Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Coupling of phosphorus recovery to an enhanced biological phosphorus removal process through a sidestream… Srinivas, Hemanth Kumar 2006

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

Item Metadata

Download

Media
831-ubc_2007-0242a.pdf [ 5.7MB ]
Metadata
JSON: 831-1.0063252.json
JSON-LD: 831-1.0063252-ld.json
RDF/XML (Pretty): 831-1.0063252-rdf.xml
RDF/JSON: 831-1.0063252-rdf.json
Turtle: 831-1.0063252-turtle.txt
N-Triples: 831-1.0063252-rdf-ntriples.txt
Original Record: 831-1.0063252-source.json
Full Text
831-1.0063252-fulltext.txt
Citation
831-1.0063252.ris

Full Text

COUPLING OF PHOSPHORUS RECOVERY TO AN ENHANCED BIOLOGICAL PHOSPHORUS REMOVAL PROCESS THROUGH A SIDESTREAM: A PILOT SCALE STUDY by H E M A N T H K U M A R SRINIVAS B.E. (Environmental Engineering), University of Mysore, India 2001 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF APPLIED SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH C O L U M B I A November 2006 © Hemanth Kumar Srinivas, 2006 ABSTRACT This research was initiated with the purpose to remove and recover phosphorus simultaneously, by coupling a sidestream for phosphorus recovery to an enhanced biological phosphorus removal (EBPR) process. Sidestream process comprised of a phosphorus release unit along with a clarifier connected to the M A P (magnesium ammonium phosphate) crystallizer. To understand the sidestream process and to investigate the optimum operating conditions, the configuration was simulated using ASM2-Delft metabolic bio-P model on A Q U A S I M platform. The simulation exercise revealed that anoxic is the best zone to take sidestream when taken individually and anoxic/anaerobic when taken in combination. In both the cases N : P molar ratio of the supernatant was more than 1:1 which is essential for the recovery of phosphorus as struvite. The minimum hydraulic retention time required in the phosphate release unit was found to be around one to two hours, above which there was no considerable amount of phosphate released. Potentially, up to 78 % of the incoming phosphorus is estimated to be recovered by implementing sidestream technology. Based on the simulation results, sidestream process was successfully implemented at U B C pilot plant for both membrane and conventional enhanced biological phosphorous removal processes (MEBPR and CEBPR). The M A P crystallizer was used to recover phosphorus as struvite. Although, the recovery efficiency obtained was not very high (approximately 60 %), the refined conditions as suggested should yield better results. Sidestream wasting method to control the solid retention time of the process indicated selective increase of phosphorus accumulating organisms in the main EBPR process. Sidestream wasting also reduced the phosphorus rich sludge wasting from the aerobic zone of the EBPR process. Magnesium (Mg) was added to the influent of the process to increase the M g 2 + concentration in the sidestream supernatant to provide better conditions for struvite formation. The CEBPR and M E B P R processes experienced poor phosphorus removal after M g addition was started. More detailed investigation is suggested to look in to the effects of magnesium on the EBPR process along with potassium. n TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS . Hi LIST OF TABLES . vi LIST OF FIGURES '. vii LIST OF ABBREVIATIONS ix ACKNOWLEDGEMENTS xi CHAPTER 1 - INTRODUCTION 1 1.1 Phosphorus Removal and Recovery 1 1.2 Sidestream Process 2 1.3 Research Objectives 3 CHAPTER 2 - LITERATURE REVIEW 6 2.1 Modeling and Simulation of Wastewater Treatment Systems 6 2.1.1 ASM2-Delft bio-P model '. 7 2.2 Principle of the EBPR Process 9 2.2.1 Comeau-Wentzel model 9 2.2.2 Mino model 10 2.2.3 Process engineering 10 2.2.4 Membrane bioreactors 13 2.2.5 Advantages and disadvantages of membrane bioreactors 14 2.2.6 M B R coupled with EBPR process 15 2.2.7 Problems with biological nutrient removal processes 15 2.3 Sidestream Process for Phosphorus Recovery 16 2.3.1 M A P crystallizer 19 2.3.2 Supplementations required 19 2.3.3 Parameters affecting struvite formation 19 2.3.4 Solubility product of struvite 20 CHAPTER 3 - MATERIALS AND METHODS 22 3.1 Process description 22 3.1.1 Site description 22 3.1.2 EBPR configuration 22 3.1.3 Membrane 25 3.1.4 Sidestream configuration 25 3.2 Model and simulation software 26 3.3 Wasting method 27 3.4 Supplementations 28 3.4.1 Sodium bicarbonate 28 in 3.4.2 V F A addition 29 3.4.3 Magnesium 29 3.5 Pumps, mixer and pipelines details 29 3.6 Sampling plan and analytical procedures 30 3.7 MAP Crystallizer 31 3.7.1 Top and external clarifiers 31 3.7.2 Seeding and harvesting procedure 32 3.7.3 Chemical supplementation, storage tanks and pumps 33 3.7.4 pH control 34 3.7.5 Sampling plan 34 CHAPTER 4 - RESULTS AND DISCUSSION 36 4.1 Simulation of Sidestream Process..... 36 4.1.1 Sidestream configuration in A Q U A S I M 36 4.1.2 Source of the sidestream 37 4.1.3 Hydraulic retention time 41 4.1.4 Effects on the main EBPR process effluent and potential P recovery 43 4.1.5 Simulation results 47 4.2 Experimental Investigation 48 4.2.1 EBPR process performance: Period I 49 4.3 Sidestream Process 49 4.3.1 Operating conditions 49 4.3.2 Phosphate release unit 49 4.3.3 Sidestream clarifier 50 4.3.4 Characteristics of the sidestream supernatant 50 4.3.5 P0 4 -P and N H 4 - N release trend 51 4.3.6 Mass of P release to V F A consumed 55 4.4 P Removal Performance 58 4.4.1 Mass of V F A consumed to mass of PO4-P removed 59 4.4.2 P uptake and release 60 4.4.3 Simulation of Sidestream wasting ; 62 4.5 N Removal Performance 66 4.6 Effect of Mg Addition 68 4.7 P Recovery Study 70 4.7.1 P removal and recovery efficiency 71 4.7.2 N removal 73 4.8 Comparison with the Simulation Results 74 CHAPTER 5 - CONCLUSIONS AND RECOMMENDATIONS 75 5.1 Conclusions ...75 5.2 Recommendations 77 REFERENCES 79 iv APPENDIX A: ASM2-DELFT BIO-P MODEL 86 APPENDIX B: INSTRUMENTAL OPERATIONAL PARAMETERS 112 APPENDIX C: DAILY MONITORING RECORDS (IN CD-ROM) 114 APPENDIX D: MASS BALANCE EQUATIONS.. 115 v LIST OF TABLES Table 2.1. IWA family of models 8 Table 3.1. Primary effluent wastewater characteristics 23 Table 3.2. Wasting calculations 28 Table 4.1. Experimental periods with sidestream operating conditions 48 Table 4.2. Influent Mg/P and K/P values studying effects of M g and K on EBPR process 69 Table 4.3. Results of M A P crystallizer runs 72 Table 4.4. Struvite recovery 73 Table A . l . Stoichiometric matrix and component composition matrix 96 Table A.2a. Stoichiometric coefficients for S N H and SPO 98 Table A.2b. Stoichiometric coefficients for SHCO and X Tss 99 Table A.3. Component composition factors belonging to the composition matrix 101 Table A.4. Stoichiometric parameters 102 Table A.5a. Kinetic parameters for Hydrolysis, X H and H A • 104 Table A.5b. Kinetic parameters for X P A o 106 Table A.6. Kinetic rate equations 108 Table A.7. Changed values of parameter due to the calibration when applied to U B C pilot plant I l l Table B . l . Instrument operational parameters for the flame atomic absorption spectrophotometer 113 Table B.2. Instrument operational parameters for the LaChat QuickChem flow injection analysis instrument 113 LIST OF FIGURES Figure 2.1. Schematic of metabolism of biological phosphorus removal organisms during anaerobic and anoxic /aerobic conditions 11 Figure 2.2. Schematic of the basic EBPR process and UCT process 12 Figure 2.3. Schematic of (a) external and (b) submerged membrane module 14 Figure 2.4. Schematic of Phostrip process (Drnevich, 1979) 17 Figure 2.5. Different configuration for recovery of phosphate from phosphorus rich sidestreams 18 Figure 2.6. Sidestream phosphorus recovery as proposed by Smolders et al. (1996).. ....18 Figure 3.1. Schematic of the M E B P R and CEBPR processes as in U B C pilot plant 24 Figure 3.2. Schematic of sidestream configuration for M E B P R 26 Figure 3.3. Schematic of the U B C M A P crystallizer 32 Figure 4.1. Schematic layout of sidestream M E B P R process 37 Figure 4.2. P0 4 -P and N H 4 - N release trend in P R U 38 Figure 4.3. N H 4 - N : PO4-P ratio in the anaerobic, anoxic and aerobic supernatants 38 Figure 4.4. PO4-P and N H 4 - N release trend for combined sidestream flow 40 Figure 4.5. PO4-P release and recovery trend for anoxic sidestream .41 Figure 4.6. PO4-P release with time for different sidestream flow 42 Figure 4.7. Batch test results for P0 4 -P release for individual and combined sidestream 43 Figure 4.8a. EBPR process failure due to anaerobic sidestream 45 Figure 4.8b. EBPR process failure due to anoxic sidestream 45 Figure 4.8c. Failure due to combined sidestream (anoxic/anaerobic) 45 Figure 4.8d. Response of biomass to changes in f\ and/2 45 Figure 4.9. Effect offi on PO4-P release (anaerobic zone) and NO3" (MEBPR effluent) 46 Figure 4.10. Estimated phosphorus recovery for different sidestream flow 47 Figure 4.1 la. PO4-P and NH4-N release trend for membrane sidestream process 52 Figure 4.1 lb. PO4-P and N H 4 - N release in conventional sidestream process 53 Figure 4.12. Nitrates concentrations in anoxic zones of M E B P R and CEBPR 53 Figure 4.13. Specific phosphate release in M E B P R and CEBPR sidestreams 54 Figure 4.14. Ammonia nitrogen in anoxic zone and P R U for membrane sidestream process 54 Figure 4.15. PO4-P in anoxic zone and P R U for membrane sidestream process 55 Figure 4.16. Ratio of PO4-P released to V F A consumed 57 Figure 4.17. Total phosphorus in the influent and orthophosphate-P in the effluent for M E B P R and CEBPR processes 59 Figure 4.18. Ratio of V F A utilized to phosphate removal for M E B P R sidestream process 60 Figure 4.19. M E B P R sidestream process: Phosphate release and uptake trend 61 Figure 4.20. CEBPR sidestream process: Phosphate release and uptake trend 62 Figure 4.21. Anaerobic and sidestream phosphate release: without sidestream clarifier wasting and with sidestream wasting 63 Figure 4.22. Box plot for aerobic TP of M E B P R sidestream process 65 Figure 4.23. Phosphate release in anaerobic zone and PRU after implementing sidestream wasting 66 Figure 4.24. Total nitrogen in the influent, and ammonium and nitrite+nitrate in the M E B P R and CEBPR effluent 67 Figure 4.25. Magnesium in P R U (MEBPR sidestream), influent and effluent for Period III 70 Figure 5.1. Mixing of sidestream supernatant with anaerobic digester supernatant 78 LIST OF ABBREVIATIONS B N R biological nutrient removal BOD biochemical oxygen demand CEBPR conventional enhanced biological phosphorus removal COD chemical oxygen demand EBPR enhanced biological phosphorus removal f\ fraction of the influent taken as sidestream fi fraction of the sidestream flow taken as clarifier underflow G L Y Glycogen Hac Acetate HRT hydraulic retention time IWA International Water Association LDPE low density polyethylene M A P magnesium ammonium phosphate M B R membrane bioreactor M E B P R membrane-based enhanced biological phosphorus removal Mg Magnesium N Nitrogen NH4-N ammonia nitrogen P Phosphorus PAOs phosphorus accumulating organisms P H A polyhydroxy-alkanoate PHB poly-hydroxybuterate P04-P ortho-phosphate PP poly-phosphate P R U phosphate release unit Qi n influent flow Qss sidestream flow QSSR sidestream return flow SRT solid retention time SSR supersaturation ratio T C A tricarboxylic acid T K N total Kjeldahl nitrogen TP total phosphorus TSS total suspended solids TUD Technical University of Delft vui University of British Columbia University of Cape Town volatile fatty acid waste water treatment plants ACKNOWLEDGEMENTS I would like to acknowledge the assistance, support and encouragement that have been provided by the following people and institution, without whom, this research would not have been possible. • M y supervisors, Don Mavinic and Eric Hall, for their encouragement, understanding and unwavering support throughout the completion of this research. • Fred Koch, for his continued help, input and direction. • Alessandro Monti, for his invaluable help and guidance from day one, in making the project understandable to me. • Eman AI-Atar, for her immense help in providing the details related to model and simulation. • Paula Parkinson and Susan Harper, for their help in the laboratory. • To my family and friends for their encouragement and support. • The University of British Columbia, for giving me the opportunity to continue my studies. INTRODUCTION CHAPTER 1 - INTRODUCTION 1.1 Phosphorus Removal and Recovery Stringent water quality standard criteria are set for phosphorus and nitrogen compounds to control the eutrophication process, which is a world wide water pollution problem. In this regard, the enhanced biological phosphorus removal (EBPR) process has been recognized as one of the most economical and sustainable processes to remove phosphorus (Munch and Barr, 2001). Membranes, coupled with EBPR processes, have been getting more attention these days to meet the increasing demand to expand the capacity of existing wastewater treatment plants (WWTPs) and to meet the stringent effluent discharge guidelines. Membrane-Assisted EBPR processes offer many advantages, like operating at higher hydraulic loading rates and superior effluent quality, when compared to conventional activated sludge processes. Phosphorus recovery, along with removal, has been discussed in the past few years (Gaterell et al, 2000; Kuroda et al, 2002). This is due to the fact that we are facing the exhaustion of phosphorus minerals by the mid 21 s t century (Saktaywin et al, 2005). The contemporary focal issue is not eliminating phosphorus per se, but recycling it. Presently, there is a need for a process of phosphorus removal that separates it from other waste components, so that it can be recycled as a fertilizer or an ingredient in other valuable phosphorus products. The emerging technologies of struvite crystallization (magnesium-ammonium phosphate) and hydroxyapatite precipitation (calcium phosphate), to recover phosphorus may serve as catalysts for removing phosphorus as a recyclable product. A promising way is to combine the EBPR process with struvite and hydroxyapatite crystallization. This will save chemicals for phosphorus precipitation, reduce the size of the treatment facility, reduce the volume of effluent to be treated (more concentrated phosphorus in present effluent), and consequently reduce overall costs (Stratful et al, 1999; Jeanmaire and Evans, 2001). 1 INTRODUCTION The EBPR process accumulates more phosphorus in the biomass compared to a conventional activated sludge process, in the form of polyphosphate. The wasted sludge, when hydrolyzed anoxically or anaerobically, releases the polyphosphate, along with potassium and magnesium, to the bulk solution. This supernatant could be processed using a fluidized bed reactor to recover phosphorus as struvite (magnesium ammonium phosphate or MAP) . Struvite is known to display excellent slow release fertilizer properties (Doyle and Parsons, 2002). In recent years, fluidized bed reactors have been commonly used to recover P in the form of struvite from anaerobic digester supernatant (Munch and Barr, 2001). Another important reason to recover phosphorus is the operational problems encountered in wastewater treatment plants. The pumping and maintenance cost is increased due to the deposition of struvite in the piping and equipment of sludge treatment processes, when the diameter of the pipe is reduced. Also, the time taken for the sludge to be moved from one place to another increases, reducing the plant capacity (Britton et al., 2005; Doyle and Parsons, 2002). 1.2 Sidestream Process To achieve simultaneous phosphorus removal and recovery in the present study, a membrane-based EBPR process was integrated with a M A P crystallizer, using a sidestream process. It is important to note that there has been no research carried out previously, studying the coupling of membrane based EBPR and M A P crystallizer processes. This work is a follow up on the comparative study done by Monti (2006) between membrane and conventional EBPR processes in the Department of Civi l Engineering, University of British Columbia (UBC). This innovative and sustainable sidestream process, which introduces membrane technology and phosphorus recovery in the biological nutrient removal (BNR) process, is expected to change the way in which BNR systems are designed and operated. In general, the sidestream method to recover phosphorus reported in the literature refers to the return sludge line, which contains a phosphorus release unit followed by phosphorus precipitation chemicals (e.g. Phostrip process) for recovery in the form of calcium phosphate or struvite. The precipitation methods are incompatible with technologies currently used in 2 INTRODUCTION the phosphate industry. They either require excessive energy input, to separate the phosphates from the added precipitation chemicals, or interfere with the industrial process. Struvite formation, as a route to phosphorus recovery, looks very promising and several struvite recovery processes are already in operation (e.g. D H V Crystalactor, Netherlands; Unitika Phosnix process, Japan). Another alternative to recover phosphorus is to take a sidestream flow, either from the anaerobic, anoxic or aerobic zones from an EBPR process, to produce a phosphorus-rich stream. For sidestream processes, it becomes important to evaluate acetate requirements, since the phosphate release in the sidestream is dependent on acetate availability. Since there exists no information on this type of sidestream operation with the coupling of a membrane based EBPR and a M A P crystallizer, it was decided to mathematically simulate the sidestream operation to gain a better understanding of the process, before implementing it at the pilot scale level. This integration, which offers an important advantage of achieving simultaneous phosphorus recovery and removal, has not been investigated. Therefore, the main question to be answered here is i f the addition of a sidestream process to existing biological nutrient removal plants, allows for a significant amount of phosphorus recovery; also, what would be the impact of the sidestream operation on the main process? 1.3 Research Objectives The coupling of the membrane-based EBPR process with the M A P crystallizer unit needs to be studied, to answer several design questions and to assess the impact on the main process performance. The source of the sidestream to obtain the sludge, the fraction of the inflow to be taken as sidestream and retention time of the phosphorus release unit are some of the key factors to be determined, before implementing the process at the pilot plant. 1. Simulation of the sidestream operation is intended to understand the following : • origin of the sidestream, • hydraulic retention time in the phosphate release unit (PRU), • percentage of the sidestream to be taken, • PO4-P and NH4 -N concentration in the PRU, and 3 INTRODUCTION • possible different combinations to access the sidestream. 2. Based on the simulation results, it was intended to demonstrate: • implementation of sidestream process for both membrane-based EBPR and conventional EBPR processes, at pilot scale, and • recovery of phosphorus as struvite using the M A P crystallizer. 3. To study the impact of sidestream process on: • phosphorus and nitrogen removal, and • acetate requirements. 4 L I T E RAT URE REVIEW CHAPTER 2 - LITERATURE REVIEW This literature review is presented in three main parts: membrane-based EBPR processes, sidestream processes, with phosphorus recovery and simulation, using the Delft bio-P model. Section 2.1 reviews the development of wastewater treatment models and its features. An overview of the Delft bio-P model is also presented. Section 2.2 reviews the development and principles of the EBPR process, along with the membrane bioreactors. Section 2.3 reviews the existing types of sidestream configurations for phosphorus recovery reported in the literature, along with the parameters affecting struvite formation in a M A P crystallizer. 2.1 Modeling and Simulation of Wastewater Treatment Systems There have been several models developed over the years. They can be broadly classified as simple and advanced models. Simple models are based on mainly four parameters, viz-a-viz soluble substrate, heterotrophic biomass and the incorporation of mainly the growth and decay operations. Some of the simplified models include Lawrence and McCarty (1970), Eckenfelder (1966), Goodman and Englande (1974), Gaudy and Kincannon (1977), and Chen and Hashimoto (1980) (Eckenfelder, 2000). The Lawrence and McCarty (1970) model has been widely recognized and has been the foundation for many design equations and procedures. Advanced models are those which are developed more recently and which include more components. Generally, simplified models incorporate only aerobic biochemical oxygen demand (BOD) or chemical oxygen demand (COD) removal, while advanced models add aerobic COD removal, nitrification, anoxic denitrification, and phosphorus removal. The models developed by the research group at University of Cape Town (UCT), South Africa have progressively included aerobic COD removal and nitrification, anoxic denitrification, and anaerobic-anoxic-aerobic biological excess phosphorus removal (Wentzel and Ekama, 1997). Some of these models have provided the basis for most of the advanced models developed later. 