.    77  References Adav, S., Lee, D., Show, K., Tay, J., 2008. Aerobic granular sludge: recent advances. Biotechnol. Adv. 26, 411–23. Angela, M., Béatrice, B., Mathieu, S., 2011. Biologically induced phosphorus precipitation in aerobic granular sludge process. Water Res. 45, 3776–86. APHA, 2005. Standard methods for the examination of water & wasterwater, 21st ed. American Public Health Association, Washington DC. Arvin, E., Kristensen, H., 1982. Effect of denitrification on the pH in biofilms. Water Sci. Technol. 14, 833–848. Bassin, J.P., Kleerebezem, R., Dezotti, M., van Loosdrecht, M.C.M., 2012. Simultaneous nitrogen and phosphate removal in aerobic granular sludge reactors operated at different temperatures. Water Res. 46, 3805–16. Beun, J.J., Hendriks, A., van Loosdrecht, M.C.., Morgenroth, E., Wilderer, P.., Heijnen, J.., 1999. Aerobic granulation in a sequencing batch reactor. Water Res. 33, 2283–2290. Bitton, G., 2011. Wastewater microbiology, 4th ed. John Wiley & Sons, Hoboken, NJ. Chen, M.Y., Lee, D.J., Tay, J.H., 2007. Distribution of extracellular polymeric substances in aerobic granules. Appl. Microbiol. Biotechnol. 73, 1463–9. de Bruin, L.M.M., de Kreuk, M.K., Van Der Roest, H.F.R., Uijterlinde, C., Van Loosdrecht, M.C.M., 2004. Aerobic granular sludge technology: an alternative to activated sludge? Water Sci. Technol. 49, 1–7. de Kreuk, M.K., Heijnen, J.J., van Loosdrecht, M.C.M., 2005a. Simultaneous COD, nitrogen, and phosphate removal by aerobic granular sludge. Biotechnol. Bioeng. 90, 761–9. de Kreuk, M.K., Kishida, N., van Loosdrecht, M.C.M., 2007. Aerobic granular sludge – state of the art. Water Sci. amd Technol. 55, 75. de Kreuk, M.K., McSwain, B.S., Bathe, S., Tay, J., Schwarzenbeck, S., 2005b. Aerobic granular sludge, water and environmental services series. IWA Pulishing, Munich. de Kreuk, M.K., van Loosdrecht, M.C.M., 2004. Selection of slow growing organisms as a means for improving aerobic granular sludge stability. Water Sci. Technol. 49, 9–17.   78  Galloway, J.N., 2003. The global nitrogen cycle, in: Schelesinger, W. (Ed.), Biogeochemistry Vol. 8. Treatise on Geochemistry. Elsevier, San Diego, CA, pp. 557–583. Giesen, A., van Loosdrecht, M., de Bruin, B., van der Roest, H., Pronk, M., 2013. Full -scale experiences with aerobic granular biomass technology for treatment of urban and industrial wastewater, in: IWW Amsterdam. Amsterdam, Netherlands. Liu, Y., Liu, Q.-S., 2006. Causes and control of filamentous growth in aerobic granular sludge sequencing batch reactors. Biotechnol. Adv. 24, 115–27. Liu, Y., Tay, J.-H., 2002. The essential role of hydrodynamic shear force in the formation of biofilm and granular sludge. Water Res. 36, 1653–65. Liu, Y.-Q., Tay, J.-H., 2008. Influence of starvation time on formation and stability of aerobic granules in sequencing batch reactors. Bioresour. Technol. 99, 980–5. Liu, Y.-Q., Wu, W.-W., Tay, J.-H., Wang, J.-L., 2007. Starvation is not a prerequisite for the formation of aerobic granules. Appl. Microbiol. Biotechnol. 76, 211–6. McSwain, B.S., Irvine, R.L., Hausner, M., Wilderer, P.A., 2005. Composition and distribution of extracellular polymeric substances in aerobic flocs and granular sludge. Appl. Environ. Microbiol. 