6 L I T E R A T U R E REVIEW The best known advanced models are the family of models developed by the International Water Association (IWA), known as Activated Sludge Models (ASM). ASM1 is considered as the reference model, since this model triggered the general acceptance of waste water treatment plants (WWTPs) modeling by the research community and later by industrial groups. It has become a reference for many scientific and practical projects and also has been implemented in simulation software. To date, IWA has developed A S M 1 , ASM2, ASM2d and ASM3 models, where ASM3 is the most advanced version. Other models like Wentzel et al. (1989b), a general model for biological nutrient removal (Barker and Dold, 1997), the E A W A G bio-P module for ASM3 (Rieger et al, 2001) are well recognized. A l l these models include the bio-P process. The Wentzel model was the basis for ASM2 and ASM2d to model biological phosphorus removal. Barker and Dold (1997) merged the ASM1 and Wentzel models, with a number of modifications. Rieger et al. (2001) proposed some modifications to the phosphorus module of ASM2d, with the developments suggested in ASM3. A l l these models were based on a black box description of EBPR organic storage, as only one lumped organic storage compound is modeled and stoichiometric coefficients of the process are empirically derived. The mathematical model developed by the Technical University of Delft (TUD), (Smolders et al, 1995c and Murnleitner et al, 1997), named the TUDP model and revised metabolic model proposed by Filipe and Daigger (1998), are also well received. Table 2.1 summarizes the important features of A S M models and the Delft bio-P model is presented in the next section. 2.1.1 ASM2- Delft bio-P model This metabolic model was introduced by the environmental engineering group from TUD. A l l storage components are modeled explicitly and a full accounting of the metabolism of phosphorus accumulating organisms (PAOs) is made. The fundamental biochemical principles used in the model are the ones proposed by the Mino model (see Section 2.2.2). A l l of the processes that generate and use energy are modeled, based on the major biochemical pathways. It is expected that process stoichiometry calculated using such fundamental principles, will be applicable to a broader set of conditions, than empirically-derived stoichiometry. In the metabolic model, eight cell-internal processes are described, involving three storage polymers: polyhydroxy-alkanoate (PHA), glycogen (GLY) and poly-phosphate (PP). The processes are anaerobic PHA storage, anaerobic PP degradation, aerobic 7 L I T E R A T U R E REVIEW and anoxic PP and G L Y formation, and aerobic and anoxic P H A degradation. The yield coefficients of these processes are determined by two metabolic yield coefficients (Smolders et al, 1994a, b). The concept of endogenous respiration or maintenance is used (van Loosdrecht and Henze, 1999). The formation of biomass (PAOs) is calculated as the net result of conversion of PHA, G L Y and PP. Table 2.1. IWA family of models. ASM1 - Only Organic carbon and N compound removal - Flow of COD is rather complex and the death regeneration cycle of the heterotrophs and the decay process of nitrifiers are strongly interrelated - Based entirely on COD for all particulate organic material, as well as the total concentration of the activated sludge ASM2 - Includes bio-P process & chemical phosphorus precipitation - Biomass has cell internal structure - Include polyphosphates, which are responsible for phosphorus uptake/release and which does not exert any COD ASM2d -It 's a minor extension of ASM2. - It includes two additional processes to account for the fact that PAOs can use cell internal organic storage products for denitrification. - Includes anoxic growth of PAOs. - Includes denitrifying PAOs. ASM3 - Includes microbiological transformation processes - Chemical precipitation process can be easily added - Conversion processes of the two groups of organisms are clearly separated and decay processes are described with identical models. - Assumption that all substrate passes storage before being metabolized in the heterotrophic microorganisms Filipe and Daigger (1998) tried to apply the TUDP model to a completely independent set of data generated by an EBPR process run by Wentzel et al. (1989a), in a continuous flow mode. The model "under-predicted" the phosphorus removal capability of the system and a new set of kinetics was proposed to overcome this problem. Regardless of the criticisms raised by Filipe and Daigger (1998), the TUD model was tested on three full-scale WWTPs (Van 8 L I T E R A T U R E REVIEW Veldhuizen et al, 1999; Wichern et al, 1999 and Brdjanovic et al, 2000). From these tests, it was concluded that: a better calibration procedure is required, along with a method to handle and evaluate large quantities of data used for simulation purposes. Meijer (2004) carried out a model evaluation and concluded that the stoichiometric basis of the metabolic bio-P model is reliable. Meijer also developed a step-wise calibration method that strongly relied on the mass balances calculated in the data evaluation procedure. It was tested on two, full-scale wastewater treatment processes and the method was shown to be quick and straight forward. 2.2 Principle of the EBPR Process The biological mechanism of removing phosphorus and nitrogen from wastewater is generally referred to as the biological nutrient removal process (BNR). The first evidence of appearance of phosphorus accumulating organisms, which are responsible for this process, in an anaerobic reactor, dates back to 1959 (Srinath et al., 1959). In the literature, there are two important models which explain the principle behind an EBPR process. 2.2.1 Comeau-Wentzel model Comeau et al. (1986) and Wentzel et al. (1986) established that, under anaerobic conditions, i f the activated sludge is mixed with the influent wastewater, then microorganisms taking up carbon sources from the influent wastewater would dominate in the reactor. PAOs are capable of growing under such conditions, in which they take up carbon and store it in the form of polyhydroxyalkanoates (PHA); this is accompanied by the cleavage of poly-P and, consequently, release of orthophosphate. Next, in the subsequent aerobic phase, PAOs can grow aerobically and take up orthophosphate to replenish the poly-P content by using the stored P H A as the carbon and energy source. The idea that the tricarboxylic acid (TCA) cycle functions under anaerobic conditions, to oxidize a part of acetate (carbon source) to C O 2 and to generate reducing power in the form of N A D H , was the key point of the Comeau-Wentzel model. This idea was put forward because PHA is a more reduced compound than acetate and thus the conversion of acetate to P H A requires reducing power. 9 L I T E R A T U R E REVIEW 2.2.2 Mino model Mino et al. (1987) proposed an alternative model, in which anaerobic degradation of intracellularly stored glycogen to acetyl-CoA, as well as its partial oxidation to C O 2 , was hypothesized to generate reducing power for P H A formation. Other studies by Arun et al. (1988), Smolders et al. (1994a), Pereira et al. (1996) and Maurer et al. (1997) supported the Mino model. Pereira et al. (1996), who found through his experiments that the T C A cycle under anaerobic conditions supplies a part of the reducing power for PHA formation, proposed a combination of the two models, to explain the results obtained by in-vivo nuclear magnetic resonance (NMR). The schematic of biological phosphorus removal, during anaerobic and subsequent aerobic/anoxic conditions, is given in Figure 2.1. 2.2.3 Process engineering A number of different design solutions have been proposed to achieve biological phosphorus removal, often coupled also with nitrogen elimination from wastewaters (Tchobanoglous et al., 2003). They are operated as full biological processes, as well as combined biological/chemical processes. A l l of them feature an anaerobic compartment favoring the growth of PAOs, which can store more polyphosphate in their cell than any other microorganisms. When the sludge enters the aerobic zone, PAOs grow on the stored substrate while other heterotrophic organisms have no substrate to survive. Thus, the anaerobic zone in an EBPR process has to be designed based on the influent wastewater characteristics, the most important one being the V F A content that contributes to the sizing of the anaerobic zone. The aerobic zone length depends on the nitrification because of the slower growth rate of nitrifying bacteria. Figure 2.2 illustrates the basic schematic of a nutrient removal process. To couple nitrogen removal and EBPR in a single process, a denitrification zone has to be added to the treatment line. 10 L I T E R A T U R E REVIEW orN2 , Figure 2.1. Schematic of metabolism of biological phosphorus removal organisms during anaerobic (above) and anoxic /aerobic conditions (below). 11 L I T E R A T U R E REVIEW It is traditionally assumed that the two processes compete for the same, substrate, supported by the evidence that the presence of nitrate in the anaerobic zone gradually deteriorates the biological phosphate removal process. In order to prevent the presence of nitrate in the anaerobic tank, researchers in South Africa pioneered a modified process layout, to first introduce nitrate containing return sludge to a denitrification reactor, after which the nitrate free sludge/water mixture is partly recycled to the anaerobic tank. This design approach was named the University of Cape Town (UCT) process (Fig. 2.2). Assuming a complete absence of nitrate in the anaerobic reactor, denitrification can be achieved through two processes. (1) heterotrophic denitrification in the anoxic zone using leftover carbon as an electron donor and nitrate (coming with the aerobic and returned sludge recycle) as the electron acceptor; (2) denitrifying dephosphatation, an innovative process carried out by a sub-group of PAOs that can use nitrate as the electron acceptor in the absence of oxygen. The latter process has been an active research area in recent years, due to its strong potential to reduce the use of COD for nitrogen removal and save oxygen requirements in the subsequent aeration zone (Bortone etal, 1996 and Kuba et al, 1996). Influent Anaerobic tank Aerobic tank Settling tank Effluent Influent Return sludge Excess Sludge • Anoxic recycle Anaerobic tank Anoxic tank Aerobic tank Settling tank Effluent Aerobic recycle 1 Excess Sludge Sludge recycle flow Figure 2.2. Schematic of the basic EBPR process (above) and UCT process (below). 12 L I T E R A T U R E REVIEW Chemical precipitation to remove phosphorus from the wastewater is often applied in conjunction with biological means, to further improve the quality of the final effluent. Generally, chemicals are added to the main sludge line with the disadvantage of accumulating precipitates in the bioprocess. As an alternative, a fraction of the return sludge is introduced to a stripper tank, in which anaerobic conditions are maintained and treated downstream with chemical precipitation (PhoStip process); or, more recently, with more sustainable methods such as crystallization, resulting in phosphorus recovery. Reducing the process SRT is claimed to be another method to improve the efficiency of EBPR systems. This results in an increase of net biomass yield and, as a consequence, more phosphorus-rich sludge to be wasted. 2.2.4 Membrane bioreactors There are two types of membrane applications in microbiological processes; the first one is a solids-liquid separator, which replaces the conventional clarifier and secondly, membranes are used for bubbleless gas transfer devices, to supply gaseous substrates (Rittmann, 1998). The first type of application is widely used in wastewater treatment systems and, in recent years, has grown to become a serious alternative to conventional treatment. Membrane bioreactors can be defined as the combination of two basic processes, biological degradation and membrane separation, into a single process in which suspended solids and microorganisms responsible for biodegradation are separated from the treated water by a membrane filtration unit (Manem and Sanderson, 1996). The configurations which are used in wastewater treatments are represented in Figure 2.3. Fig 2.3a is the type where activated sludge is pumped to tubular or flat sheet modules where it flows at high velocities, with consequently quite a high pressure drop and thus quite a high transmembrane pressure. This is known as a typical cross flow filtration. Fig 2.3b represent membrane bioreactor (MBR) with immersed membranes, where hollow fibers are immersed in the aerated tank and treated water is generated by applying suction to the inner part of the fibers. 13 LITERATURE REVIEW (a) (b) Influent Sludge recirculation Influent Effluent Bioreactor Effluent \ Membrane Bioreactor Figure 2.3. Schematic of (a) external and (b) submerged membrane module. The cross-flow filtration requires higher pressure, due to the pressure drop along the membrane length with a high tangential velocity. This results in a large recycle ratio and thus increases energy consumption. The immersed membrane works at smaller applied pressure, without recirculation, but requires aeration for membrane scouring. Thus, the immersed membranes have found more applications around the world. 2.2.5 Advantages and disadvantages of membrane bioreactors The advantages of using a membrane bioreactor are as follows. Superior quality of the effluent with very high removal of pathogens. Greater reliability and flexibility in use, as hydraulic and solids retention times can be separated completely, providing optimum control of the biological reaction. - The selection of bacteria and microorganisms is no longer dependent on their ability to form biological floes and thus to settle, which means that all the species present in the system will have the same residence time (Ben Aim and Semmens, 2003). - MBRs can operate at higher solids retention times as sludge settling characteristics no longer are important (Yamamoto et al, 1989, Yasui et al, 1996, Sakai et al, 1997). - MBRs are compact systems. The main disadvantage is the high capital and operating cost of membrane bioreactors, but with the introduction of submerged membrane bioreactors and reduction in membrane production cost, the M B R process stands a better chance of adoption when compared to 14 L I T E R A T U R E REVIEW conventional activated sludge systems. The other major disadvantage is that membranes are prone to fouling, a process that gradually reduces the permeate production rate and thus requires the cleaning of membranes, periodically. 2.2.6 MBR coupled with EBPR process The coupling of M B R and EBPR is a lucrative area, given the fact that the effluent produced will have only soluble phosphorus. Also, in the M B R process alone, phosphorus removal is carried out through a chemical process. There are only a few studies that have focused on M B R and EBPR processes. Recently, Adam et al (2002), Patel et al (2005) and Fleisher et al (2005) studied the EBPR coupled with M B R systems. Traditionally, M B R technology has been employed for carbon and nitrogen removal biological processes (Cote et al, 1998; van der Roest et al, 2002; Rosenberger et al, 2002). One of the most recent works of coupling M B R and EBPR comes from the research group at the University of British Columbia. Monti (2006) carried out a comparative study of M E B P R and CEBPR processes. The main observations of this study were that the membrane-assisted process maintained satisfactory nitrification and phosphorus removal performances under high rate conditions, even at an hydraulic retention time of five hours. The process performance was not compromised even at higher solids retention time (20 days). 2.2.7 Problems with biological nutrient removal processes BNR processes face problems with the disposal of excess sludge, which are phosphorus rich. When further digested, especially anaerobically, most of the phosphate is re-released, which must then be removed again from the wastewater in the aerobic/anoxic zones. Various studies show that from 26 % to 90 % of the phosphorus entering the head of the treatment plant is due to this release i.e. phosphorus in the return liquors (Mavinic et al, 1998; Jardin and Popel, 1994). This means that phosphorus is not removed from the system but just circulated in a loop. Also, a number of treatment plants have reported the occurrence of unintentional struvite formation in plant piping and other equipment, e.g. pumps, valves, filter belts etc. (Snoeyink and Jenkins, 1980; Ohlinger et al, 1998). The precipitation occurs when the combined concentrations of M g + , N H 4 + and P0 4 " 3 exceed the struvite solubility limit. This problem is more severe in anaerobic digestion systems. These deposits can damage the pumping equipment and block the sludge pipes resulting in costly maintenance 15 L I T E R A T U R E REVIEW and repairs (Ohlinger et al., 1998). Several solutions like the installation of water softening devices, precipitating phosphorus by the addition of ferric chloride, dilution of digester sludge with secondary effluent, adding scale inhibitors, acidifying the waste stream and redesigning the plants, have been taken or suggested to overcome the problem (Borgerding, 1972; Williams, 1999). However, all these measures are costly and, at times, the problem is not completely eliminated. 2.3 Sidestream Process for Phosphorus Recovery There was little information found in the literature about the introduction of membrane technology and phosphorus recovery in BNR process. However, there exist different variants of sidestream processes to recover phosphorus from wastewater treatment plants, in which phosphorus is recovered by treating part of the return sludge under anaerobic conditions in a stripper. This method of phosphorus recovery, using a phosphorus-rich sidestream, is most common in Europe. The recovery of phosphorus is very much compatible with biological phosphorus removal due to the 'problems' associated with it, (as discussed before). So, in order to recover phosphorus, one does not need to go out of the way in terms of installation of a new facility. Recovery of phosphorus would not only be useful for phosphorus recycling, but also will help in producing less sludge in biological nutrient removal plants (Woods et al, 1999). In the following paragraphs, an overview of the existing different types of sidestream processes is presented. To achieve a sidestream with high phosphate concentration, it is necessary that the process be operated with biological phosphorus removal. Several reviews on biological phosphorus removal (Arvin, 1985; Balmer and Hultman, 1988; Morse et al, 1998) discuss sidestream technology. The most well known e xample is the PhoStrip process, which is illustrated in Fig 2.4. Since phosphorus recovery was not a significant objective at that time, the lime-precipitated sludge was recycled to the pre-sedimentation basin. 16 L I T E R A T U R E REVIEW Pre-Sedimentation Influent wastewater^ Secondary Sedimentation Effluent Clear Water Excess_Sludge Part of return Sludge Figure 2.4. Schematic of Phostrip process (Drnevich, 1979). The main purpose of using a sidestream in this case was that considerably less lime (ca 10-20%) would be used, compared to what was needed for precipitation in the main stream. Instead of recycling precipitated calcium phosphate to the pre-sedimentation chamber, calcium phosphate can be extracted as a product. The calcium phosphate produced by the Dutch wastewater treatment plant Geestmerambacht is used by the phosphate industry (Gaastra et al, 1998). Similar sidestream processes have been constructed in Darmstadt in Germany (Hillenbrandt et al, 1999) and Haarlem Waarderpolder in Holland (Brdjanovic et al, 2000). Klapwijk et al, (2001) calculated a 60 % phosphate recovery in the PhoStrip process and a sidestream process. Figure 2.5 shows these two processes along with Renphos process (Rensink et al, 1997), where the phosphorus-poor sludge is taken out from the process as excess sludge. The acetate requirement has to be evaluated carefully, since acetate is required to release the phosphate from the sludge. Smolders et al, (1996) made a steady state analysis and showed that acetate requirements for a sidestream process were much lower than the requirements needed in a mainstream process. The configuration is showed in Fig 2.6. For this kind of configuration, since there are is no influent acetate used, acetate COD has to be added to the sidestream. 17 L I T E R A T U R E REVIEW Separation •Excess P Precipitation ! Anaerobic Anoxic Aerobic \ zone Zone — • Zone Exc Sludge -•Recovered phosphate Phosphorus recovery and excess sludge from a sidestream Return Sludge P Precipitation Separation Stripper Recovered phosphate PhoStrip ! Activated Sludge ! process with I biological Effluent[ phosphorus ! removal P poor sludge r RenPho Phosphorus recovery from I activated sludge PhoStrip- with PhoStrip or Excess Sludge RenPho Excess l Figure 2.5. Different configuration for recovery of phosphate from phosphorus rich sidestreams. Influent Effluent Excess Sludge Acetate Figure 2.6. Sidestream phosphorus recovery as proposed by Smolders et al. (1996) As indicated earlier, to date, phosphorus has been recovered from wastewater as calcium phosphate, as struvite (magnesium ammonium phosphate), by precipitation as aluminum and iron phosphates and with ion exchange technologies. Phosphorus recovery in the form of calcium phosphate produces the same chemical as that in mined phosphate ore and which be easily recycled in the phosphate industries. The D H V crystalactor™ is the best known example to recover phosphorus in this form. The only example which recovers P as A I P O 4 is by Thermphos International, located in Vlissigen, Holland. It can be recycled in the P 18 L I T E R A T U R E REVIEW industry using the thermal route. The iron phosphate form cannot be used in the existing P industry processes and probably has low or zero fertilizer value (Jeanmaire, 2001). The REM-NUT process is an example where ion exchange technology is employed, followed by struvite precipitation. 2.3.1 MAP crystallizer In the present study, a previously-developed, pilot scale M A P crystallizer was used. In the literature, crystallization reactor designs have ranged from simple glass beakers at the bench scale, to more complicated reactor designs, such as fluidized beds and agitated crystallizers. Using the U B C M A P crystallizer reactor, up to 91% recovery of removed phosphorus has been achieved in the past (Britton, 2002). 2.3.2 Supplementations required Struvite forms in a theoretical Mg:N:P molar ratio of 1:1:1. Supplementation of magnesium from an external source, would be necessary, since fairly low levels of this cation would exist in the supernatant produced to fulfill stoichiometric requirements of the maximum possible struvite formation. External magnesium sources previously investigated include magnesium chloride (Giesen, 1999; Britton et al., 2004), magnesium oxide (Schuiling and Andrade, 1999; Munch and Barr, 2001) and seawater (Giesen, 1999). Since struvite formation is greatly facilitated under alkaline conditions (pH = 8 - 1 0 ) , the addition of base is also required (Momberg and Oellermann, 1992). Adding magnesium hydroxide is known to increase the pH (Munch and Barr, 2001) which favors struvite formation. This is because struvite is highly insoluble in alkaline solution. This also helps to maintain saturated conditions, which is required to run the crystallization process. The other source of base is sodium hydroxide (Ohlinger et al, 2000; Seckler et al, 1996). 2.3.3 Parameters affecting struvite formation Some of the important parameters affecting struvite formation are outlined below. a) Mg:P ratio: Various studies have shown that an increase in Mg:P molar ratio increased the P-removal efficiency. The optimized Mg: P molar ratio was about 1.3:1 (Yaffer et al, 2002; Ohlinger et al, 2000). 19 L I T E R A T U R E REVIEW b) Turbulence: This is a very important parameter in struvite crystallization. Nucleation is controlled by solubility chemistry, while growth rate is believed to be limited by low turbulence or low mixing energy (Adnan, 2002). The lack of a universal quantification for turbulence poses a big problem. The Reynolds number has been (Dastur, 2001) used to measure degree of turbulence; however, due to the fact that the Reynolds number would change once the reactor would start to fill up with growing crystals, its usefulness is limited. c) Recycle ratio: The most important function of recycle is to dilute the strong wastes so that the process of crystallization remains in the metastable zone. It also helps in achieving a desired up flow velocity inside the reactor. d) Seeding: Seeding the reactor is of vital importance since it provides particles on which the deposition of struvite can occur. It avoids the lag period for nucleation to occur and growth can proceed quickly (Dastur, 2001). Quartz, phosphate rock, bone charcoal and struvite have all been used as seeding materials (Momberg and Oellermann, 1992). It has been also suggested that seeding is required at the start up and the ongoing process eventually becomes self-seeding (Much and Barr, 2001). e) Temperature: Burns and Finlayson (1982) reported that struvite is less soluble at 25° C and more soluble at 38° C. One more study by Borgerding (1972) indicated that, as the temperature increases from 0 to 20° C, the solubility increases to a maximum after which struvite solubility decreases with increasing temperature. At U B C , Adnan (2002) found that struvite was less soluble at 10 than at 20° C. 2.3.4 Solubility product of struvite Extensive studies have been conducted for calculating the solubility product (Ksp) value for struvite (Ohlinger et al, 1998; Burns and Finlayson, 1982). Reported values for - log Ksp range from 12.6 to 13.8 (Dastur, 2001). The dispersed Ksp values for struvite could be due to a number of factors. For example, since solubility products may be derived by using approximate solution equilibria, the effects of ionic strength are often neglected. Also, mass balance and electro neutrality equations are not always used and different chemicals species are selected for the calculations (Dastur, 2001). In order to overcome the complexities associated with the calculation of a solubility product for struvite, a simple, well-defined concept of conditional solubility is used for practical purposes. These conditional solubility 20 L I T E R A T U R E REVIEW constants, are equilibrium constants that are true for given experimental conditions and are defined by (Dastur, 2001), P s = [Mg + 2] t o t a , .[NH 4-N] t o t a l .[P04-P]totai = K s p ...(2.1) a M g + 2 a N H 4 + a P0 4 " 3 y M g + 2 y N H 4 + y P0 4 " 3 Where, P s is conditional solubility product, a = ionization fraction of the respective components, and Y = activity coefficient for respective ion species. 2.3.5 Supersaturation ratio In order to quantify struvite precipitation potential, a term supersaturation ratio (SSR) is used (Snoeyink and Jenkins, 1980). It is given by, SSR = P s /P s . e q (2.2) Where, Ps-eq = conditional solubility product at equilibrium, and Ps = conditional solubility product analysis given by the measured molar concentrations of total dissolved magnesium, ammonia and orthophosphate species. Theoretically, SSR >1 implies that precipitation is possible, SSR = 1 indicates that the system is at equilibrium and SSR < 1 indicates that precipitation is not possible and the system is undersaturated. 21 M A T E R I A L S A N D METHODS CHAPTER 3 - MATERIALS AND METHODS 3.1 Process description 3.1.1 Site description The pilot plant at the University of British Columbia is located on the south campus and has been in operation since 1981. It consists of membrane-based enhanced biological phosphorus removal process (MEBPR) and conventional enhanced biological phosphorus removal process (CEBPR) treatment trains, which are based on the University of Cape Town (UCT) model. The north-south trunk sewer line, which runs to South West Marine Drive, serves as the source for wastewater to the pilot plant. During the study, this trunk collected wastewater mainly from the Acadia Park and Hampton Place housing complexes. The wastewater was pumped into two storage tanks with a capacity of 15,000 liters at 3 am, 9 am, 3 pm and 9 pm, daily. 3.1.2 EBPR configuration Wastewater from the two storage tanks is fed to a primary clarifier. Primary effluent is collected in a 20 liter holding tank, from which it is fed continuously to both of the treatment trains. A l l of the sampling for influent was collected from this primary effluent holding tank. The general characteristics of the influent wastewater are presented in Table 3.1. The process consisted of three zones, an anaerobic zone, which is followed by a anoxic zone and then the aerobic zone (Tchobanoglous, et al, 2003). To stimulate the growth of phosphorus accumulating organisms, a recycle line returned biomass from the anoxic to the anaerobic zone. In order to achieve denitrification, a recycle line carrying nitrates was provided from the aerobic zone to the anoxic zone. In the case of conventional EBPR processes, a recycle line from the secondary clarifier to the anoxic zone is provided to minimize the effect of nitrates entering the anaerobic zone. A l l of the recycle flows were set to the influent flow rate, which was 5.4 m 3/d. The effluent from the secondary clarifier was collected in a holding tank, which was kept in a complete mix condition. The operating conditions for both processes 22 M A T E R I A L S A N D METHODS were identical. The hydraulic retention time (HRT) of 10 hours and solids retention time (SRT) of 20 days was maintained throughout the study. Figure 3.1 shows the schematic of the membrane and conventional EBPR processes, as set in the U B C pilot plant. Both trains had the same total reactor volume of 2228 liters and the volume fraction of each zone was 0.11, 0.28 and 0.61 for the anaerobic, anoxic and aerobic reactors, respectively. Both trains received the same primary influent. The only difference between the two treatment lines was in the method of separating the solids from the liquid. A secondary clarifier was used in the case of the conventional EBPR process and a membrane module replaced the clarifier in the case of the membrane-based EBPR process. Acetate (5000 mg/L) flow was maintained to the anaerobic zones of each process, to enhance the phosphorus release. This was required due to the insufficient carbon present in the influent wastewater. Table 3.1. Primary effluent wastewater characteristics. Parameter Mean Min-Max TSS (mg/L) 90.1 16-180 COD t o t (mg/L) 291 122 - 590 COD s 6 , (mg/L) 186 80 - 467 Acetate (mg/L) 18.2 0.0 - 43.4 Propionate (mg/L) 5.2 0.0 - 24,4 Tot V F A ' s (mg COD/L) 19.00 0.0-76.57 T K N (mg N/L) 35.9 25.8-47.3 N H 4 - N (mg N/L) 26.7 9.1-39.2 N 0 3 - N (mg N/L) Not detec. TP (mg/L) 4.2 2.1-7.9 P O 4 - P (mg/L) 3.0 1.1-6.7 Mg 1.2 T°C 20.2 14.0-24.0 pH 7.2 6.4-7.8 23 M A T E R I A L S A N D METHODS Permeate Tank 00 a Membrane module MIXER Aerobic Anoxic MIXER Anoxic Fine bubble aerator o Figure 3.1. Schematic of the MEBPR (above) and CEBPR (below) processes as in UBC pilot plant. 24 M A T E R I A L S A N D METHODS 3.1.3 Membrane Two coupled membrane modules were used for this study, which were directly immersed in the aeration zone. The surface area of each module was 12 m 2 and the membranes had a nominal pore size of 0.04 /im. The hollow fiber membrane was provided by Zenon Environmental Inc. The membranes were operated in permeation mode for 9.5 minutes and back flush mode for 0.5 minute. Permeate flow was set slightly higher than 5.4 m3/d, to accommodate this rapid back flush. To maintain the constant water level in the reactor at all the times, permeate was also recycled back to the aerobic zone, as required. The air flow was maintained at 0.34 m3/min, with intermittent coarse-bubble aeration set for 20 seconds on and 10 seconds off, provided to the module. This air flow aids in reducing the cake formation on the membrane surface and in between the membrane fibers. The amount of air provided to agitate the membranes was also sufficient to satisfy the biochemical oxygen demand, while keeping the aerobic zone dissolved oxygen (DO) level between 2.5 and 3.5 mg/L. DO was measured using a VMR® meter that was calibrated at least twice a week. When the transmembrane pressure exceeded the recommended value of 65 kPa , the module was replaced by a module that had been previously restored through soaking in solutions of concentrated citric acid and sodium hypochlorite. 3.1.4 Sidestream configuration The sidestream set up (Fig. 3.2) was identical for both the membrane and conventional EBPR processes. The sidestream (Qss) from the EBPR process entered the phosphorus release unit (PRU), whose volume was 120 liters. The P R U was well mixed with the aid of a mechanical mixer, so that the acetic acid added was distributed throughout the unit. The outlet was connected to a clarifier of volume 57 liters. The strength of the, acetic acid was the same as that added to the anaerobic zone. The acetate flow was maintained at 0.015 mL/min. This was arrived by a trial and error method, ensuring that sufficient amount of V F A was left over (in the range 40 - 50 mg/L) in the supernatant produced. This was required to confirm that V F A was not limiting and that maximum phosphorus release was obtained. Supernatants from both sidestream clarifiers were collected in a single collection tank, and this formed the feed to run the crystallizer. Underflow from the clarifier was returned to the anoxic zone 25 M A T E R I A L S A N D METHODS (QSSR), at a rate of 20 % of the incoming sidestream flow. The dotted line in the Figure 3.2 represents the multiple zone sidestream scenario. PRU Clarifier Qss Anaerobic Zone Sidestream Supernatant Qssr Struvite AeroBicTRecycle Flow Anoxic Zone Anoxic Recycle Flow MAP Crystallizer Membrane Module Aerobic Zone •+ Effluent Sidestream process MEPBR process Figure 3.2. Schematic of sidestream configuration for M E B P R 3.2 Model and simulation software The ASM2-Delft metabolic bio-P model was implemented and calibrated on A Q U A S I M platform (simulation software) previously for both M E B P R and CEPBR processes of the U B C pilot plant (Al-Atar, 2006). It was calibrated using the protocol suggested by Meijer et al, (2001) and later reemphasized by Roeleveld and van Loosdrecht (2002) and Hulsbeek et al, (2002) ( Eman AI-Atar, Dept. of Chemical and Biological Engineering, U B C , Vancouver, B.C., pers. comm..). A l l stoichiometric parameters and kinetic coefficients used in the simulation are presented in Appendix A (Meijer, 2004). The simulation software used was the second version of A Q U A S I M (Richeirt, 1999). The M E B P R configuration applied in A Q U A S I M was extended to include a sidestream process which consisted of P R U and clarifier as shown in Fig 3.2 (also see Section 4.1.1). A l l simulations were calculated over a period of 600 days, to attain steady state conditions. The pilot plant was modeled according to the schematic given in Figure 3.2. A l l zones were constructed as completely mixed reactors. In the sidestream clarifier, particulate matter was separated from the liquid. Under steady state conditions, the portion of the flow going to the sidestream (Qss) was the only variable parameter in this study. 26 M A T E R I A L S A N D METHODS The sidestream flow was calculated as the percentage of the inflow (Qin). QsS=/l-Qin •• • (3.1) Where f\ is the fraction of influent to be taken as sidestream and which varied from 0.01 to 1. QSSR is the flow going back to the anoxic zone from the sidestream clarifier and is given as follow. QSSR =fi Qss ..(3.2) Where is the fraction of QSs to be taken as underflow and was set to 0.2 through out this study. 3.3 Wasting method For the present study, to control the SRT of the process, sludge wasting by three different means was considered. The first one was the traditional method of wasting from the aerobic zone. The second one was foam wasting from the anoxic zone for the membrane-based EBPR process. Foam production in the anoxic zone, is an inherent characteristic of a membrane-based EBPR process. It has to be dealt with on a daily basis; i f not, it leads to a thick layer of foam on the surface of the anoxic zone, causing an odor problem. There were two options to deal with the foam. One was to re-suspend the foam back into the anoxic zone and, secondly, to skim off the foam manually on a daily basis. The latter choice was adopted, due to two reasons. Resuspension of foam would have required lot of mixing energy (especially, in case of a full scale WWTP's) and, secondly, skimming of the foam allows to quantify the foam production and to further analyze the foam. This harvesting was considered as a form of wasting from the process and was accounted for in the wasting calculation. The foam was skimmed into a bucket and was mixed thoroughly, to make it a consistent liquid. Total solids were measured for this foam and then the volume of liquid to be wasted (harvested) was calculated, using the formula given in Table 3.2. After the sidestream was implemented for the membrane-based EBPR process, there was poor settling of sludge in the sidestream clarifier. In order to obtain a clear supernatant, it was decided that the clarifier would be drained once daily. Initially, it was drained into the anoxic 27 M A T E R I A L S A N D METHODS zone, but later it was removed from the system as waste sludge (referred to as sidestream wasting), to control the SRT. A l l of the wasting calculations are presented in Table 3.2. Table 3.2. Wasting calculations Traditional method Quantifying foam wasting Wasting through sidestream clarifier Membrane EBPR process Q w = Y,(VS)i/SRT»Saer Where V = Total Reactor Volume (m3). It's the sum of anoxic, anaerobic and aerobic volume reactors. S = Total suspended solids (TSS) in the process (g/m3). It's the sum of TSS in anoxic, anaerobic and aerobic reactors. Qfw - (Qw fbam/Saer) Mass of J^iVSy/SRTS^Sss S s s is TSS in the sidestream clarifier return line. Conventional EBPR process £ (VS)l /SRT* Saer - Qtn . Sou, I So, Q i n = Inflow (m3/d) Sout = TSS in the effluent (g/m3) 3.4 Supplementations 3.4.1 Sodium bicarbonate Approximately 1.5 kg of sodium bicarbonate was added to each influent storage tank daily, as the alkalinity of the influent wastewater was too low. This addition acted as buffering against the loss of alkalinity during the nitrification process. This addition was done based on the previous experience and the final concentration in the influent after addition was 28 M A T E R I A L S A N D METHODS approximately 100 mg as CaCCVL (Frederic Koch, Research Associate, Department of Civil Engineering, U B C , pers. comm.). 3.4.2 VFA addition Volatile fatty acid (VFA) was added to the anaerobic zone as a carbon source, to maximize the phosphorus release. Throughout the study, acetate (5000 mg/L) was added at the rate of 22 mL/min. The V F A in the unsupplemented influent ranged between 30 - 40 mg/L. Previous experience suggested that an additional 30 mg/L of V F A was sufficient to ensure maximum release of phosphorus in this zone (Frederic Koch, Research Associate, Department of Civi l Engineering, UBC, pers. comm.). Sodium acetate was used to prepare the solution. A batch of 140 liters of solution was prepared once every two days. 3.4.3 Magnesium During the magnesium addition period, commercially available magnesium chloride was used to prepare the solution. A 200 L tank was used to the store the magnesium chloride solution and was pumped using a peristaltic pump to the primary clarifier, so that complete mixing was achieved. Magnesium addition was carried out for a period of more than two sludge ages. The pumping rate was 60 mL/m, which gave a concentration of 18 mg/L in the primary influent. The concentration in the feed tank was 2000 mg/L. 3.5 Pumps, mixer and pipelines details The anaerobic and anoxic zones were kept completely mixed by mechanical mixers (Dayton , model no 42128). The aeration process in the aerobic zones also acted as a mixing mechanism. The P R U in the sidestream was kept mixed, using a Dayton® DC motor. A l l recycle and feed pumps used were Moyno Inc® progressing cavity pumps (model no 33260). Baldor® industrial motors were used for both recycles and feed lines. SP 500 VS Drive® controllers were used to adjust the speed of the pumps. In the pilot plant, all connecting pipelines were of Tigerflex® brand. Peristaltic pumps were used for V F A supply and for pumping sidestream and its return flow. A Masterflex ® brand, 14" head were used for V F A and 18" head for sidestream processes. The pipeline connecting 29 MATERIALS AND METHODS the anoxic zone and the P R U was made of Fisherbrand® transparent tubing having lA ID and 1/16 wall. 3.6 Sampling plan and analytical procedures The pH was monitored daily in all zones and in the primary holding tank. The pH was measured using a porTable WP pH Testr BNC, which was calibrated at least three times a week. The dissolved oxygen concentration, along with temperature, was measured using a porTable V W R SP50D. In addition, all the flow rates were measured fortnightly, using a graduated cylinder and stop watch. A systematic sampling and analytical program was designed for monitoring the process performance. Grab samples were collected from both the influent and effluent streams. For the influent, total chemical oxygen demand, soluble chemical oxygen demand, orthophosphate, ammonium nitrogen and volatile fatty acids were analyzed five days per week. Total phosphorus (TP) and total Kjeldahl nitrogen (TKN) were analyzed twice a week. For the effluent, the same parameters were measured, except for volatile fatty acids and total chemical oxygen demand. Mixed liquor samples were collected three times a week before starting the sidestream, to establish a baseline and then were collected twice a week. These samples were analyzed for total suspended solids, V F A (anaerobic zones), ammonia nitrogen, ortho phosphate and nitrates. T K N and TP were analyzed for mixed liquors of the aerobic zone twice a week. Magnesium levels in influent and effluent were measured three times a week, during the magnesium addition period. For sidestream supernatant, orthophosphate, ammonia nitrogen and V F A were measured five times a week. Initially, soluble COD, TSS, T K N and TP were also measured, in order to characterize the supernatant. Magnesium levels were measured three times a week, during the period of magnesium addition to the influent. Except for the M E B P R effluent, all other samples were filtered using 1.2 /*m filter paper (Fisherbrand®) to measure the soluble parameters. The analytical methods used for the experiments were as given in Standard Methods (APHA et al, 1998). Total orthophosphate 30 M A T E R I A L S A N D METHODS (method # 4500PG) and total ammonia-nitrogen (method # 45OO-NH3 H ) were analyzed using flow injection analysis (model Lachat QuickChem® 8000). Nitrate was analyzed using Standard Method 4500 N O 3 T . Total magnesium was measured using atomic absorption spectrophotometry (model Varian Inc. SpectrAA220® Fast sequential atomic absorption spectrophotometer), using Standard Method 3111. Instrument operational parameters are detailed in Appendix B. T K N and TP were digested and analyzed in general accordance with A P H A et al. (1998) Method 4500-N OrgD. The processed samples where transferred into Lachat tubes and stored in a 4° C refrigerator, prior to analysis and were analyzed using the Lachat analyzer (Method 10-115-01-1-Z). 3.7 MAP Crystallizer A previously built pilot scale magnesium ammonium phosphate (MAP) crystallizer was used in this study to recover phosphorus in the form of struvite. The reactor design was as shown in Figure 3.3, which is based on the fluidized bed reactor concept. The sidestream supernatant was stored in a huge tank (~ 3000 L) and processed batch wise. The crystallizer, once started, was in operation for three to four days. The feed to the crystallizer consisted of sidestream supernatant, a mix of magnesium chloride and ammonium chloride solution and caustic soda, all of which were fed to the injection port of the crystallizer. In the injection port, the solutions were mixed to achieve the required super saturation ratio. The dimensions and material details of the crystallizer are outlined in the Figure 3.3. The total volume of the reactor was 24.5 liters. 