71. Metcalf & Eddy, Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2002. Wastewater engineering: treatment and reuse, 4th Editio. ed. McGraw-Hill, New York. Miller, M.B., Bassler, B.L., 2001. Quorum sensing in bacteria. Annu. Rev. Microbiol. 55, 165–199. Oehmen, A., Lemos, P.C., Carvalho, G., Yuan, Z., Keller, J., Blackall, L.L., Reis, M. a M., 2007. Advances in enhanced biological phosphorus removal: from micro to macro scale. Water Res. 41, 2271–300. Oehmen, A., Saunders, A.M., Vives, M.T., Yuan, Z., Keller, J., 2006. Competition between polyphosphate and glycogen accumulating organisms in enhanced biological phosphorus removal systems with acetate and propionate as carbon sources. J. Biotechnol. 123, 22–32. Oehmen, A., Teresa Vives, M., Lu, H., Yuan, Z., Keller, J., 2005. The effect of pH on the competition between polyphosphate-accumulating organisms and glycogen-accumulating organisms. Water Res. 39, 3727–37. Pijuan, M., 2013. Mechanisms of aerobic granular sludge formation, in: WEF/IWA Nutrient Removal and Recovery. Vancouver, BC.   79  Prepas, E., Charette, T., 2003. Worldwide eutrophication of water bodies : causes , concerns , controls, in: Schlesinger, W. (Ed.), Environmental Geochemistry Vol 9. Treatise on Geochemistry. Elsevier, San Diego, pp. 311–331. Qin, L., Liu, Y., Tay, J.-H., 2004. Effect of settling time on aerobic granulation in sequencing batch reactor. Biochem. Eng. J. 21, 47–52. Ren, T., Yu, H., Li, X., 2010. The quorum-sensing effect of aerobic granules on bacterial adhesion, biofilm formation, and sludge granulation. Appl. Microbiol. Biotechnol. 88, 789–97. Ruttenberg, K.C., 2003. The global phosphorus cycle, in: Schlesinger, W. (Ed.), Biogeochemistry Vol. 8. Treatise on Geochemistry. Elsevier, San Diego, CA, pp. 585–643. Schwarzenbeck, N., Erley, R., Wilderer, P. a, 2004. Aerobic granular sludge in an SBR-system treating wastewater rich in particulate matter. Water Sci. Technol. 49, 41–6. Seviour, T., Pijuan, M., Nicholson, T., Keller, J., Yuan, Z., 2009. Gel-forming exopolysaccharides explain basic differences between structures of aerobic sludge granules and floccular sludges. Water Res. 43, 4469–78. Tan, C.H., Koh, K.S., Xie, C., Tay, M., Zhou, Y., Williams, R., Ng, W.J., Rice, S. a, Kjelleberg, S., 2014. The role of quorum sensing signalling in EPS production and the assembly of a sludge community into aerobic granules. ISME J. 8, 1186–97. Tay, J.-H., Liu, Q.S., Liu, Y., 2004a. The effect of upflow air velocity on the structure of aerobic granules cultivated in a sequencing batch reactor. Water Sci. Technol. 49, 35–40. Tay, J.-H., Liu, Q.-S., Liu, Y., 2001. The role of cellular polysaccharides in the formation and stability of aerobic granules. Lett. Appl. Microbiol. 33, 222–226. Tay, J.-H., Pan, S., He, Y., Tiong, S., Tay, L., 2004b. Effect of rganic loading rate on aerobic granulation . I : reactor performance. J. Environ. Eng. 1094–1101. Tay, J.-H., Pan, S., He, Y., Tiong, S., Tay, L., 2005. Effect of organic loading rate on aerobic granulation . II : characteristics of aerobic granules. J. Environ. Eng. 130, 1102–1109. Verawaty, M., Tait, S., Pijuan, M., Yuan, Z., Bond, P.L., 2013. Breakage and growth towards a stable aerobic granule size during the treatment of wastewater. Water Res. 47, 5338–49. Wang, Z.