3.7.1 Top and external clarifiers The purpose of the top clarifier was to arrest the fines from washing out of the reactor. Two outlets were provided to this clarifier. The lower one was connected to the external clarifier through a vertical 2.5 cm inside diameter clear PVC pipe and the upper outlet was used in case of contingency situations. The main function of the external clarifier was to recycle fine particles back to the reactor. The diameter and the height of the top clarifier were 20.3 and 45.7 cm, respectively, made up of clear acrylic pipe. The external clarifier had a square pyramidal bottom with a 45° slope. The approximate external clarifier volume was 54 liters, 31 M A T E R I A L S A N D METHODS with the recycle flow withdrawn from a port on the side, approximately 15 cm below the water surface. The effluent drain line was equipped with a three way valve, to allow flow measurements. The tubing used for drain line and flow measurements was 1.27 cm outside diameter, LDPE (low density polyethylene) tubing. Top clarifier/ Seed hopper 2" pH -Probe Injection port 1.5' To drain (sampling point) Harvesting zone Caustic Recycle flow Figure 3.3. Schematic of the U B C M A P crystallizer 3.7.2 Seeding and harvesting procedure The crystallizer was seeded at the start of each run. The top clarifier served as a seed hopper. Previously-grown crystals from the LuLu Island WWTP were used as seed material. Seeding was required to overcome the time lag of the nucleation period (Dastur, 2001). Dry seed of 32 M A T E R I A L S A N D METHODS two different sizes was poured into the reactor through the top clarifier while the system was running with water and they were removed from the harvesting zone, after ten to fifteen minutes. These seeds were cloth dried and their wet weight was noted. This procedure was repeated at least three times, and the initial average wet weight of seed materials was calculated. After noting the average wet weight, the seed was again poured in to the reactor to run with supernatant. The crystals produced were harvested at the end of each run from the harvesting zone, which was located at the bottom of the reactor. The nominal diameter of the harvesting section was 106 cm. Using ball valves, the harvesting section was isolated from the remaining sections. Crystals were then collected in a bucket. Also, the reactor was flushed using water from the top clarifier to collect the fine materials deposited in the top clarifier. Harvested crystals were cloth dried, and the final wet weight was measured. For each run, the same sieve sizes were used to sieve the seed crystals and the harvested seeds. The sieve size openings used were 2.0 mm and 0.5 mm. 3.7.3 Chemical supplementation, storage tanks and pumps Commercially available magnesium chloride (MgCl2.6H 20) was used as the supplement for magnesium, as the wastewater at U B C pilot plant contained very low levels of magnesium. Theoretically, magnesium, ammonium and phosphate are required in 1:1:1 molar ratio for struvite formation. The required concentration in the feed was calculated from the Potts1 model. Sodium hydroxide (NaOH) was used as an alkali to control the pH. Ammonium chloride (NH4CI2) was also added, since the sidestream supernatant had ammonium in the range of 10 to 12 mg/L only; this would result in low crystallizer efficiency and high pH requirements. 'This assembled model is an unpublished work developed at U B C by Daniel Potts. This model uses an iterative algorithm (powered by Excel Solver) to find a solution to the problem. The problem consists of two chemical equilibrium equation linked by dilution and struvite precipitation. The main problem variables are recyle ratio, pH and struvite precipitation. Feed is mixed with recycle flow to make a "mixed influent". The constituents of the mixed influent are calculated to achieve the target influent SSR and comply with numerous solution equilibria. The struvite removal is calculated to achieve the target SSR in the effluent. Resultant effluent concentrations are the fed back into the mixed influent concentrations. Equilibrium constants are adjusted for temperature in all cases. 33 M A T E R I A L S A N D METHODS Magnesium chloride and ammonium chloride solution was pumped at the rate of 80 mL/s, using a MasterFlex L/S variable speed peristaltic pump with standard pump heads. The tubing used for this purpose was 0.63 cm outside diameter LPDE tubing. The tank was 157 L in capacity and the solution was prepared each day, as long as the run lasted. Caustic solution was dosed into the reactor automatically, whenever the measured pH was lower than the setpoint pH. The caustic solution was stored in the 50 L tank, and was prepared as, and when, required. The strength of the caustic solution used through out this study was 0.1N. The flow rate to the crystallizer was calculated using the Potts model. The recycle flow and feed flow were measured once daily using a stop watch and a graduated cylinder. The influent flow rate was measured at the drain line originating from the external clarifier. The recycle flow was measured from the pipe connecting the top and external clarifiers. For both the feed and effluent recycle, a Moyno Model 500 progressive cavity pump, with Vz HP motor and adjusTable drive speed, was used with 1.27 cm outside diameter LDPE tubing. 3.7.4 pH control The pH was monitored continuously inside the reactor. The addition of caustic to the reactor depended on the measured value of the pH and the setpoint value of the pH. The pH probes were calibrated before each run. Model HD-pH-1 by SA® was used for monitoring the pH inside the reactor and subsequent dosing of caustic to the reactor. The setpoint pH value, based on the Potts model calculation, was adjusted whenever there was an increase in cloudiness in the external clarifier. This cloudiness was the fines of struvite and occurred due to over saturation inside the reactor. The pH in the external clarifier was monitored using an Oakton continuous pH monitor. 3.7.5 Sampling plan Feed samples and effluent samples were analyzed for total magnesium [Mg ], total ortho-phosphate [PO4 -P] and total ammonium nitrogen [ N H 4 - N ] . At least two samples were collected every 24 hours. Effluent samples were collected from the external clarifier and the influent samples (sidestream supernatant) were collected from the storage tank. Figure 3.3 shows the sample collection points. A l l samples were filtered through 1.2 Fischerbrand® filter paper, in order to avoid plugging of instruments used in the analysis. 34 M A T E R I A L S A N D METHODS Samples for N H 4 - N were preserved using two drops of 5% v/v sulphuric acid. Samples for P O 4 - P were preserved by adding one drop of phenyl mercuric acetate. A l l samples for M g were preserved by adding 0.05 mL concentrated nitric acid. A l l samples were stored at 4°C, until analysis. Analytical methods used for analysis were as given in Section 3.6. 35 RESULTS A N D DISCUSSIONS CHAPTER 4 - RESULTS AND DISCUSSION 4.1 Simulation of Sidestream Process The main purpose of simulating the sidestream process was to develop a clear picture as to where to draw the sidestream, to determine how much to take and to see i f there would be any negative impact on the mainstream EBPR process. The process configuration as applied in A Q U A S I M (simulation software) is shown in Figure 4.1. A simulation exercise was carried out only for the M E B P R based process; it was intended only for discussion and understanding the process, before implementation at the pilot plant. Appendix A describes, in detail, the ASM2 bio-P model components and processes used, as given by Meijer (2004). A l l of the kinetic rate equations and stoichiometric coefficients used are also tabulated in the Appendix A (Meijer, 2004). Any values changed, as a result of calibration, are also tablulated1. 4.1.1 Sidestream configuration in AQUASIM The EBPR process at the pilot plant was based on the UCT process and it consisted of an anaerobic zone followed by anoxic and aerobic zones (see Chapter 2 and 3). The M E B P R process of the pilot plant was configured on A Q U A S I M as depicted in Figure 4.1 (AI-Atar, 2006). The A Q U A S I M platform allows defining the spatial configuration of the system as a set of compartments which can be connected to each other by links (Richeirt, 1996). Thus (as shown in Fig. 4.1), R l was constructed as an anaerobic zone, R2 as an anoxic zone and R3 as an aerobic zone with internal anoxic recycle (Ql) and aerobic recycle (Q2). The various modelled processes are given in Appendix A (Meijer, 2004; AI-Atar, 2006). Compartment R5 (phosphorus release unit) and R6 (clarifier) were added to this existing configuration with the sidestream return flow (Q3), as shown in Figure 4.1. 'As indicated earlier model calibration was carried out by Eman AI-Atar, Dept. of Chemical and Biological Engineering, U B C . 36 RESULTS A N D DISCUSSIONS The phosphorus release unit (R5) was simulated in a similar fashion to the anaerobic zone (Rl) of the main EBPR process. The clarifier was similar to that of membrane unit (R4) except that a residual X e ff component was included to include the loss of sludge in the overflow. In case of membrane, sludge was separated completely from the liquid. A l l compartments were simulated as completely mixed reactors. A l l the compartments were connected using advective links which represented water flow, except for the membrane and clarifier compartments. The link connecting the membrane and clarifier units with other compartments were diffusive links which allowed defining selective penetration of certain substances (Reichert, 1996). Throughout the simulation exercises the volume of the PRU was kept constant at 50 L and f was varied between from 0.01 to 0.5, unless otherwise mentioned. The main EBPR process was operated at 20 days SRT and 10 hours HRT for all simulations carried out. Sidestream process To MAP j crystallizer • Influent Main EBPR process Figure 4.1. Schematic layout of sidestream M E B P R process 4.1.2 Source of the sidestream Simulations were run for two sets of conditions. The sidestream flows were taken individually from each zone and from multiple zones in combination. When the sidestream was taken from only one zone, it was found that the anoxic zone produced a supernatant (clarifier overflow is referred to as supernatant) that was most suitable for struvite production. Figures 4.2 and 4.3 show the phosphate and ammonia release and their molar ratio, respectively. The supernatant produced by the anoxic 37 RESULTS A N D DISCUSSIONS I 1 1 1 1 1 1 1 ;—i 0 1 2 3 4 5 6 7 8 PRU Hydraulic Retention Time (hrs) Figure 4.2. P O 4 - P and N H 4 - N release trend in P R U 38 RESULTS A N D DISCUSSIONS zone (referred to as anoxic supernatant from here afterwards) had the optimum N H 4 - N and PO4-P release (Fig. 4.2). To generate the results shown in Fig. 4.2, the volume of the P R U was kept constant at 50 L and f\ was varied from 0.02 to 0.2 to get different retention times. For the anoxic zone supernatant the molar ratio of NH4-N(tot):P04-P(tot) was 1:1 (Fig. 4.3), which favored struvite formation when compared with the aerobic zone supernatant. The anaerobic zone supernatant also presented favorable characteristics for struvite formation; however, P O 4 - P release from the anoxic zone sidestream was higher than the anaerobic zone (Fig 4.3), meaning that more phosphate was diverted for recovery from the main process. Both the aerobic and anoxic zones are good sources of phosphorus, due to the phosphate uptake phenomenon occurring in these two zones. Also the total suspended solids (TSS) concentrations in these two zones were higher when compared to the anaerobic zone which means more release of P O 4 - P . The TSS was 4850 mg/L, 2525 mg/L and 1525 mg/L in the aerobic, anoxic and anaerobic zones, respectively. The anaerobic zone has both phosphorus and ammonia coming from the influent and anoxic recycle lines. For any single zone sidestream flow, the P O 4 - P release was highest with an aerobic zone sidestream, due to its high suspended solids content. However, there was not enough ammonia available due to the nitrification process going on in this zone. Thus, the supernatant produced by the anoxic zone would be most appropriate for struvite formation. For multiple zone sidestreams, the anaerobic/anoxic combination produced the most suitable results for phosphate recovery, compared to the anaerobic/aerobic combination (Fig. 4.4). Figure 4.4 shows that when the anaerobic fraction was higher, the N H 4 - N release was higher. It increased further with the anaerobic/anoxic combination, compared to the anaerobic/aerobic combination. From Figure 4.4, it can be seen that for the anaerobic/aerobic combination, the N H 4 - N release was lowest when the anaerobic fraction was least and P0 4 -P release was highest when the aerobic fraction was highest. This confirms that there was little or no ammonia in the aerobic zone. Since ammonia is present in both the anoxic and anaerobic zones, the variations in the N H 4 -N release were minor, when compared to the anaerobic/aerobic combination. The aerobic/anoxic combination would produce a super-rich phosphorus supernatant, but without much ammonia. Thus, the addition of ammonia would be required to produce struvite. The other option is to mix this supernatant with the 39 RESULTS A N D DISCUSSIONS ammonia-rich supernatant, usually obtained from the anaerobic digestion of return sludge. The advantage of the anaerobic/anoxic combination is that more N H 4 - N and P O 4 - P were available from the system, when compared with the other combination. In other words, the required molar ratio for struvite formation was achieved. Therefore, the anaerobic/anoxic combination was concluded to produce the most suitable supernatant for struvite formation. The sidestream process presents two situations for choosing the value off\ (fraction of inflow to be taken as sidestream). When f\ is low, a high PO4-P release is obtained, but recovery is low, and vice versa. This is illustrated in Figure 4.5, for the anoxic sidestream. As f\ increases, the P recovery percentage increases; as more biomass (at higher f\ value) enters the PRU, the release of P O 4 - P is reduced due to the reduced retention time in the PRU. However, at the same time, more biomass containing phosphate enters the sidestream process, meaning improved recovery efficiency. If the selected recovery technology for phosphorus is struvite crystallization, then a high concentration of phosphorus is required in the supernatant. 0.06 0.05 0.04 v_7 0.03 A 0.02 4 0.01 4 0.00 I Anaerobic 3 Aerobic 1 Anoxic -#— PO4-P O NH4-N 6 8 10 12 14 Different combinat ions Figure 4.4. PO4-P and NH4-N release trend for combined sidestream flow 40 RESULTS A N D DISCUSSIONS 200 0 -\- 1 — i 1 1 1 1 1 1 1 1 1 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 A Figure 4.5. PO4-P release and recovery trend for anoxic sidestream. Potentially, a high concentration also means high phosphate recovery efficiency in the M A P crystallizer. The operators at the wastewater treatment plants can then decide on which sidestream flow to use based on the phosphorus recovery efficiency targeted. 4.1.3 Hydraulic retention time From Fig. 4.6, it can be seen that for both individual and multiple sidestreams the maximum PO4-P release was reached with one hour of retention time in the phosphate release unit. Subsequently, there was no significant increase in PO4-P release. In Fig 4.6, the retention time was kept constant for each/i value i.e., 0.05, 0.1 and 0.2, by varying the volume of the PRU. The initial concentration of PO4-P in the anoxic zone (4.5 mg/L) of the main EBPR process is also plotted on the y axis (at zero HRT in PRU). The initial concentration in the anaerobic zone was 12 mg/L and the initial concentration of PO4-P for multiple sidestream (anoxic/anaerobic), was calculated to be approximately 8.2 mg/L. In Fig. 4.6, for conditions f\ = 0.05 (Anoxic/Anaerobic) and f\ - 0.1 (Anoxic), the phosphate release is similar. The 41 RESULTS A N D DISCUSSIONS difference is in the ammonia release; for the combined sidestream process, it was around 20 mg N / L (see Fig. 4.4) and for the individual sidestream, it was around 12 mg N / L (see Fig. 4.2). As discussed earlier, the anoxic/anaerobic combination would be preferred as the required molar ratio of P O 4 - P and N H 4 - N for struvite formation can be achieved. From Fig. 4.6 it appears that, at higher multiple sidestream flows there is no actual release of P O 4 - P in the PRU. A batch test was also carried out using sludge from the U B C pilot plant (Fig 4.7), operating at similar conditions to those used in this model. From Figure 4.7, it can be seen that the maximum PO4 -P was released by 1.5 hours of retention time, after which there was no significant additional phosphate release. • /. = 0.05 (Anoxic) 0 /, = 0.05 (Anoxic/Anaerobic) — /. = 0.1 (Anoxic) v /. = 0.1 (Anoxic/Anaerobic) m — /. = 0.2 (Anoxic) • /. = 0.2 (Anoxic/Anaerobic) 40 0 H 1 1 1 1 1 0.0 0.5 1.0 1.5 2.0 2.5 PRU Hydraulic Retention Time (hrs) Figure 4.6. PO4 -P release with time for different sidestream flow. 42 RESULTS A N D DISCUSSIONS 100 Time (hrs) Figure 4.7. Batch test results for PO4-P release for individual and combined sidestream. 4.1.4 Effects on the main EBPR process effluent and potential P recovery estimation At lower f values (< 0.2), effluent P O 4 - P and N H 4 - N values of the main EBPR process remained unaffected. When f\ was increased and keeping fa (fraction of sidestream returning to main process) zero, there was an increase in the effluent P O 4 - P and N H 4 - N values of the EBPR process. When/2 was set at 0.1, the process regained its stability. Thus, the EBPR process required a minimum fi to be set at 0.1, otherwise there was deterioration in the effluent quality. This was expected since, at higher values of f, more phosphorus accumulating organisms (PAOs) containing polyphosphate (along with the bulk liquid) are removed from the main EBPR system; this means a significant amount of phosphorus is removed. The leftover phosphorus in the main EBPR process is insufficient for physiological requirements of cell metabolism and growth of biomass. Also, at higher f (at f\ = 0.2, Qss is 1080 L) values the SRT of the main EBPR process would fall below the recommended SRT (2-4 days) for biological phosphorus removal process (Tchobanoglous, 2003). Thus, keeping fi zero and increasing f\ resulted in failure of the EBPR process. This is illustrated in Figure 4.8 a, b, c and d. For the anaerobic sidestream process, the failure was not significant when fi was small, but when fi = 0.5, there was a sharp increase in the effluent values of phosphorus and ammonia (Fig. 4.8a). For the anoxic sidestream and combined sidestream, failure 43 RESULTS A N D DISCUSSIONS occurred when f\ was 0.3 and 0.2, respectively (Fig. 4.8b and 4.8c). The failure for the anoxic and combined sidestream process occurred at the lower values offi because a higher portion of the PAOs and autotrophic organisms are wasted out of the EBPR process, when compared to the anaerobic sidestream. Fig. 4.7d illustrates the influence of sidestream flow and sidestream return flow on the biomass. The concentrations of PAOs and autotrophic organisms declined in the EBPR process as the value of/} increases. When fi is 0.5 and fi is zero, the PAOs concentration nearly reaches zero. This means that the phosphate release and uptake phenomenon was not occurring. When the sidestream return flow (QSSR) was introduced, it acts like seeding for the process and the EBPR process regains its stability. Thus, QSSR is very critical in sustaining the EBPR process with this sidestream. QSSR also influences the PO4-P release in the P R U (sidestream flow). Fig. 4.9 demonstrates the effect of QSSR on PO4-P release. An increased value of fi indicates that more biomass (mainly PAOs) are recirculated to the main EBPR process from the sidestream. This results in an increase uptake of phosphate in the anoxic zone and, in turn, results in a high release of phosphate in the PRU. A reduction in nitrates ( N O 3 " ) was one more important observation made. From Figure 4.9, it can be seen that, as the/2 increases (f\ was kept constant at 0.05), there was further reduction in the amount of nitrates in the effluent. This suggests that denitrification is aided by the extra carbon coming into the process through QSSR- However, this reduction is marginal as predicted by the model. The value of 72 can be based on the flow required to run the M A P crystallizer, the amount of P O 4 - P release required and possibly on the level of N O 3 " required. One more significant impact of the sidestream process is the decrease in' the total phosphorus in the aerobic zone of the EBPR process (Fig. 4.5); this means less P O 4 - P is released during the sludge treatment process. This could also reduce problems like formation of struvite in the pipelines of sludge handling equipment. 44 RESULTS A N D DISCUSSIONS 0 1 2 3 4 5 6 7 8 9 Different combinations fx k P0 4-P N H 4 - N Figure 4.8 (a) EBPR process failure due to anaerobic sidestream. 0 1 2 3 4 5 Different combinations 7 l / l 1/2 - P 0 4 - P - N H 4 - N Figure 4.8 (a) EBPR process failure due to anoxic sidestream. 1 2 3 4 Different combinations A h PO4-P NH4 -N 0 1 2 3 4 5 6 Different combinations 7 8 9 ft f2 — • — PAO's •• " O - " Autotrophs — ••— Polyphosphates Figure 4.8 (c) Failure due to combined sidestream (anoxic/anaerobic) Figure 4.8 (d) Response of biomass to changes in f\ and/ 2 45 RESULTS A N D DISCUSSIONS 0 J 1 1 1 1 1 i 1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 fi Figure 4.9. Effect offi on P O 4 - P release (anaerobic zone) and N O 3 " (MEBPR effluent) A total phosphorus mass balance was used to evaluate the potential phosphate recovery. The phosphorus content in the effluent, activated sludge, and side stream supernatants were calculated as a percentage of the influent total phosphate load. When/2 was kept constant at 0.2 and varying f\ between 0.01 and 1, 30 to 78 % of the incoming phosphorus could be diverted for the recovery process (Fig. 4.10). One additional factor to be considered here is the efficiency at which the M A P crystallizer is operated. The effluent from the struvite crystallization process could be redirected to the main EBPR process, along with traces of phosphorus, ammonia and magnesium. For calculation purposes, the M A P crystallizer process is assumed to operate with 100 % efficiency, even though, practically, this would never be the case. This situation would have to be recognized in any final process design calculation. 46 RESULTS A N D DISCUSSIONS 90 10 J 1 , 1 1 1 1 1 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Figure 4.10. Estimated phosphorus recovery for different sidestream flows. 4.1.5 Simulation results Simulation results revealed that the anoxic zone was most suitable for taking a sidestream flow when the flow was taken from a single zone. When taken from multiple zones, the anaerobic/anoxic zone combinations yielded the best conditions for struvite formation. The hydraulic retention time required in the phosphate release unit was two hours. Simulation also predicted that the main EBPR process would collapse under a high sidestream flow rate (Qss> 50 % of Qin) i f the sidestream return flow (QSSR) is not maintained. The sidestream return flow also influenced the anaerobic phosphate release and effluent nitrates level. Finally, the ASM2 bio-P model predicted that 78 % of the incoming phosphorus could be diverted for recovery as struvite through a sidestream process. 47 RESULTS A N D DISCUSSIONS 4.2 Experimental Investigation The entire experimental investigation lasted more than a year. It was divided into three periods. Period I occurred before implementing the sidestream process. During this period, it was ensured that both M E B P R and CEBPR processes were running efficiently with the desired SRT and HRT. Period I lasted for nearly five months. Period II took place after implementing the sidestream, which remained in place for the rest of the study. Period II was subdivided into two parts, the first, without sidestream wasting and the second with sidestream wasting (see Section 3.3 and 4.3). Finally, for approximately 40 days (two sludge ages), magnesium was added to the influent to increase the M g concentration in the sidestream supernatant. This was called Period III. These time divisions were the same for both the conventional and membrane-based processes. For the conventional process, there was no wasting from the sidestream process carried out during Period II. Table 4.1 summarizes the operating conditions for the different periods. Values of fi and fi were selected based on the simulation results (see Section 4.3). Table 4.1. Experimental periods with sidestream operating conditions Operating Conditions Period fx /a® Influent Mg HAc Days mg/L mg/L I 0 0 1.2 0 01 -132 II a) Without sidestream wasting 0.025 0.2 1.2 49 132-225 b) With sidestream wasting (only for MEBPR) 0.025 0.2 1.2 49 225-379 III 0.025 0.2 18 49 287-331 "The main EBPR process was operated at 10 hrs HRT and 20 days SRT throughout this study. 'f2 is the fraction of Q Ss 48 RESULTS A N D DISCUSSIONS 4.2.1 EBPR process performance: Period I The M E B P R and CEBPR processes exhibited satisfactory removal of nitrogen and phosphorus during Period I, at the set point HRT and SRT. Average PO4-P removal was 99 % and average N H 4 - N removal was 99.5 % for the M E B P R process, while the corresponding values for CEBPR were 91.6 % and 97.4 %, respectively. During this period, the foam from the anoxic zone of the EBPR process was wasted, as described in Chapter 3. This wasting was accounted for in the SRT calculation. Previously, at the pilot plant, this foam was re-suspended by mixing. 4.3 Sidestream Process 4.3.1 Operating conditions Based on the simulation results, the sidestream was taken from the anoxic zone. The required HRT in the phosphate release unit was a minimum of one hour, with fi set to 0.2. For simulation purposes, the volume of the P R U used was 50 liters. At the pilot plant, the volume of the P R U was increased to 120 liters. The fi value was set at 0.025 (0.0925 m3/d). Thus, the HRT was increased to nearly 30 hrs. These conditions were adopted, considering the practical difficulties at the plant in setting up a very small P R U with a clarifier for the chosen fi. Also, a bigger volume of PRU would allow increasing the fi value while maintaining the required hydraulic retention time, if needed in the future. 4.3.2 Phosphate release unit The P R U works similar to the anaerobic zone of the EBPR process. Under anaerobic conditions, acetate is taken up by the PAOs and is stored as poly-hydroxybutyrate (PHB). This action requires energy which is provided by the hydrolysis of cell internal glycogen and poly-phosphate (PP). Thus, under anaerobic conditions with the presence of acetate, PAOs release a large amount of phosphate to the bulk liquid (Comeau et al., 1986). The same process was fostered in the PRU, which produces a phosphorus-rich supernatant. In the case of the anaerobic zone, phosphate release is dependent on the presence of V F A in the influent. For the PRU, an external source of carbon was added in the form of acetate, as the sidestream from the anoxic zone did not contain any acetate. Using acetate as the carbon source ensured maximum phosphate release (Liao et al., 2003). , 49 RESULTS A N D DISCUSSIONS An external source of acetate was also added for the anaerobic zones of both EBPR processes. This was required because previous studies (Monti, 2006) suggested that an additional 30 mg/L of V F A would ensure smooth operation of the EBPR process. Monti (2006) observed that when the bio-P failure occurred in both M E B P R and CEPBR processes, it was linked to the V F A to TP ratio in the influent. When the external carbon was added in the form of sodium acetate, both the CEBPR and M E B P R processes responded by delivering low P O 4 - P concentrations in the treated effluent. 4.3.3 Sidestream clarifier The purpose of the sidestream clarifier was to produce suspended solids-free supernatant. As expected, for the M E B P R sidestream process, the settling in the clarifier was very poor. The sludge volume index of the anoxic zone was around 350 mL/g, whereas a value of 100 mL/g is considered representative of a good settling sludge (Tchobanoglous et al., 2003). This is expected in an membrane-based EBPR process, because membrane bioreactor mixed liquors are subjected to coarse bubble aeration in the aerobic tank; this can destroy the flocculating nature of the sludge, which results in poor settling (Monti, 2006). This settling problem was not observed in the conventional EBPR process. Poor settling resulted in an increased total suspended solids (TSS) content of the supernatant, thus decreasing the efficiency of the phosphate recovery. To overcome this problem, the sludge settled was drained once daily from the sidestream clarifier. In the beginning, the sludge was drained back to the anoxic zone (Period II without sidestream wasting). Later, it was wasted from the system to maintain the SRT of the process (Period II, with sidestream wasting). 4.3.4 Characteristics of the sidestream supernatant The sidestream supernatant produced contained soluble COD in the range of 20 to 50 mg/L. and the pH ranged from 6.89 to 7.34. The TSS content was very low. Initially for the M E B P R process, TSS was very high due to the poor sludge settling in the clarifier, but it came down to the above range after sidestream wasting was implemented. The average values of TP and T K N measured were 21 and 29 mg/L, respectively. It also contained residual V F A from the PRU. The membrane sidestream supernatant was mucky when compared to the conventional sidestream supernatant, which was much clearer. The average concentrations of P O 4 - P and N H 4 - N were 72 mg P/L and 11 mg N / L , respectively, for the 50 RESULTS A N D DISCUSSIONS membrane-based sidestream supernatant. For the conventional side, the corresponding average values were 79 mg P/L and 10 mg N/L . The M g measured was in the order of 1- 2 mg/L. During Period III, the measured concentration was in the order of 35 to 40 mg/L. 4.3.5 P04-P and NH4-N release trend The trend in the phosphate release for both the conventional and membrane-based sidestreams, are given in Figure 4.11a and b. For the membrane sidestream, the initial low phosphate release was mainly due to the lack of proper mixing in the P R U and the limiting V F A . The other possible reason could be the higher levels of nitrate in the anoxic zone of the M E B P R process (Fig. 4.12), such that the HAc's are utilized first by nitrates. If HAc added is limiting, then P O 4 - P release will be decreased further. In general, the P O 4 - P release in the M E B P R and CEBPR sidestream processes were almost equal, with a mean specific phosphate release of 0.025 (mg P/mg TSS) and 0.023 (mg P/mg TSS) respectively. This is demonstrated in the Figure 4.13, where the specific P release for the membrane and conventional process is plotted. The unusual peak of phosphate in the initial stages (Fig. 4.1 la) may be due to the release in the clarifier, under improper mixing conditions. In the conventional sidestream, N H 4 - N release was higher (> 20 mg N/L) during the first week of its implementation. This could be attributed to two reasons. Firstly, there was a shock load of ammonia in the influent (101 mg N/L), (which was very unusual at the pilot plant) which occurred during this time (See section 4.5). This caused a breakthrough in the nitrification and denitrification processes, which was evident in the high concentration of N H 4 - N in both the aerobic and anoxic zones of CEBPR process. Also, during this period, M g addition was also started, after which the process experienced poor removal of phosphate and ammonia. In both the processes, the ammonia measured in the P R U was approximately equal to the ammonia present in the anoxic zone, in the initial few days (approximately 25 days). This indicated that, there was no actual "release" of additional ammonia nitrogen in the sidestream process beyond the concentration present in the anoxic zone. In Figure 4.14, it can be seen that there was a gradual increase in the ammonia nitrogen concentration after fifty days of PRU operation and the increase was major after eighty days. This additional release can be 51 RESULTS A N D DISCUSSIONS attributed to biomass decay, which is associated with conversion of organic nitrogen to ammonia nitrogen. The observed increase happened gradually, as the biomass decay under anaerobic conditions is slow, when compared to aerobic and anoxic conditions (Siegrist et al., 1999). Similarly, the PO4-P in the anoxic zone and PO4-P release in P R U is plotted in Fig. 4.15. The mean value of anoxic PO4-P was around 7.2 mg/L and the mean value of sidestream PO4-P release was 72 mg/L, therefore it can be concluded from the Fig 4.15 that there is considerable amount of PO4-P release in the PRU. X O 300 Figure 4.1 la. PO4-P and N H 4 -N release trend for membrane sidestream process 52 RESULTS A N D DISCUSSIONS Figure 4.1 lb. P O 4 - P and N H 4 - N release in conventional sidestream process 53 RESULTS A N D DISCUSSIONS H 0.030 0.028 H t»o 0.026 H § 0.024 H W 3 0.022 H g 0.020 H 0.018 0.018 0.020 0.022 0.024 0.026 0.028 0.030 CEBPR sidestream [mg P/ mg TSS] Figure 4.13. Specific phosphate release in M E B P R and CEBPR sidestreams 30 i 60 s 25 H 20 4 15 4 io 4 5 H -O— Sidestream - • — Anoxic 1 1 1 1 1 1 1 1 1 1 1 1 1 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 Time [Days] Figure 4.14. Ammonia nitrogen in anoxic zone and P R U for membrane sidestream process 54 RESULTS A N D DISCUSSIONS 160 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 Time [days] Figure 4.15. P O 4 - P in anoxic zone and P R U for membrane sidestream process 4.3.6 Mass of P release to VFA consumed As discussed earlier, the amount of V F A present in the influent affects the efficient working of an EBPR process. Usually, the municipal wastewaters are often low in V F A content and require the addition of V F A . This can be achieved by having a pre-fermented sludge stream or by the addition of HAc directly to the anaerobic zone. So while designing a BNR process, knowing the ratio of phosphate release to V F A consumed, has its advantages. Smolder et al. (1994) reported that along with the P O 4 - P / V F A ratio, the pH of the wastewater should be considered due to its strong influence in the solution. The mass of P released per unit mass of V F A utilized was calculated through a mass balance around the anaerobic zone. Following mass balance equations were used to calculate the mass of V F A utilized and mass of P O 4 - P released. 55 RESULTS A N D DISCUSSIONS Mass of V F A consumed ( M V F A ) M V F A = CJVFA X V F A + QIN X 1 N V F A + QANX XAMXVFA - QANA X A N A V F A > + QS^ SVFA X S S V F A - QSSOUT XSSVFAOUT- • -4.1 -V -Anaerobic zone (EBPR process) PRTj (sidestream process) Where, Q V F A ~ Acetate flow added by external means (mL/min) X V F A = Concentration of V F A in the added flow (mg/L) Q I N = Influent flow to the pilot plant (L/min) X I N V F A = Concentration of V F A present in the influent (mg/L) Q A N X = Anoxic recycle flow to the anaerobic tank (L/min) X A N X V F A - Concentration of V F A present in the anoxic recycle flow (mg/L) Q A N A = Flow entering anoxic zone from anaerobic zone (L/min) X A N A V F A - Concentration of V F A present in the flow entering anoxic zone from anaerobic zone (mg/L) QSSVFA = Acetate flow added to PRU (mL/min) X S S V F A = Concentration of added V F A to the P R U (mg/L) QSSOUT = Outgoing flow from P R U (L/min) X S S V F A O U T = Concentration of V F A in the outgoing flow (mg/L) Mass of P O 4 - P released ( M P 0 4 ) Mpo4 = QINXINPO4 + QANX XANXPO4 _ QANA X A N A P 0 4 QSSVFA X S S V F A + Qss ^SSPO4_ QSSOUT XSSPO40UT 4.2 " ^ ^ v ' Anaerobic zone (EBPR process) P R U (sidestream process) where, Q I N = Influent flow to the pilot plant (L/min) XINPO4 = Concentration of P O 4 - P present in the influent (mg/L) Q A N X = Anoxic recycle flow to the anaerobic tank (L/min) X A NXP04 = Concentration of P O 4 - P present in the anoxic recycle flow (mg/L). Q A N A = Flow entering anoxic zone from anaerobic zone (L/min) X A N A P O 4 - Concentration of P O 4 - P present in the flow entering anoxic zone from anaerobic zone (mg/L) 56 RESULTS A N D DISCUSSIONS Qss = Sidestream flow to the PRU (L/min) XSSPO4 = Concentration of P O 4 - P in the sidestream flow (mg/L) QSSOUT = Outgoing flow from P R U (L/min) XSSPO40UT= Concentration of P O 4 - P in the outgoing flow from P R U (mg/L) While calculating for the situation without sidestream process, the sidestream component in the equations (4.1) and (4.2) was eliminated. As shown in Fig 4.16, the P O 4 - P to V F A ratio was relatively consistent before and after implementing the sidestream process. This ratio did not change much statistically, with and without the sidestream. Although the acetate requirements increased due to the consumption in the PRU, this acetate consumption was much less when compared to the anaerobic zone consumption. The average mass of V F A utilized in the PRU was about 2-3 mg/d and the corresponding mass of phosphate released was around 1-2 mg/d. In the case of the anaerobic zone, the average mass of V F A consumed was 50 mg/d, with phosphate release varying from 15 to 20 mg/d. For the M E B P R process, the ratio of P released per unit mass of V F A consumed was found to be 0.73 and 0.71, without the sidestream and with the sidestream process, respectively. For the CEBPR process the corresponding values were 0.50 and 0.40, respectively. The advantage with the sidestream system is the stored phosphate, which is removed from the biomass, resulting in the low T K N and TP of the sludge produced, and the removed phosphate is recovered as struvite. The ratio observed in all periods in the present study is in close conformity with other EBPR systems reported in the literature; some of the values reported are 0.22-0.40 (Kuba et al, 1997) and 0.85 (Rabinowitz et al, 1986). A previous study done by Monti (2006) at the U B C pilot plant, reported 0.5 for the M E B P R process and around 0.6 for CEBPR process. The dosage of V F A added to the P R U depended on the flow of the sidestream. It increased with the flow (f\), but also with the increased potential of more phosphate recovery as struvite. 57 RESULTS A N D DISCUSSIONS I 1 CO C o U l-i PH •* o PH 1.5 1.0 0.5 0.0 M E B P R Without Sidestream CEBPR With •> Sidestream Figure 4.16. Ratio of P O 4 - P released to V F A consumed 4.4 P Removal Performance The time series of influent TP and effluent phosphate of both processes is shown in Figure 4.17. Phosphorus removal performance was characterized by consistent removal, both in Period I and II. There was an increase in the effluent phosphorus during the Mg addition period. The percent removal of P (PCVPjn - PCVPout), remained approximately the same during Periods I and II (Appendix C). The possible explanation for the failure during Period III is discussed in Section 4.6. The mass of P O 4 - P removed from the system increased during period II, because of the sidestream process recovering phosphate as struvite. The failure in the conventional side during Period II (point 1 in the Fig 4.17) was due to the plugging of the pipeline which connected the aerobic zone and the clarifier. The phosphorus removal efficiencies for both sides were comparable, before and after the implementation of the 58 RESULTS A N D DISCUSSIONS sidestream system. PH I o PH H 0 50 100 300 350 400 150 200 250 Time [day] Figure 4.17. Total phosphorus in the influent and orthophosphate-P in the effluent for M E B P R and CEBPR processes. 4.4.1 Mass of VFA consumed to mass of PO4-P removed The ratio of the mass of V F A utilized in the influent (amount of influent V F A consumed in the anaerobic zone) per unit mass of P O 4 - P removed was calculated for the data set, before and after implementing the sidestream process (Fig. 4.18). The ratio for the sidestream process was much lower (15.9 mg/mg) when compared with the EBPR process alone, (20.3 mg/mg). In an EBPR process, the phosphate uptake in the aerobic and anoxic zone is greater than the phosphate released in anaerobic zone and thus the net removal of phosphate from the bulk liquid occurs. With a sidestream EBPR process, phosphate is also removed (recovered) as struvite. The phosphate released in the PRU is not taken up by the biomass, but removed from the process as a useful product, struvite. Thus, more mass of P O 4 - P is removed from a sidestream EBPR process when compared with an EBPR process without a sidestream. 59 RESULTS A N D DISCUSSIONS Another point to be considered here is the efficiency of the M A P crystallizer to recover phosphate, as struvite. In this study, the struvite recovery rate efficiency observed was about 60 %. The reason for the low recovery efficiency is the shorter duration crystallizer runs (> 90 % recovery efficiency is possible, see Section 4.8). Ideally, the remaining phosphate, which is in the crystallizer effluent, should be directed back to the main EBPR process. This was not done in the present study, as the sidestream supernatant was processed batch wise to recover struvite. If the M A P crystallizer can be operated at higher efficiency to recover struvite, then the ratio will drop further. o < 1 2 3 Without sidestream After sidestream Figure 4.18. Ratio of V F A utilized to phosphate removal for M E B P R sidestream process. 4.4.2 Puptake and release A mass balance calculation was performed around the anaerobic zone and the PRU, to determine the mass of phosphate released and to assess the uptake of phosphate around the anoxic and aerobic zones. The masses were divided by the influent flow rate, to express the release and uptake in milligrams of P per liter of influent flow (Monti, 2006). The mass balance was performed on daily basis and the equations used on the spread sheet are given in Appendix D. The time series of the uptake and release trend for both the M E B P R and CEBPR processes is shown in Figures 4.19 and 4.20, respectively. From Figure 4.19 and 4.20 it can be seen that anoxic phosphate uptake was predominant and found to be 60 RESULTS A N D DISCUSSIONS contributing approximately 80 % of phosphate uptake, indicating the ability of PAOs to utilize nitrate as an electron acceptor for phosphate uptake (Hu et al., 2002). It can be seen from Figure 4.19, for the M E B P R process, that the phosphate release and uptake was increased dramatically, when the sidestream wasting was implemented (Period II). The observed increase was nearly 375 %, when compared with the process without wasting from the sidestream clarifier. As discussed earlier, to avoid sludge settling problems in the clarifier, it was decided to waste the sludge from the clarifier instead of from the aerobic zone of the main process. This resulted in the increased phosphate release in the. anaerobic zone and corresponding uptake in the anoxic and aerobic zones. During the M g addition period (period III), the release and uptake was reduced in both M E B P R and CEBPR processes, but the system returned to normal conditions after the addition was stopped (see Figure 4.19 and 4.20). For the CEBPR process, when the process failed due to the plugging of the pipe line connecting the aerobic zone and the clarifier, both the anoxic uptake and anaerobic release was affected (Fig 4.20). 