-W., Li, Y., Zhou, J.-Q., Liu, Y., 2006. The influence of short-term starvation on aerobic granules. Process Biochem. 41, 2373–2378.   80  Water Environment Federation, 2010. Nutrient removal WEF manual of practice no. 34. McGraw-Hill, Alexandria, VA. Winkler, M.H., 2012. Magic granules. Delft University of Technology. Winkler, M.-K.H., Bassin, J.P., Kleerebezem, R., de Bruin, L.M.M., van den Brand, T.P.H., van Loosdrecht, M.C.M., 2011. Selective sludge removal in a segregated aerobic granular biomass system as a strategy to control PAO-GAO competition at high temperatures. Water Res. 45, 3291–9. Winkler, M.-K.H., Kleerebezem, R., Strous, M., Chandran, K., van Loosdrecht, M.C.M., 2013. Factors influencing the density of aerobic granular sludge. Appl. Microbiol. Biotechnol. 97, 7459–68. Zhou, Y., Hao, T.C., Yanghui, X., 2013. Cell-cell communication during granulation, in: WEF/IWA Nutrient Removal and Recovery. Vancouver, BC.     81  Appendices     82  Appendix A: Additional Photographs     83     Figure 34. Aerobic granular sequencing batch reactor setup with feed tank     84          Figure 35. Examples of granules grown during study   85  Appendix B: MDL Calculations    86  Sample PO4 Concentration (mg/L)      Spike Concentration (mg/L) 1  -    0.05 2 0.0743       3 0.0596       4 0.0592       5 0.0432       6 0.0416       7 0.0383       8 0.0403       9 0.0371       10 0.0388                                 Number of Points 9      Mean 0.048044      Standard Deviation 0.013098      S/N Ratio 3.66802              Results   Data Check MDL 0.037938   Is Spike < 10*MDL ? OK LCL 0.025626   Is MDL < Spike ? OK UCL 0.072681   2.5 < S/N < 10 ? OK LOQ 0.130982          00.010.020.030.040.050.060.070.08PO4 MDL with 95% Confidence Intervalmg/L   87  Sample NH3 Concentration (mg/L)      Spike Concentration (mg/L) 1 -     0.05 2 0.05       3 0.06       4 0.04       5 0.05       6 0.06       7 0.05       8 0.05       9 0.04       10 0.05                                       Number of Points 9      Mean 0.05      Standard Deviation 0.007071      S/N Ratio 7.071068              Results   Data Check MDL 0.020481   Is Spike < 10*MDL ? OK LCL 0.013834   Is MDL < Spike ? OK UCL 0.039237   2.5 < S/N < 10 ? OK LOQ 0.070711          00.010.020.030.040.05NH3 MDL with 95% Confidence Intervalmg/L   88  Sample NO2 Concentration (mg/L) NOx Concentration (mg/L) NO3 Concentration (mg/L)  Spike Concentration (mg/L) 1 0.01 0.09 0.08107  0.05 2 0.01 0.08 0.0664     3 0.02 0.08 0.0543  Column Efficiency 4 0.00 0.07 0.07933  1 5 0.01 0.08 0.07274     6 0.02 0.07 0.0475     7 0.01 0.07 0.06614     8 0.00 0.07 0.07105     9 0.01 0.09 0.07576     10 0.00 0.07 0.06635                                     Number of Points 10      Mean 0.068064      Standard Deviation 0.010582      S/N Ratio 6.432009              Results   Data Check MDL 0.029857   Is Spike < 10*MDL ? OK LCL 0.020536   Is MDL < Spike ? OK UCL 0.054507   2.5 < S/N < 10 ? OK LOQ 0.105821         00.010.020.030.040.050.06NO3 MDL with 95% Confidence Intervalmg/L 

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.24.1-0166061/manifest

Comment

Related Items