150 S 100 o 200 0 50 100 150 200 250 300 350 400 Time [Days] Figure 4.19. M E B P R sidestream process: phosphate release and uptake trend. 61 RESULTS A N D DISCUSSIONS 60 S > -a > <u -*-» & O OH 60 40 H 20 -20 4 -40 4 -60 Anaerobic release Anoxic uptake Aerobic uptake 50 100 150 200 250 300 350 400 450 Time [Days] Figure 4.20. CEBPR sidestream process: phosphate release and uptake trend. 4.4.3 Simulation of Sidestream wasting To explain this phenomenon, the wasting from the sidestream process was simulated. It was achieved by making changes in A Q U A S I M . For operation without sidestream wasting, the amount of sludge to be wasted (QeXcess) was linked to the aerobic zone of the main EBPR process (AI-Atar, 2006). For operation with sidestream wasting, QeXcess was linked to the sidestream clarifier. The simulation exercise revealed the following. 1. PAOs dominated the system with sidestream wasting. The PAOs concentration in the system (including in the PRU), increased by 55 %. Cell internal storage of polyphosphate (Xpp) and glycogen (XGLY ) increased by 355% and 47 %, respectively. The TSS in all zones remained approximately the same, under both conditions. 62 RESULTS A N D DISCUSSIONS 2. With sidestream wasting, the anaerobic phosphate release and sidestream phosphate release was similar (see Fig. 4.21). 3. Heterotrophic organisms were reduced by 99.8%. Autotrophic organisms remained approximately the same, under both conditions. 4. P0.4-P and N H 4 - N removal efficiency increased. 5. During the sidestream wasting period, more V F A was consumed in the anaerobic zone; as a result, there was an increase in PHA content also. I / W V W W Soluble phosphate Anaerobic P Biliasi Soluble phosphate Anatrobic P Mtm • 30°. „ timc(d) 500 600 Figure 4.21. Anaerobic and sidestream phosphate release: Before sidestream clarifier wasting (above) and after implementation of sidestream wasting (below). 63 RESULTS A N D DISCUSSIONS The sidestream process acted as continuous wasting from the process, at a very slow rate (0.025% of Qi„ which was 0.0953 m3/d). At the pilot plant, after wasting from the sidestream clarifier and the foam, an average of 15 to 20 L/d was also wasted from the aerobic zone to maintain the desired SRT of 20 days. As a result, this increased the net retention time of the solids inside the reactor (i.e., there is no increase in the SRT, but there is no washout of biomass from the system at once); thus, a large amount of PAOs (which are phosphate-rich) were not wasted from the main EBPR system as only 15 to 20 L/d was wasted. The sidestream waste sludge containing PAOs were depleted out of phosphate in the P R U (i.e., phosphate-poor sludge). This may explain the increased poly-phosphate concentration in the system. Another factor contributing to the increased population of PAOs is the sidestream return flow, which contains poly-P organisms. Most of the time, the remaining acetate in the anaerobic zone during the sidestream wasting period, was negligible or zero (Appendix C); this supports the fact that PAOs were dominant in the system, as the amount of poly-P bacteria produced in a system results predominantly from the anaerobic uptake of volatile fatty acid. The concentration of poly-P organisms is, in general, only a small fraction of the total biomass concentration in the system. The measured P content of the total biomass in the system is the average P content of the poly-P organisms and other non-poly-P organisms (Smolders et al, 1996). The TP measured for the aerobic zone during this period was comparatively high (Fig. 4.22), indicating the presence of more poly-P organisms. 64 RESULTS A N D DISCUSSIONS 350 i 200 0 1 2 3 Without sidestream wasting With sidestream wasting Figure 4.22. Box plot for aerobic TP of M E B P R sidestream process. Percentiles shown are 5 t h, 10 t h, 25 t h, 75 t h, 90 t h and 95 t h. The horizontal line inside the box represents median. As the simulation results predicted, the anaerobic release and sidestream phosphate release were not equal, but the anaerobic release was higher compared to the sidestream release (Fig. 4.23). Also, the advantage of wasting from the sidestream clarifier is that the sidestream sludge has lower T K N and TP values. T K N and TP were measured for a short duration and the average values obtained were 40.2 mg/L and 74 mg/L, respectively. Also, the wasting from the aerobic zone was reduced significantly, from an average 80 L/d to 15 L/day. 65 RESULTS A N D DISCUSSIONS 0 -250 J 1 1 1 1 1 1 1 1 1 1 0 50 100 150 200 250 300 350' 400 450 Time [Day] Figure 4.23. Phosphate release in anaerobic zone and P R U after implementing sidestream wasting 4.5 N Removal Performance The time series of influent T K N and the effluent nitrate and ammonia concentrations of both processes are presented in Figure 4.24. The CEBPR process exhibited more nitrogen removal than the M E B P R unit, due to the extra sink for nitrate existing in the clarifier sludge blanket; this is in line with other studies reported in the literature (Siegrist and Gujer, 1994; Monti 2006). Both the M E B P R and CEBPR processes achieved complete and sTable nitrification during this study. On February 20 t h of 2006, the pilot plant experienced a shock load of ammonia (TP was approximately 60 mg/L) in the influent (point 3 in Fig 4.24). The M E B P R process was more resistant to this shock load except for a single breakthrough, after which the process regained its stability. The CEBPR process was more vulnerable to this shock load, and took longer to 66 RESULTS A N D DISCUSSIONS perform satisfactorily. During this period, nitrification and denitrification failed, which was evident with a high value of N H 4 - N both in the anoxic and aerobic zones. This also increased the N H 4 - N release in the CEBPR sidestream process. Point 2 in Fig 4.22 is the process failure occurring due to the clogging in the pipe line connecting the aerobic zone and secondary clarifier. For the M E B P R process during Period II (point 1 in Fig. 4.24), when the sidestream clarifier sludge was drained to the anoxic zone, the nitrate concentration decreased, due to the extra carbon entering the system; the nitrate concentration increased again when the wasting from the sidestream started. The same pattern was difficult to identify in the CEBPR sidestream process, because the addition of magnesium and sidestream operation started simultaneously. 50 100 150 200 ' 250 Time[Days] Figure 4.24. Total nitrogen in the influent, and ammonium and nitrite+nitrate in the M E B P R and CEBPR effluent 67 RESULTS A N D DISCUSSIONS 4.6 Effect of Mg Addition Magnesium and potassium ions are also released along with phosphate, under anaerobic conditions in the presence of acetate. Magnesium ions are beneficial for the struvite crystallization process. To increase the concentration of M g ions in the supernatant, it was decided to add magnesium chloride to the both CEBPR and M E B P R process influent (Period III). The concentration was increased from 1.2 mg/L to 18 mg/L. As expected, there was an increase in the M g release (Fig. 4.25) in the phosphate release unit, but the EBPR effluent concentrations of phosphate and ammonia also increased during this period. This period lasted for more than two sludge ages. The deterioration of the effluent quality was almost immediate, as it was observed after the first day of addition. The following observations were made during this period. 1. The effluent concentrations of N H 4 - N and P O 4 - P increased to about 2 mg/L for both pilot processes. The phosphorus release in the anaerobic zone experienced an average decrease of 37 % during this period (Fig 4.18 and 4.19), despite maintaining the acetate source. The corresponding uptake was also reduced. 2. The M g concentration increased in the sidestream supernatant, reaching a maximum of 40 mg/L. This confirmed that magnesium is co-transported, along with phosphate, as observed by others. (Wentzel et al, 1989; Pattarkine et al, 1990). 3. The effluent concentration of Mg was approximately the same as the influent concentration during this period. The removal of M g during this period was very low, around 6 -12 %. The removal of Mg during the previous period was 50-60 %. 4. Both pilot processes regained stability after nearly a month from the day the M g addition was stopped. Its been established that, for a successful EBPR process, both M g 2 + and K + are needed in specific molar quantities (Rickard and McClintock, 1992; Brdjanovic et al, 1996) in phosphate release and uptake and they are unable to substitute for each other. Potassium and magnesium concentrations were not measured routinely at the pilot plant. Sporadic measurements done in the previous years indicate an influent K value in the range of 8 to 10 68 RESULTS A N D DISCUSSIONS mg/L. The average value of Mg measured in the influent during this study was around 1.2 mg/L. The EPBR process performance was good before adding, and after stopping, the magnesium addition. Thus, for the pilot plant conditions, it is appropriate to assume that the influent Mg/P (0.34) and K/P (2.85) ratios were sufficient for the successful operation of an EBPR process. When the Mg/P ratio was increased, the process failed. This might be an indication of a relationship between the cations for the phosphorus release and uptake to proceed. The fact that the removal/uptake of magnesium decreased during the addition period (6-12 % from 50-60%) indicates that the magnesium uptake was influenced by the potassium concentration. This finding is in line with the other findings reported in the literature (Miya et al, 1987; Rickard and McClintock, 1990). Rickard and McClintock (1992) showed that the EBPR process failed when either potassium or magnesium was limited in the influent. Brdjanovic et al, (1996) studied the dynamic effects of potassium limitation on biological phosphorus removal and found that a severe shortage of potassium in the influent caused an absence of phosphorus removal. Table 4.2 gives the Mg/P and K/P ratio reported in these two studies, along with the present study. Table 4.2. Influent Mg/P and K/P values studying effects of M g and K on EBPR process. Mg/P K/P Process Remarks Influent Influent performance 0.34 2.85 Good Present Study Pilot Scale study 5.14 2.85 Failed 0.44 0.58 Good Rickard and A/O continuous study McClintock, 0.46 0 Failed with synthetic (1992) wastewater 0 0.57 Failed Brdjanovic et al (1996) - 0.0009 1.25 Failed Good Anaerobic-aerobic sequenced batch reactor system using sludge from full scale plant 69 RESULTS A N D DISCUSSIONS 50 40 H • Influent •O Effluent MEBPR Effluent CEBPR Time [Days] Figure 4.25. Magnesium in P R U (MEBPR sidestream), influent and effluent for period III. Recently, Wu et al. (2006) reported that, when MgCb was used for the M g 2 + source, there was a decrease in the anaerobic phosphate release but it stimulated the aerobic uptake. A more detailed study is required to address these issues and is beyond the scope of this work. 4.7 P Recovery Study The M A P crystallizer was operated in a batch mode. The sidestream supernatant was stored in a tank (~ 3000 L), after which it was fed to the crystallizer. The storage tank was stirred thoroughly to determine the concentrations of P O 4 - P and N H 4 - N and M g before starting the crystallizer. In the formation of pure struvite, equimolar amounts of ammonia, phosphate magnesium are removed. Thus, to attain P removal, phosphate ions should be limiting. Since, the ammonium nitrogen and magnesium concentrations in the sidestream supernatant obtained were low when compared to the P O 4 - P ; it was decided to supplement the supernatant with ammonium chloride and magnesium chloride. Both ammonium chloride and magnesium chloride were mixed into the supernatant at the inlet of the reactor to provide 70 RESULTS A N D DISCUSSIONS Mg and ammonium ions at the concentration calculated by the Potts model, for optimum P removal. Table 4.3 summarizes the operational conditions and results obtained in the two runs. 4.7.1 P removal and recovery efficiency For the first run, the removal efficiency varied from 39 to 61 %. This removal efficiency was calculated by comparing the theoretical mass of struvite grown, based on the amount of phosphate removed, against the amount harvested (Table 4.4). For the first run, the recovered struvite efficiency was very low (around 30 %). This was due to a supersaturation ratio that was greater than required. It was observed that the effluent in the clarifier was turning cloudy, which was an indication that nucleation was occurring, but there was no agglomeration. Most of the struvite was deposited as fines in the clarifier and in the hopper, and these were difficult to recover. Some fine struvite escaped the reactor and settled in the bottom of the reactor; these losses were not taken in to account. Also, some losses invariantly occurred during the recovery process, from harvesting to sieving. Thus in reality, the actual mass of struvite produced was higher than that harvested, when these losses are taken into consideration. In the second run, the removal efficiency of P O 4 - P varied from 46 to 57 %. The pH decreased from 8.4 to 8.2, when the effluent in the clarifier turned cloudy. The struvite recovery efficiency was 50 %, whereas a recovery efficiency of over 80 % has been reported during previous studies (Parvez, 2004; Adnan, 2002). 71 Table 4.3. Results of MAP crystallizer runs. RUN Recycle ratio SSR pH Effluent (mg/L) Influent (mg/L) P removal % Theoretical molar ratio Mg ! NH 4 -N ! P O 4 - P Mg N H 4 - N P O 4 - P N : P Mg:P I 1:3 5 8.7 35.1 17.6 25.6 112 28 59.9 57.26 1.03 2.38 I 1:3 5 8.6 33.5 17.2 36.2 112 28 59.9 39.56 1.03 2.38 I 1:3 5 8.6 34 19 33.1 112 28 59.9 44.74 1.03 2.38 I T 1:3 5 8.7 44 21.9 28 112 28 59.9 53.25 1.03 2.38 I 1:3 5 8.7 47 21.9 26.3 112 28 59.9 56.09 1.03 2.38 I 1:3 5 8.7 42.5 20.8 29.4 112 28 59.9 50.91 1.03 2.38 I 1:3 5 8.7 42.5 20.1 25.1 112 28 59.9 58.09 1.03 2.38 I 1:3 5 8.7 45.5 21.4 23.3 112 28 59.9 61.10 1.03 2.38 II 1:4 4 8.4 38 29.4 31.8 60 40 75 57.60 1.18 1.02 II 1:4 4 8.4 34.7 30.7 32.8 60 40 75 56.27 1.18 1.02 II 1:4 4 8.2 39.3 30.2 39.3- 60 40 75 47.60 - 1.18 1.02 II 1:4 4 8.2 37.9 30.9 40.3 60 40 75 46.27 1.18 1.02 II 1:4 4 8.2 40 30.9 36.8 60 40 75 50.93 1.18 1.02 II 1:4 4 8.2 30.7 15.8 33.2 60 40 75 55.73 1.18 1.02 Supersaturation ratio as computed by Potts model. ' Required concentrations were computed based on the Potts model RESULTS A N D DISCUSSIONS Table 4.4. Struvite recovery. Run I Run II Initial Seed weight 612 g 1091 g Final weight 915 g 1752 g Struvite production (Potts model prediction) 1059 g @ 80 % removal rate 1335 g @ 76 % removal rate Struvite production obtained 303 g 661 g Recovery rate 30% 50% The low removal efficiency may be attributed to a number of factors, including the difficulty in maintaining the flow and desired supersaturation ratio (SSR). Operating the M A P crystallizer online, instead of batch processing, would avoid the flow variation by maintaining the constant water head in the feed tank. This would help in maintaining the desired SSR at the inlet. Since the sidestream flow was very low, the batch processing lasted only around 3 days. Prolonged runs should yield better performance, as experienced by previous studies (Frederic Koch, Research Associate, Department of Civi l Engineering, U B C , pers. Comm..). 4.7.2 N removal In this study, upto 30 % ammonia removal was achieved in both runs. This was expected to be lower than the P removal, because ammonia, magnesium and phosphate should be removed in equimolar amounts. As discussed earlier, ammonium chloride was added to provide ammonium ions to optimize the P removal. In principle, the addition of ammonium chloride should be avoided. The best alternative is to mix this supernatant with the centrate obtained from the anaerobic digestion of return sludge, which is usually high in ammonia. This would enhance the recovery /removal of ammonia. Also, i f desired, the process also could be engineered for better ammonia removal by keeping the ammonium ions as the limiting one. 73 RESULTS A N D DISCUSSIONS 4.8 Comparison with the Simulation Results Simulation, using the ASM2 bio- p model, was carried out to understand the consequences of implementing the sidestream process at the pilot plant. In general, the values obtained for phosphate release in the P R U was little lower when compared to the actual release. Simulation results predicted maximum of 35-40 mg/L of P O 4 - P release in P R U and at the plant more than 70 mg/L of P O 4 - P release was obtained. However, the ammonia values obtained in the sidestream process were close to the model prediction. Simulation predicted around 9 to 10 mg/L of N H 4 - N in PRU, where as at the pilot plant around 10 to 12 mg/L of N H 4 - N was measured in the PRU. The simulation process served as an excellent tool in achieving the objectives set in the beginning of the study. Other factors, such as process failure predicted by the model, were not tested at the plant. 74 CONCLUSIONS A N D R E C O M M E N D A T I O N S CHAPTER 5 - CONCLUSIONS AND RECOMMENDATIONS The sidestream operation to recover phosphorus was successfully implemented at a pilot scale. Although, the recovery efficiency obtained was not very high, the refined conditions as suggested should yield a better performance. The ASM2 bio-P model was useful in predicting the most suitable conditions to operate the sidestream process. The objectives formulated in the beginning of the study were met. The following conclusions and recommendations were drawn from this investigation. 5.1 Conclusions 1. The anoxic zone was more suitable for taking a sidestream flow when taken individually. When taken in combination, anaerobic/anoxic zone yielded better conditions for struvite formation. The hydraulic retention time required in the phosphate release unit was two hours. 2. Simulation results revealed that the EBPR process will collapse under high sidestream flow rate (Qss> 50 % of Qjn) i f the sidestream return flow (QSSR) is not maintained. At higher sidestream flows, it was found that there is not enough phosphate left in the main process for biomass to be sustained. The introduction of return flow helps the process to regain its stability. 3. The sidestream return flow also influenced the anaerobic phosphate release and effluent nitrates level. With the increased return flow, there was increase in the phosphate release and a reduction in the effluent nitrate level, due to the extra acetate entering from the sidestream process. 4. The ASM2 bio-P model predicted 78 % of the incoming phosphorus could be diverted for recovery as struvite, using a sidestream process. 75 CONCLUSIONS A N D R E C O M M E N D A T I O N S 5. The sidestream process was successfully implemented at the pilot scale for both the M E B P R and CEBPR processes. Phosphorus and nitrogen removal performances were not compromised with the introduction of a sidestream process. Average removal efficiencies of more than 90 % were achieved in both the processes. 6. The results obtained at the plant, and then verified by the ASM2 bio-P model, indicate that the PAOs concentration in the EBPR process can be selectively increased by wasting from the sidestream clarifier, to control the SRT of the process. By doing this, there was a large increase in the phosphate release in anaerobic zone and corresponding uptake in the anoxic and aerobic zone. 7. Sidestream wasting resulted in reduced sludge wasting from the bioreactor. The average TP and T K N value for the sidestream sludge was 40.2 mg P/L and 74.1 mg N/L , when compared to the aerobic sludge which had average values of 195 mg P/L and 380 mg N/L . 8. Magnesium addition to the influent increased the phosphate and ammonium in the effluent, possibly due to potassium limitation in the influent. The process re-gained its stability after the addition was stopped. An influent Mg/P and K/P ratio of 0.34 and 2.85 respectively, was found to be sufficient to operate an E B P R process successfully under pilot plant conditions. 9. The ratio of V F A consumption per unit mass of P removed was decreased with a sidestream process, due to the additional phosphate being removed (recovered) as struvite. The V F A consumption in the sidestream process was much less, in comparison to the anaerobic zone of the EBPR process, with a higher release of phosphate. 76 CONCLUSIONS A N D R E C O M M E N D A T I O N S 10. The average P removal efficiency of 50 % was achieved with sidestream supernatant, using a M A P crystallizer and 50 % of the phosphorus removed was recovered as struvite pellets. 11. The ASM2 bio-P model served as an excellent research tool and facilitated discussion and understanding of the process. 12. It was observed that anoxic phosphate uptake was more predominant in the EBPR process. As confirmed earlier, the CEBPR process exhibited a better denitrification due to the anoxic sludge blanket in the secondary clarifier. It was also observed that the M E B P R based process was more resistant to shock loads when compared to the CEBPR process. 5.2 Recommendations 1. A sidestream process could be extended to include a sludge busting operation, which will result in a reduction of excess sludge generation (Saktaywin et al, 2005). 2. A sidestream sludge wasted is rich in carbon and can be used to recover carbon. 3. Sidestream supernatant does not contain a high concentration of ammonia. This phosphate rich sidestream could be combined with the ammonium rich streams obtained from return sludge digestion, for the recovery of phosphates as struvite (Fig. 5.1). 77 CONCLUSIONS A N D R E C O M M E N D A T I O N S P Release Tank Clarifier Sidestream Supernatant To Crystallizer T • Wasting Supernatant Figure 5.1. Mixing of sidestream supernatant with anaerobic digester supernatant 4. Future efforts must concentrate on duplicating these efforts, refining them to recover most of the P from the sidestream process. Longer duration runs or a continuous online system should yield better results. 5. A detailed study on the effects of adding magnesium to the EBPR process can be studied. This would also benefit the sidestream process, with an increase in Mg concentration in the supernatant. 6. Different configurations and operating conditions can be tested at the pilot plant, such as running a combined sidestream process and operating at higher sidestream flow. 7. A microbial community study can be carried out, to confirm the selective increase of PAOs during the sidestream wasting. 78 REFERENCES REFERENCES Adam, C , Gnirss, R.,Lesjean, B., Buisson, H., Kraume, M . (2002) Enhanced biological phosphorus removal in membrane bioreactors. Water science and Technology 46(4-5), 281-286. Adnan, A. (2002) Pilot-scale Study of Phosphorus Recovery Through Struvite Crystallization. M.A.Sc Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, B.C. Canada. AI-Atar, E. (2006) Dynamic Modeling and Optimization Membrane Enhanced Biological Phosphorus Removal Process for Municipal Wastewater Treatment. Ph.D. Thesis, University of British Columbia, Vancouver, B.C. Canada. In preparation APHA-AWWA-WEF (1998) Standard methods for the examination of water and wastewater. 20 th Ed., Washington, D.C. Arvin E. (1985) Biological removal of phosphorus from wastewater. CRC Crit. Rev. in Env. Control, 15 (1), 25-64. Barker, P.S., Dold, P.L. (1997) General model for biological nutrient removal activated-sludge systems: model presentation. Water Environ. Res. 69, 969-984. Ben Aim, R.M., Semmens, M.J. (2003) Membrane bioreactors for wastewater treatment and reuse: a success story. Water Science and Technology 47 (1), 1-5. Borgerding, J. (1972) Phosphate deposits in digestion systems. Journal of the Wastewater Pollution Control Federation. 44, 813-819. Bortone, G., Saltarelli, R., Alonso, V., Sorm, R., Wanner, J., Tichle, A. (1996) Biological anoxic phosphorous removal- The dephanox process. Water Science and Technology 34 (1-2), 119-129. Brdjanovic, D., van Loosdrecht, M.C.C., Versteeg, P., Hooijmans, C M . , Alaerts, G.J., Heijnen, J.J (2000) Modeling COD, N and P removal in a full-scale WWTP Haarlem Waarderpolder. Water research 34 (3), 846-858 Brdjanovic, D., Hooijmans, C. M. , van Loosdrecht, M . C. M . , Alaerts, G. J., Heijnen, J. J. (1996/10) The dynamic effects of potassium limitation on biological phosphorus removal. Water Research 30(10), 2323-2328. Britton, A., Koch, F.A., Mavinic, D.S., Adnan, A., Oldham, W.K., Udala, B. (2005) Pilot scale struvite recovery from anaerobic digester supernatant at an enhanced biological phosphrous removal wastewater treatment plant. Journal of Environmental and Engineering Science 4, 265-277. 79 REFERENCES Britton, A.T. (2002) Pilot-scale struvite recovery trials from a full scale digester supernatant at the city of Penticton advanced wastewater treatment plant. M.A.Sc. Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, B.C., Canada. Burns, J., Finlayson, B. (1982) Solubility product of magnesium ammonium phosphate hexahydrate at various temperatures. The Journal of Urology. 128, 426-428. Comeau, Y. , Hall, K.J., Hancock, R.E.V., Oldham, W.K. (1986) Biochemical model for enhanced biological phosphorus removal. Water Research 20(12), 1,511-1521. Cote, P., Buisson, H., Praderie, M.(1998) Immersed membranes activated sludge process applied to the treatment of municipal wastewater. Water Science and Technology 38 (4-5), 437-442. Dastur, M.B. (2001) Investigation into the Factors Affecting Controlled Struvite Crystallization at the Bench-scale. M.A.Sc. Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, B.C., Canada. Doyle, J.D., Parsons, S.A. (2002/9) Struvite formation, control and recovery. Water Research 36 (16), 3925-3940. Eckenfelder W W, Jr. (2000) Industrial Water Pollution Control, 3 r d Ed. McGraw-Hill, New York. Fattah, K.P. (2004) Pilot-scale Struvite Recovery Potential from Centrate at Lulu Island Wastewater Treatment Plant. M.A.Sc Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, B.C. Canada. Filipe, C.D.M., Daigger, G.T. (1998) Development of a revised metabolic model for the growth of phosphorus-accumulating organisms. Water Environmental Research 70, 67-79. Fleischer, E.J., Broderick, T.A., Daigger, G.T., Fonseca, A.D., Holbrook, R.D., Murthy, S.N. (2005) Evaluation of membrane bioreactor process capabilities to meet stringent effluent nuturient discharge requirements. Water Environment Research 11 (2), 162-178. Gaterell, M.R., Gay, R., Wilson, R., Gochin, R.J., Lester, J.N., (2000) An economic and environmental evaluation of the opportunities for substituting phosphorus recovered from wastewater treatment works in existing U K fertilizer markets. Enviromental Technology 21, 1067-1084. Giesen, A. (1999) Crystallization process enables environmental friendly phosphate removal at low costs. Environmental Technology. 20, 769-775 Siegrist, H., Brunner, I., Koch, G., Linh Con Phan., Van Chieu Le (1999) Reduction of biomass decay rate under anoxic and anaerobic conditions. Water Science and Technology 39 (1), 129-137. 80 REFERENCES Hao, X. , van Loosdrecht, M . C. M. , Meijer, S. C. F., Heijnen, J. J., Qian, Y. (2001) Model-Based Evaluation of Denitrifying P Removal in a Two-Sludge System. Journal of Environmental Engineering 127 (2), 112-118. Hillenbrandt T., Bohm E., Bartl, J. (1999) Evaluation of the expanded Phostrip process and comparision with other methods for phosphorus removal. Wasser Abwasser, 140 (2), 86-92 Hu, Z., Wentzel, M.C., Ekama, G.A. (2002/11) Anoxic growth of phosphate-accumulating organisms (PAOs) in biological nutrient removal activated sludge systems. Water Research 36 (19), 4927-4937. Jaffer, Y., Clark, T.A., Pearce, P., Parsons, S. A. (2002/4) Potential phosphorus recovery by struvite formation. Water Research 36 (7), 1834-1842. Jardin, N. , Popel, H.J. (1994) Phosphate release of sludges from enhanced biological P-removal during digestion. Water Technology. 30, 281-292. Jardin, N . , Popel, J.J. (1996) Behaviour of waste activated sludge from enhanced biological phosphorus removal during sludge treatment. Water Environment Research: 68,965-973. Jeanmaire, N . (2001) Recycling of removed phosphorus: analysis of the potential interest in wastewater treatment plants. Centre European d'Etudes des Polyphosphates (CEEP). World wide web http://www.nhm.ac.uk/mineralogy/phos/Jeanmaire_report.doc Jeanmaire, N. , Evans, T. (2001) Technico-economic feasibility of P-recovery from municipal wastewaters. Environmental Technology 22, 1355-1361. Klapwijk B., Rulkens W., Temmink H. (2001) Recovery of phosphorus from sewage, 2nd Int. Conf. on Recovery of Phosphates from Sewage and Animal Wastes, Holland, NL, 12-13. Kuba, T., Van Loosdrecht, M.C.M., Brandse, F.A., Heijnen, J.J. (1997/4) Occurrence of denitrifying phosphorus removing bacteria in modified UCT-type wastewater treatment plants. Water Research 31(4), 777-786. Kuba, T., Van Loosdrecht, M.C.M., Heijnen, J.J. (1996) Phosphorus and nitrogen removal with minimal cod requirement by integration of denitrifying dephosphatation and nitrification in a two-sludge system. Water Research 30 (7), 1702-1710. Kuroda, A., Takiguchi, N. , Gotanda, T., Nomura, K., Kato, J., Ikeda, T., Ohtake, H. (2002) A simple method to release polyphosphate from activated sludge for phosphorus reuse and recycling. Biotechnology andBioengineering 78 (3), 333-338. Lesjean, B., Gnirsss, R., Adam, C , Kraume, M. , Luck, F. (2003) Enhanced biological phosphorus removal process implemented in membrane bioreactors to improve phosphorous recovery and recycling. Water Science and Technology 48 (1), 87-94. 81 REFERENCES Liao, P.H., Mavinic, D.S., Koch, F.A. (2003) Release of phosphorus from biological nutrient removal sludges: A study of sludge pretreatment methods to optimize phosphorus release for subsequent recovery purposes. Journal of Environmental Engineering (2) 369-381 Manem, J., Sanderson, R. (1996) Membrane bioreactors. In: Water Treatment membrane processes (Malleviable, J., Odenaal, P., and Weisner,M.R., Eds.j.McGraw-Hill,NewYork,17.1-17.31. Maurer, M . , Gujer, W., Hany, R., Bachmann, S. (1997) Intracellular carbon flow in phosphorus accumulating organisms from activated sludge systems. Water Research 31 (4), 907-917 Mavinic, D.S., Koch, F.A., Hall, E.R., Abraham, K., Niedbala, D.(1998) Anaerobic co-digestion of combined sludges from a BNR wastewater treatment plant. Environmental Technology 19, 35-44. Meijer, S. C. F., van Loosdrecht, M . C. M. , Heijnen, J. J. (2001/8) Metabolic modelling of full-scale biological nitrogen and phosphorus removing wwtp's. Water Research 35 (11), 2711-2723. Meijer, S.C.F. (2006) Theoretical and Practical Aspects of Modeling Activated Sludge Processes. Ph.D. Thesis, Delft University of Technology, The Netherlands. Mino, T., Tsuzuki, Y. , Matsuo, T. (1987) Effect of phosphorus accumulation of acetate metabolism in the biological phosphorus removal process. Proc. IAWPRCInt. Conf. on Biological Phosphate Removal from Wastewaters, Rome, Adv. Water Pollution Control, ed., R. Ramadori. 27-38. Miya, A., Kitagawa, M. , Tanaka, T. (1987) Behavior of magnesium in biological phosphate removal. Proc. IAWPRC Int. Conf. on Biological Phosphate Removal from Wastewaters, Rome, Adv. Water Pollution Control, ed., R Ramadori. 135-146. Momberg, G., Oellerman, R. (1992) The removal of phosphate by hydroxyapatite and struvite crystallization in South Africa-. Water Science and Technology 26, 987-996. Munch, E., Barr, K. (2001) Controlled struvite crystallization for removing phosphorus and anaerobic digester sidestreams. Water Research 35,151-159. Murnleitner, E., Kuba, T., van Loosdrecht, M.C.M., Heijnen, J.J. (1997) An integrated metabolic model for the aerobic and denitrifying biological phosphorus removal. Biotechnology and Bioenggineering 54 (5), 434-450. Ohlinger, K., Young, T., Schroeder, E. (2000) Postdigestion struvite precipitation using a fluidized bed reactor. Journal of Environmental Engineering 126, 361-368. Ohlinger, K., Young, T., Schroeder, E. (1998) Predicting struvite formation in digestion. Water Research 32, 3607-3614. 82 REFERENCES Pereira, H., Lemos, P.C., Reis, M.A.M. , Crespo, J.P.S.G., Carrondo, M.J.T., Santosh, H.(1996) Model for carbon metabolism in biological phosphorus removal processes based on in vivo C-13-NMR labeling experiments. Water Research 30(9),2128-2138. Rabinowitz, B., Oldham, W.K. (1986) Excess biological phosphorus removal in the activated sludge process using primary sludge fermentation. Canadian Journal of Civil Engineering 13 (3), 345-351. Reichert, P. (1994) Aquasim- A tool for simulation and data analysis of aquatic systems. Water Science and Technology 30 (2), 21-30. Rensink, J.H., van der Ven, J., van Pamelen, G., Fedder, F., Majoor E. (1997). The modified Renphosystem: A high biological nutrient removal system. Water Science and Technology 35 (10), 137-146. Rickard, L.F., McClintock, S.A. (1992) Potassium and magnesium requirements for enhanced biological phosphorus removal from wastewater. Water Science and Technology 26 (9), 2203-2206 Rieger, L., Koch, G., Kuhni, M. , Gujer, W., Siegrist, H. (2001) The EAWAG bio-p module for activated sludge model no. 3. Water Research 35 (16), 3887-3903. Rittmann, B.E. (1998) Opportunities with membrane bioreactors. Proc. Microfdtration Conference. San Diego, CA: National Water Research Institute, Nov. 12-13. Rosenberger, S., Kruger, U., Witzig, R., Manz, W., Szewzyk, U. , Kraume, M . (2002) Performance of a bioreactor with submerged membranes for aerobic treatment of municipal waste water. Water Research 36 (2), 413-420. Sakai, Y. , Fukase, T., Yasui, H., Shibata, M . (1997) An activated sludge process without excess sludge production. Water Science and Technology 36 (11),163-170. Saktaywin, W., Tsuno, H., Nagare, H., Soyama, T., Weerapakkaroon, J. (2005) Advanced sewage treatment process with excess sludge reduction and phosphorus recovery. Water Research 39 (5), 902-910. Schuiling, R., Andrade, A. (1999) Recovery of struvite from calf manure. Environmental Technology. 20, 765-768. Seckler, M . M. , Bruinsma, O. S. L., Van Rosmalen, G. M . (1996) Calcium phosphate precipitation in a fluidized bed in relation to process conditions: A black box approach. Water Research 30 (7), 1677-1685. Siegrist, H., Gujer, W. (1994) Nitrogen removal in activated sludge systems including denitrification in secondary clarifiers. Water Science and Technology 30 (6), 101-111. Smolders, G.J.F., Van Loosdrecht, M.C.M., Heijnen, J.J. (1996) Steady-state analysis to evaluate the phosphate removal capacity and acetate requirement of biological phosphorus removing mainstream and sidestream process configurations. Water Research 30 (11), 2748-2760. 83 REFERENCES Smolders, G.J.F., Klop, J.M., van Loosdrecht, M.C.M., Heijnen, J.J. (1995c) A metabolic model of the biological phosphorus removal process . 1. Effect of the sludge retention time. Biotechnology andBioengineering 48 (3), 222-233. Smolders, G.J.F., Vandermeij, J., van Loosdrecht, M.C.M., Heijnen, J.J. (1994a) Model of the anaerobic metabolism of the biological phosphorus removal process -stoichiometry and pH influence. Biotechnology and Bioengineering 43 (6), 461-470. Smolders, G.J.F., Vandermeij, J., van Loosdrecht, M.C.M., Heijnen, J.J; (1994b) Stoichiometric model of the Aerobic metabolism of the biological phosphorus removal process. Biotechnology and Bioengineering 47 (3), 277-287. Snoeyink, V., Jenkins, D. (1980) Water Chemistry, John Wiley & Sons, New York. Srinath, E.G., Sastry, C.A., Pillai, S.C. (1959) Rapid removal of phosphorus from sewage by activated sludge. Experienta 15 339-340. Stratful, I., Brett, S., Scrimshaw, M.B., Lester, J.N. (1999) Biological phosphorus removal, its role in phosphorus recycling. Enviromental Technology 20, 681-695. Tchobanoglous, G., Burton, F.L., Stensel, H.D. (2003) Wastewater Engineering: Treatment and Reuse, 4 t h Ed. McGraw-Hill Higher Education, New York. Van der Roest, H.F., van Bentem, A.G.N., Lawrence, D.P. (2002) MBR-technology in municipal wastewater treatment: challenging the traditional treatment technologies. Water Science and Technology 46 (4-5), 273-280. van Veldhuizen, H.M., van Loosdrecht, M.C.M., Heijnen, J.J. (1999) Modelling biological phosphorus and nitrogen removal in a full scale activated sludge process. Water Research 33 (16), 3459-3468. Wentzel, M.C., Dold, P.L., Ekama, G.A., Marais, G.V.R. (1989a) Enhanced polyphosphate organism cultures in activated sludge systems. Part II: Experimental behavior. Water South Africa 15 (2), 71-87. Wentzel, M.C., Dold, P.L., Ekama, G.A., Marais, G.V.R. (1989b) Enhanced polyphosphate organism cultures in activated sludge systems. Part III: Kinetic model. Water South Africa 15 (2), 89-102. Wentzel, M.C., Ekama, G.A. (1997) Principles in the design of single-sludge activated-sludge systems for biological removal of carbon, nitrogen, and phosphorus. Water Environmental Research 69, 1222-1231. Wentzel, M.C., Lotter, L.H., Loewenthal, R.E., Marais, G.v.R. (1986) Biochemical behavior of Acinetobacter spp. In enhanced biological phosphate removal - a biochemical model. Water South Africa 12 (4), 209-244. Wichern M . , Obenaus F., Rosenwinkel K.-H (1999) Comparison of the models ASM2d and TU-Delft in simulation praxis, 6. Simba-anwendertreffen, Magdenburg. 84 REFERENCES Williams, S. (1999) Struvite precipitation in the sludge stream at slough wastewater treatment plant and opportunities for phosphorus recovery. Environmental Technology 20,743-747. Woods, N.C., Sock, S.M., Daigger, G.T (1999) Phosphorus recovery technology modeling and feasibility evaluation for municipal wastewater treatment plants. Environmental Technology 20 (7), 663-679. Wu, Q., Bishop, P. L., Keener, T. C. (2006) Biological Phosphate Uptake and Release: Effect of pH and Magnesium Ions. Water Environment Research 78 (2), 196-201. Yamamoto, K., Hiasa, M . , Mahmood, T., Matsuo, T. (1989) Direct solid-liquid separation using hollow filter membrane in an activated sludge aeration tank. Water Science and Technology 21 (4-5), 43-54. Yasui, H., Nakamura, K , Sakuma, S., Iwasaki, M . , Sakai, Y . (1996) A full scale operation of a novel activated sludge process without excess sludge production. Water Science and Technology 34 (3-4), 395-404. 85 APPENDIX A APPENDIX A: ASM2-DELFT BIO-P M O D E L MODEL COMPOUNDS AND PROCESSES MODEL STOICHIOMETRY AND KINETICS 86 APPENDIX A Description of the model compounds In the model, soluble components are indicated with S?, particulate components with X?. Only soluble components carry an ionic charge. The X? fractions are assumed to be associated with the activated sludge, whereas S? fractions can only be transported with the water phase. Soluble (S?) compounds S 0 Dissolved oxygen (g02"m"3). Dissolved oxygen can directly be measured and is subject to gas exchange. On the COD balance oxygen accounts for - l g COD-g 1 0 2 . SF Readily biodegradable fermenTable organic substrate (gCOD'm"3). This fraction of the biodegradable substrate is readily available for heterotrophic organisms. SF is converted to SA in the fermentation process. SF is converted to SA in the fermentation process. SF does not have an ionic charge. SA Volatile fatty acids, VFA (gCOD'm3). VFA can directly be measured. It is assumed all lower fatty acids (C2 to C 5 ) are available as SA- The ionic charge is -1/64. The TOC / COD ratio of S A is 0.4 gC/gCOD. In the bio-P process S A is stored under anaerobic conditions as PHA. SNH Ammonium plus ammonia (NH 3 + N H 4 + , gN.m"3). Ammonium can directly be measured. For the balance of the ionic charge (+1 / 14 mole) SNH is assumed to be all NH4 + . SNO Nitrate plus nitrite (NO3" + NO2", gN.m"3). For all stoichiometric calculations SNO is assumed to be N O 3 " only. The charge of SNO is equal to -1 / 14 mole. On the COD balance nitrate accounts for -2.86 gCOD* g"'N. 87 APPENDIX A SN2 Di-nitrogen (gNui") formed in the denitrification process. In the model di-nitrogen accumulates in the liquid and can be used as a measure for the total denitrified load. Spo Inorganic soluble phosphorus, primary ortho phosphate (gP'm"3). For the balance of the ionic charge it is assumed that Spo consists of 50% H 2 P O 4 " and 50% H P O 4 2 " and therefore has an electric charge of 1.5/31 mole. Si Inert soluble organic material (gCODm"3). Si is assumed to be non biodegradable. Generally Si is a part of the influent. As in the default of ASM2d, in the TUDP model no Si is formed in the hydrolysis process (fsi=0). In low loaded activated sludge plants the effluent soluble organic fraction (SA+SF+SI) largely consists of Si. SHCO Alkalinity of the wastewater mole (HCCVin"3). Alkalinity is used to approximate the conservation of ionic charge (equal to -1 mole) in biological reactions. Alkalinity is introduced in order to predict possible low pH conditions, which might inhibit some biological processes. Particulate (X?) compounds Xi Inert particulate organic material (gCOD'm3). Under normal operational conditions this organic fraction is not biodegraded. Xi may be a part of the influent. In the process Xi flocculates into the activated sludge. Also Xi may be formed in the process from lysis of heterotrohic and autotrophic organisms. Xi is assumed to contain a fraction ammonium (iNxi) and phosphate (iPxi). X s Slowly biodegradable particulate substrate (gCOD'm"3). This substrate must undergo cell external hydrolysis to SF before being available as substrate. Xs may be part of the influent whereas it also may be formed in the process from 88 A P P E N D I X A lysis of heterotrophic and autotrohpic organisms. Xs is assumed to contain a fraction ammonium (iNxs) and phosphate (iPxs)-XH Heterotrophic organisms (gCODTn"3). These organisms represent the general heterotrophic population. XH grows aerobically, whereas most a large fraction of XH also can use nitrate as an electron acceptor (i.e., denitrification). XH grows on both SA and SF- NO cell-internal storage is modeled. XH is assumed to contain a fraction ammonium (INBM) and phosphate (IPBM)- These fractions are equal for all biomass fractions (i.e., XH, XPAO and XA). XPAO Phosphate accumulating organisms (gCOD'm3). PAO's represent the hetrotrohpic population that under anaerobic conditions stores Poly-Hydroxyalcanoates (XPHA)- Under denitrifying and aerobic conditions PAO's use PHA as substrate for growth while storing polyphosphate (X P P ) and glycogen (XGLY)- A l l processes related to PAO's are based on metabolic descriptions of the cell internal processes. The state of reduction of XPAO is assumed to be 0.334 gOgCOD"1. Xpp Poly-phosphate (gP*m"3). This is a cell-internal storage product of PAO's. It is assumed only to be associated with PAO's, however not included in the latter's mass. It forms part of the particulate phosphorus and may be analytically observed. For stoichiometric considerations, poly-phosphate is assumed to have the composition of (Ko.33Mgo.33P03)n. Since the model does not account for K + and M g 2 + , poly phosphate is represented as ( H P O 3 V An electric charges of-1/31 must be included to compensate for this term. XPHA Cell-internal storage of poly-hydroxyalkanoates in PAO's (gCOD-m"). This storage compound is only associated with PAO's. it is however not included in the mass of PAO's. X P H A can not be directly be related to analytically 89 APPENDIX A measured PHA values, as also non PAO's are known to accumulate PHA. Therefore XPHA is only a functional component required for modeling biological phosphorus removal. For stoichiometric considerations, XPHA is assumed to have the composition of Poly-Hydroxy-Buterate (C^HeChV The state of reduction of X P HA is assumed to be 0.334 gC'gCOD"1. XGLY Cell-internal storage of glycogen in PAO's (gCOD*m3). This storage compound is only associated with PAO's. XGLY can not be directly be related to analytically measured glycogen values, as also non PAO's are known to accumulate glycogen. Therefore XGLY is only a functional component required for modeling biological phosphorus removal. For stoichiometric considerations, XGLY is assumed to have the composition of (C6HioOs)n. The state of reduction of XGLY is assumed to be 0.375 gOgCOD"1. XA Autotrophic nitrifying . organisms (gCOD'm"3). Nitrifying organisms are responsible for nitrification. They are obligate aerobic, chemolitho-autotrohpic. It is assumed that nitrifiers oxidase ammonium (SNH) directly to nitrate (SNO)- Nitrite ( N O 2 ) as intermediate compound is not considered in the model. XTSS Total Suspended Solids (g*m~3). TSS is an analytical measurement typically followed on a day to day basis by plant operators. XTSS is calculated in the model from the model stoichiometry. It is not a balanced compound and therefore relies on the TSS fractions (ijss) of the particulate compounds (X?) and XTSS in the influent. Description of the Modelled Processes (1-3) Hydrolysis (rH N , r° , r„ 0 ,gCOD x s -d~') . In these three process the particulate bio-degradabel substrate Xs is converted to soluble substrate (SF). The 90 APPENDIX A hydrolysis process is dependent on the electron donor (oxygen or nitrate). Under anoxic and anaerobic conditions the hydrolysis rate is reduced according to the reduction factors r|NO and T|fe. In the TUDP model hydrolysis is a function of the total heterotrophic population (XH + XPAO)-(4) Aerobic heterotrohic growth ( r^gCODj^ 'd - 1 ) on readily biodegradable fermenTable organic substrate (SF). In presence of oxygen XH grows on SF-The formation of storage products (PHA) by regular heterotrohps (XH) is not modeled. (5) Aerobic heterotrohpic growth (rs° , gCOD X H «d - 1 ) on Volatile Fatty Acids (SA). In presence of oxygen (So) as electron acceptor XH grows on VFA (SA). The formation of storage products (PHA) by regular heterotrophs is not modelled. (6) Anoxic heterotrophic growth (r^0, gCOD X H -d- ') on readily biodegradable fermenTable organic substrate (SF). The same process as aerobic growth (process 5) except nitrate (SNO) serves as the main electron acceptor. It is not clear weather all heterotrophs are capable of denitrification. Experimentally a reduced anoxic activity is observed compared to aerobic conditions. This is expressed with a reduced factor (T|NO) in the kinetic rate equation. (7) Anoxic heterotrohpic growth (rs^°, gCODX H«d"') on Volatile Fatty Acids (SA)-The same as for aerobic growth (process 6) except nitrate serves as the main denitrification. Experimentally a reduced anoxic activity is observed compared to aerobic conditions. This is expressed with a reduction factor (T|NO) in the kinetic rate equation. (8) Anaerobic fermentation ( r£ N ,gCOD S F ' d - 1 ) . Anaerobic formation of Volatile Fatty Acids (SF). The process biodegradable fermenTable organic substrate 91 APPENDIX A (SF). The process is induced by regular heterotrohpic organisms (XH). In most plants anaerobic conditions are relative short. Therefore in the degradation of particulate substrate to VFA (XS-»SF—»SA), anaerobic Fermentation (SF-»SA) is often the limiting process. (9) Heterotrophic Lysis ( r H L ,gCOD X H «d _ 1 ) . Lysis of regular heterotrophs (XH) to particulate substrate (Xs) and particulate inert material (Xi). In the model the concepts of lysis (processes 9 and 21) is used for the regular heterotrophs, whereas the maintenance concept (processes 11 and 15) is used for phosphate accumualtign bacteria (XPAO)-(10) Anaerobic uptake of Volatile Fatty Acids by PAO's ( r ^ . gCOD^-d - 1 ) . VFA is taken up under anaerobic conditions and stored as PHA (XPHA)- The energy needed for this process is provided from the conversion of cell-internally stored glycogen (XGLY) and poly-phosphate (X P P ) . Hereby, PAO's release large amounts of orthophosphate (SPO). The Yield of S A taken up and X P H A formed (YPHA) is a function of the pH. (11) Anaerobic maintenance (r* N ,gP«d _ 1 ). In absence of an electron acceptor cell-internal stored poly-phosphate (X P P ) is converted to ortho-phosphate (Spo) yielding in energy that used for cell maintenance. (12) Anoxic uptake of Volatile Fatty Acids by PAO's (r s* N ,gCOD S A-d _ 1). VFA is taken up under anoxic conditions and stored as PHA (XPHA)- The energy needed for this process is provided from the conversion of cell-internally stored poly-phosphate (X P P ) , which is released by PAO's as ortho-phosphate (Spo). An glycogen is used in this process. 92 APPENDIX A (13) Aerobic consumption of PHA (r°H A,gCOD x s«d ' ) . Growth of PAO's on cell internally stored PHA (XPHA)- The net growth of PAO is assumed to be a function of the PHA degradation, poly-phosphate formation and growth. In the model however, XPHA is converted to XpAo> whereas glycogen (XGLY) and poly-phosphate (Xpp) are only formed from XPAO- AS a result of this mathematical description, processes 13, 14 and 15 can not be read independently. (14) Aerobic poly-phosphate formation (r^,,gP«d _ 1) . Ortho-phosphate (Spo) is aerobically taken up by XPAO and cell-internally stored as polyphosphate (Xpp). This process is much faster then nitrification and therefore rarely limiting in wwtp's. Under aerobic and anoxic conditions Xpp and XGLY are restored to provide the anaerobic uptake of substrate (SA). (15) Aerobic glycogen formation (r<?LY, gCOD X G L Y »d _ 1 ) . In presence of an electron acceptor (So), X P H A is partly converted to glycogen (XGLY)- Under aerobic and anoxic conditions Xpp and XGLY are restored to provide the anaerobic uptake of substrate (SA)-(16) Aerobic maintenance (r^, gCOD X P A O -d" ' ) . In the maintenance concept PHA is oxidized yielding in energy for cell maintenance. This is mathematically described by the oxidation of XPAO with oxygen (So)- The formation of inert material (XT or Si) from processes like decay, predation and lysis are reconciled in the model via the process yields. (17) Anoxic PHA degradation (rpNH°A, gCOD X P H A -d~') . Oxidation PHA ( X P H A ) with nitrate (XNO) as the main electron acceptor. It is not clear whether all PAO's are capable of denitrification. Experimentally a reduced denitrification rate is observed compared to aerobic conditions. In the model this is expressed with 93 APPENDIX A a reduction factor (T|NO) in the kinetic rate equation. As a result of the mathematical description of the metabolic model, processes 17, 18 and 19 can not be read independently. (18) Anoxic poly-phosphate formation (r p^°,gP«d _ 1) . Ortho-phosphate (SPO) is stored as poly-phosphate (Xpp) by XPAO with nitrate (SNO) as the main electron acceptor. As for the anoxic PHB degradation (process 16), the anoxic poly-phosphate formation rate is reduced with a factor TJNO compared to aerobic conditions (process 13). (19) Anoxic formation of glycogen (r^jgCODxQLy'd"1). In presence of nitrate (SNO), XPHA is partly converted to glycogen (XGLY). XGLY is restored to provide reduction equivalents and energy in the anaerobic uptake of substrate (SA). As for the anoxic PHB degradation (process 16), the anoxic glycogen formation is reduced with the a factor T|NO compared to aerobic conditions (process 14). (20) Anoxic maintenance (r^ 0 ,gCOD X P A O 'd _ l ) . Maintenance of PAO's analogue to aerobic maintenance (process 15) with nitrate (XNO) as the main electron acceptor. (21) Autotrophic growth (rA ,gCOD X A «d~') . Autotrophic growth of nitrifying organisms (XA) in which the energy yielding reaction is the oxidation of ammonium (XNH) to nitrate (XNO) with oxygen (So) as the main electron acceptor and C O 2 as the carbon source for biomass formation. (22) Autotrophic lysis (r^gCOD^'d"') . Lysis of autotrophic nitrifying organisms (XA) yielding in inert material (Xi) and particulate substrate (Xs). 94 APPENDIX A Model Stoichiometry The model stoichiometry is presented Table A . l . In the matrix, horixontally 22 processes are represented, the 18 modelled compounds are ordered vertically. The stoichiometric coefficients (C?,i..22, Table A.2a) are derived from the composition balances. The yields and maintenance factors of the metabolic model are derived from the P/O (8) ration being the amount ATP produced from N A D H 2 and the ATP use for maintenance (BIATP) (Tables A.4 and A.5b). These parameters therefore can not be calibrated individually. Model Kinetics The kinetic rate equations of the TUD model are presented in Table A.6. The kinetic parameters are listed in Tables A.5a/b. For the calibration of the glycogen formation rate we suggest to use the tree-step method as proposed in this chapter. Table A.7 gives the changes in the parameters as a result of calibration done at UBC pilot plant. 95 Table A . 1. Stoichiometric matrix and component composition matrix 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 t« <U Component —> So S F S A S N H S N O SN2 Spo s , S H C O X i Xs X H X p A 0 Xpp XpHA X G L Y X A X T S S <—Proc a o 60 S Q O U 60 S Q O O 60 S z 60 a z 60 a z 60 B PM 60 a Q O U 60 a "o a a Q O U 60 a a Q Q o o U U 60 60 a Q o u 60 a CM 60 a Q o u 60 a Q o U 60 a Q o O 60 a ~60 1 rh° Aerobic hydrolysis gCODxs/d 1-fsi CN,I CP.I fsi C e , i -1 C T S S J 2 Anoxic Hydrolysis gCODxs/d 1-fsi CN,I - CP.I fsi C e , i -1 C T S S J 3 AO r h Aanerobic Hydrolysis gCODxs/d 1-fsi CN,I - Cp.i fsi C e , i -1 C T S S J Regular Heterotrophic Organisms X H 4 ASF Aerobic Growth on SF gCODxH/d -(1/YH-1) - 1 / Y H CN,4 Cp,4 C e ,4 1 CTSS,4 5 rs°A Aerobic Growth on S A gCOD X H /d -(1/YH-1) - 1 / Y H CN,5 Cp,5 CE,5 1 CTSS,5 6 NO R SF Anoxic Growth on SF gCODxH/d - 1 / Y H CN,6 ( 1 / Y H - 1 ) 2.86 ( 1 / Y H - 1 ) 2.86 Cp,6 C E ,6 1 CTSS,6 7 NO F SA Anoxic Growth on S A gCOD X H /d - 1 / Y H CN,7 ( 1 / Y H - 1 ) 2.86 i ( l /Y„- l ) 2.86 Cp,7 CE >7 1 CTSS,7 8 _ AN r fe Fermentation gCODsp/d -1 1 CN,8 Cp,8 CE,8 CTSS,8 9 Heterotrophic lysis gCODxH/d CN,9 Cp,9 CE,9 fxi.H l-fxi,H "I CTSS,9 Phosphorus Accumulating organisms XPAO 10 AN 1 S A Anaerobic Storage of S A gCOD S A /d -1 V A N 1 PO C e,io V A N I P 0 •y AN * S A 1 -1 1 S A C T S S J 0 11 AN R M Anaerobic Maintenance gP/d 1 C e , l l -1 C T S S J I 12 NO R SA Anoxic Storage of S A gCOD S A /d -1 (1-Y S N A ° ) 2.86 ( 1 - Y S N A ° ) -2.86 V N O 1 PO Ce>12 V N O I P O V N O * S A C T S S J 2 13. NO PHA Anoxic PHA Consumption gCODpHA/d CN,13 ( 1 - 1 / Y ™ ) 2.86 (1-1/Y P N H° A ) 2.86 Cp,l3 Ce,13 1 / Y N 0 1 1 1 PHA -1 C T S S J 3 14. - N O lpp Anoxic storage of poly-P gP/d CN,14 (1/Y P N P°) 2.86 ( l / Y p N p ° ) 2.86 Cpj4 C e,14 - 1 / Y N 0 1/ lpp 1 C T S S J 4 15. NO GLY Anoxic Glycogen Formation gCODcLY/d CN,15 ( l / Y G T v - l ) 2.86 (l /Y G N L °v-l) 2.86 Cp,i5 Ce,15 - l / Y N O L / 1 GLY 1 C T S S J 5 O o s-l OH I Component —> 1 So O oo • 2 SF Q O U oo 3 S A Q O U oo 4 S N H Z 00 5 S N O Z 00 6 SN2 z oo 7 Spo OH 00 8 S, Q O O oo 9 S H C O a is a 10 11 12 13 x , X s X H XpAO S a a a gCOD/ gCOD/ gCOD/ gCOD/ 14 Xpp OH 00 15 16 17 18 X P H A X G L Y X A X T S S a Q O U oo Q O U 00 a o o u oo oo 16. r N 0 Anoxic Maintenance gCODPAO/d M 17. 18. 19. 20. L P H A i p p GLY Aerobic PHA Consumption Aerobic Storage of poly-P Aerobic Glycogen Formation r ° Aerobic Maintenance g C O D P H A / d gP/d gCODcLY/d gCODpAo/d i /Y° -1 1 1 1 P H A 1 -1/Y° 1 / I p p 1-1/YG°LY -1 CN,16 CN,17 CN,18 GN,19 CN,20 -1/2.86 1/2.86 Cp,i6 Cp,n Cp,i8 Cp,i9 Cp,20 Ce,16 CE,17 CE,18 Ce,19 CE,20 -1 1/Y° 1 / 1 P H A 1/Y° / I p p 1/YG°LY Autotrophic Nitrifying organisms XA C T S S J 6 CTSS.I 7 CTSS.I 8 CTSS.I 9 CTSS,2 0 21 r ° Autotrophic Growth gCODXA/d l-4.57/Y A CN,21 1/YA Cp,21 CE,21 1 CTSS,2 1 22. f A L Autotrophic Lysis gCODx A /d CN,22 Cp,22 CE,22 fxi.A l-fxi,A -1 CjSS.2 2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Component —> So SF S A S N H S N O SN2 Spo s, SHCO x , X s X H X p A 0 Xpp XpHA X G L Y x A X T S S ^Composition a o 00 Q O O 00 Q O U 00 • Z 00 Z 00 Z o a 00 a Q o u 00 a a a Q o u 00 a Q o u 00 en a Q o U 00 a Q o U 00 a 0^  00 C") a Q o u 00 CO a Q o u 00 HI a Q o u 00 a 1 COD gCOD -1 1 1 -2.86 1 1 1 1 1 1 1 1 2 T O C / C O D g C / g C O D 0.4 CN,I 0.334(a) 0.334 0.375 3 Nitrogen gN lN,SF IN.SA 1 1 1 1N,SI lN,XI 1N,XS IN.XH iN.BM IN.BM 4. Phosphorus gP ip.SF ip.SA 1 ip.si ip.xi ip.xs ip.XH ip.BM 1 ip.BM 5. Ionic charge Mole -1/64 +1/14 -1/14 -1.5/31 - i -1/31 6. TSS g ITSS.XI ITSS.XS ITSS.BM ITSS.BM iTSS.PP iTSS.PHA ITSS.GLY ITSS.BM —1 S7 APPENDIX A Table A.2a. Stoichiometric coefficients for SNH and Spo SN H(gN.m- 3) SPO (gP.m"3) CN,I = 1N,XS -iN.si'fsi -JN,SF' (1 - fsi) C p , i = ip,xs -jp,si 'fsi,H -ip,sF'(l _fsi ,H) CN,4 = IN.SF / Y H - i N , B M C p ,4 = ip,sF / Y H - i p , B M CN,5 = -1N,BM CP,5 = - ip ,BM CN,6 = IN.SF / Y H - i N , B M CP,6 = ip,SF / Y H - ip ,BM CN,7 = 1N,BM C p,7 = - ip .BM CN,8 = 1N,SF Cp,8 = ip,SF CN,9 = iN,BM - i N , X l ' f x i , H - i N , X F * (1 — fxi.n) CP)9 = ip,BM —ip,xi 'fxi.H -ip.xs '(1 -fxi.Fi) CN.IO = 0 c .« — V A N ^p,10 Ipo CN,II = 0 Gp,u = l CN,12 = 0 r _ vN 0 p^,12 Ipo C N . B = 1N,BM /Yp^ Cp, 13 = - I P B M / Yp^ A CN,I4 = 1N,BM /Yp^,0 C p , i 4 = ip.Bivi/Yp'p0 -1 CN,15 = iN,BM / Y Q ^ , Cp,i5 = IP.BM/YQ^Y CN,16 = IN.BM Cp,16 = ip.BM CN,17 = -1N,BM /Yp^A C p , i 7 = _ i p , B M / Y p ^ A CN,18 = IN.BM /Yp ,^ Cp,18 = ip.BM/Yp^ -1 CN,19 = i]M,BM /Y° L Y Cp,i9 = i p ,BM /Y° L Y CN,20 = IN.BM Cp,20 = lp,BM CN,2I = -IN.BM — 1/YA Cp,21 = - i p , B M CN,22 = IN.BM ~ i N , X I ' fxi,A - IN .XS ' (1 - fxi.A) Cp,22 = ip,BM - i p , X I 'fxi,A -ip,XS'(l _fxi ,A) 98 APPENDIX A The Integrated Metabolic Activated Sludge Model Table A.2b. Stoichiometric coefficients for SHco and XTss-S H C O (mole.m") X r s s (g.m" ) c e - =cN 1/14-c p l-(1.5/31) c e 4 = c N 4 / 1 4 - c p 4 . (1.5/31) c , 5 = cN>5/14 - c p 5- (1.5/31) + 1/(64.YH) c,e = c N , 6 /H-c p 6 -(1.5/31) + (1/YH -1)/(14.2.86) c e . 7 =c N > 7/l4-c p 7-(1.5/31) +(1/YH -1)/(14.2.86) + 1/(64.YH) c, 8 =cN 8/14-c p 8-(1.5/31)-1/64 ce.9 =cN 9/14-c p 9-(1.5/31) c „ o = YpAN(1.5/31) + l/64 + Y A N /31 c e l l =1.5/31 + 1/31 c , i 2 = YPA0N(1.5/31) +1/64 + YPA0N/31 + (1 - YSAAN)/(14.2.86) c „ 3 = c N y i 4 - c p '(1.5/31)-(1/YP N H°A -1)(14.2.86) 'TSS. l -•TSS,4 'TSS.5 -•TSS,6 TSS.XS TSS.BM TSS.BM TSS.BM ^TSS,6 ~~ 'TSS.BM C TSS,9 'TSS.XI ^XI.H ~*~ 'LTSS.XS 0 f X I . H ) 'L ' T S S . B M - i • f - Y ' ^ M + i ' Y A N + i *n-Y A N"> •'TSS.lO 'TSS.PP V PO / TSS.PHA SA T 'TSS.GLY V 1 1 SA / c - i * ( - Y A N W i * Y ^TSS.H ~ 'TSS.PP V PO / TSS.PHA 1i SA = ; / v N O - i TSS.13 'TSS.BM^PHA 1TSS.PHA APPENDIX A X r e s (g.nf3) - 'TSS.H •'TSS, 15 / Y N 0 TSS.BM' 1 PP 'TSS.PP / y N O _ j TSS.BM GLY 'TSS.GLY 'TSS.16 — TSS.BM _ : / v N U - i WsS.n — 'TSS.BM' PHA 'TSS.PHA •'TSS, 19 / Y N 0 - i TSS.BM' 1 PP 'TSS.PP / y N O • TSS.BM' 'GLY 1 TSS.GLY •'TSS^l — 'TSS.BM CTSS,22 'TSS.XI ^XI.H + 'TSS.XS 0 ^Xl.H) ' l S H C O (mole.m3) c,,4 = c N , , 4/l 4 - c P I 4 - (1.5/31) +1/(14.2.86. )-1 /31 Ce,5=cN^14-c P ] 1 5-(1.5/31) + (l/Y G N L° Y-l)/(14.2.86) c e.i 6=cN, 1 6/14-cPv(l-5/31) + l/(14.2.86) c,,7 =c R 1 7/14-c -(1-5/31) c=.i8 =cN18/14-cP18-(1.5/31)-l/31 c e l 9 =c N - 1 9 / l 4- c p l 9 - (1.5/31) c , 2o =cN20/14-cP20-(1.5/31) c , 2 , = cN>21/I4 - c P 2 1- (1.5/31) - 1/(14.YA) ce.22 =cN22/14-cp22-(1.5/31) Table A.3. Component composition factors belonging to the composition matrix (Table 1, bottom). Component composition factors in, ip and iiss 1 Nitrogen content of inert soluble COD, Si •N,SI 0.01 gN-g'CODs, 2 Nitrogen content of soluble substrate, Sp iN.SF 0.03 gN-g'CODsF 3 Nitrogen content of inert particulate COD, Xi lN,XI 0.03 gN-g'CODxi 4 Nitrogen content of particulate substrate, X s lN,XS 0.03 gN-g'CODxs 5 Nitrogen content of biomass, X H , XpAo, X A U T 1N,BM 0.07 gN-g'CODBM 6 Phosphorus content of inert soluble COD, Si ip,si 0 gP-g'CODsi 7 Phosphorus content of soluble substrate SF ip,SF 0.01 gP-g'CODsF 8 Phosphorus content of inert particulate COD, Xi ip,xi 0.01 gP-g'CODx, 9 Phosphorus content of particulate substrate, X s ip,xs 0.01 gP-g'CODxs 10 Phosphorus content of biomass, X H , X P A o, X A U T ip,BM 0.02 gP-g'CODBM 11 Ratio Total Suspended Solids to Xi iiss,xi 0.75 gTSS-g'CODxi 12 Ratio Total Suspended Solids to X s ITSS.XS 0.75 gTSS-g'CODxs 13 Ratio Total Suspended Solids to Biomass X H , X P A O , X A U T (CH2.09O0.54N0.2P0.OI5) 1TSS,BM 0.90 gTSS-g'CODBM 14 Ratio Total Suspended Solids to X P P (Mg2^Kj"/3P03) 1TSS,PP 3.23 gTSS-g'Ppp 15 Ratio Total Suspended Solids to X P H A (C4H602)i/4 iTSS,PHA 0.6 gTSS-g'CODpHA 16 Ratio Total Suspended Solids to XGLY (CeHioOs)^ ITSS.GLY 0.84 gTSS-g-'CODoLY 17 Ratio COD to Oxygen (So) icoD.o -1 gCOD-g '02 18 Ratio COD to Nitrate (SN 0) icOD.NO -2.86 gCOD-g'NsNo APPENDIX A Table A.4. Stoichiometric parameters 1 Fraction of inert COD generated in hydrolysis Regular Heterotrophic Organisms XH fsi 0 gCODsrg"1 COD(XH-XPAO) 1 Heterotrophic yield for growth on substrate Y H 0.63 gCODxH-g'COD 2 Fraction of inert COD generated in biomass lysis Autotrophic Nitrifying Organisms X A fxi,H 0.10 gCODxrg'CODxH 1 Autotrophic yield for growth Y A 0.24 gCODX A-g" 1NSNH 2 Fraction of inert COD generated in biomass lysis fxi,A 0.10 gCOD x ,-g' 1 COD X A 1 ATP produced per NADH or P/O ratio 5 1.85 Mole ATP-mole"'NAH 2 Observed biomass ratio TOC over COD a 0.334 gCpAo'E^CODpAo 4 Anaerobic yield for phosphate release Y A N 1 PO 0.184xpH-0.94 «0.35 gPsPo'g'CODsA 5 Yield for anaerobic formation of PHA from SA V A N 1SA 1.50 gCODpHB'g'CODsA 6 Observed yield for anoxic phosphate release V N 0 1 PO 0.23 gPsPo-g'CODsA 7 Yield for anoxic formation of PHA from S A V N 0 *SA (8.3'Y PN0°-4.9 + 8-5) Q n ] 9-8 gCODpHB'g'CODsA o APPENDIX A 8 Anoxic yield for degradation of XPHB V N O 1 PHA 19 + 3- V gCODpHB'g'CODpAo / a „ 4 + 9-5 sl.72 9 Anoxic yield for formation of XGLY V N 0 1 GLY 12.6 + 2- V / a » 1 . 1 8 gCODpGLY'g'CODpAO 9 + 6-5 10 Anoxic yield for formation of Xpp •y NO I Pp 57 + 9- V gPppg'CODpAo / a , 28 + 4-5 *3.02 11 Aerobic yield for degradation of XPHB Y ° 1 PHA 9.4 + 3- Y gCODpHB'g'CODpAo / a , 2 + 9-5 = 1.39 12 Aerobic yield for formation of XGLY 1 GLY 12.6 + 4- y / a wl . l l GCODcLY-g'CODpAo 9 + 12-5 13 Aerobic yield for formation of Xpp Y ° I pp 57 + 18- ¥ / a 28 + 4-5 = 4.42 gPpp-g'CODpAo Table A.5a. Kinetic parameters for Hydrolysis, X H and H A . Hydrolysis of Particulate Substrate 1 Hydrolysis rate(T) kh 3 .0xe( 0 0 4 0 6 ( T " 2 0 ) ) g C O D x s - g ' C O D ^ ^ - d - 1 2 Anoxic hydrolysis reduction factor n.N0 0.8 -3 Anaerobic hydrolysis reduction factor T|fe 0.2 -4 Saturation / inhibition coefficient for oxygen K 0 0.2 g02-m"3 5 Saturation / inhibition coefficient for nitrate K N O 0.5 gNsNo'm"3 6 Saturation coefficient for particulate COD^T ) K x 0_]xe(-0.110(T-20)) ' g C O D x s ' g - ' C O D ^ ™0 ) Heterotrophic Micro-Organism's X H 1 Maximum heterotrophic growth rate(T' 6.0xe(0.069.(T-20)) gCODxH-g'CODxH-d-' 2 Maximum fermentation rate(T) qfe 3_0xe(0.069.(T-20)) gCODsF-g'CODxH-d"1 3 Heterotrophic decay rate(T) b H 0_4xe(0.069.(T-20)) gCODxH-g'CODxH-d'1 4 Reduction factor for denitrification 0.8 -5 Saturation / inhibition coefficient for oxygen K 0 0.2 g02"m"3 6 Saturation coefficient for growth on SF K F 4.0 gCODsF-m"3 7 Saturation coefficient for fermentation of SF Kfe 20.0 gCODsF-m"3 8 Saturation coefficient for growth on acetate K A C 4.0 gCODsA-m'3 9 Saturation / inhibition coefficient for nitrate K N O 0.5 gNsNo'm"3 10 Saturation coeff. for ammonium (nutrient) K N 0.05 gNsNH'm"3 11 Saturation coefficient for Phosphate (nutrient) K P 0.01 gPSpo*m"3 12 Saturation coefficient for alkalinity ( H C O 3 " ) K H C O 0.1 MoleHC03"-m"3 Autotrophic Micro-Organism's XA 1 Autotrophic growth rate(T) HA L0XE(0.105.(T-20)) GCODxA-g'CODxA-d" 2 Autotrophic decay rate(T) bA 0.15xe( a , 0 5 ( T - 2 0 ) ) gCODxA-g'CODxA-d" 3 Saturation coefficient for oxygen KA,O 0.5 g02"-m"3 4 Saturation coefficient for Ammonium K-NH 1.0 gNsNH'm"3 5 Saturation coefficient for Phosphate (nutrient) K P 0.01 gNSPo-m"3 6 Saturation coefficient for alkalinity ( H C O 3 ) KHCO 0.5 moleHC03"-m"3 (T) Temperature dependant parameters Table A.5b. Kinetic parameters for X P A o-Phosphorus Accumulating Organisms X P A O 1 Maximum anaerobic acetate uptake rate(T) qAc 80xe(0090(T-20)) gCODsAg'CODpAo-d"1 2 Anaerobic maintenance rate mAN 0-5xe(0.069(T-20)) gPpp-g'CODpAo-d"1 3 Maximum anoxic acetate uptake rate ( T ) NO MAC 1.5-qA Cxe ( a 0 9 0- ( T-2 0 ) ) gCODsA-g'CODpAo-d'1 4 P H A degradation rate(T) kpHA 5.51xe ( 0 1 2 ,- ( T- 2 0 ) ) gCODpHAg-'CODpAo-d"1 5 Glycogen formation rate(T) kGLY 0.93 xe ( 0 1 , 8 - ( T - 2 0 ) ) gCODGLY-g'CODpAo-d1 6 Poly-phosphate formation rate *T) kpp 0.10xe ( 0 0 3 1- ( T- 2 0 ) ) gPppg'CODpAo-d"1 7 Observed oxygen consumption for moc 0.096 g02-g- ,CODpAO-d-1 8 Aerboic maintenance rate(T) m 0 3 - 5 - 5 ° c 7 - 0 . 0 6 xe ( 0 0 6 9 - ( T - 2 0 ) ) 3.2+ 9/ / a gCODpAo-g'CODpAo-d1 9 Anoxic maintenance rate(T^ mNo 6 - 5 - 5 ° c - 0 . 0 6 x e ( a o 6 9 - ( T - 2 0 ) ) 6.3 + y / a GCODpAo-g'CODpAo'd"1 10 Saturation reduction factor for PP formation gpp 0.22 -11 Reduction factor for dentrifying P removal r|NO 0.8 -12 Saturation coefficient for poly-P formation Kpo 1.0 gPspo-m"3 A P P E N D I X A 13 Saturation coefficient for growth on acetate K A C 4.0 gCODsA-m"3 14 Saturation / inhibition coefficient for nitrate K N O 0.5 g N s N O - m " 3 15 Saturation / inhibition coefficient for Ko 0.2 G02-m"3 16 Saturation coefficient for fpHA KFPH 0.2 gCODpHA-g'CODpAo 17 Saturation coeff.for Phosphate (nutrient) K P 0.02 gPS P 0-m" 3 18 Saturation coefficient for N H 4 (nutrient) K N 0.05 g N s N r f m " 3 19 Maximum poly-phosphate fraction of PAO's jmax 0.35 gPpp-g'CODpAo 20 Maximum glycogen fraction of PAO's 0.5 gCODcLYg'CODpAo 21 Saturation coefficient for poly-P Kpp 0.01 gPpp- m"3 22 Saturation coefficient for glycogen K G L Y 0.01 gCODcLY'ni3 23 Saturation coefficient for PHA KpHA 0.01 GCODpHA-m"3 24 Saturation coefficient for fGLY KfGL 0.01 gCODoLY-g'CODpAo 25 Saturation coefficient for fpp Kfpp 0.01 gCODppg'CODpAo 26 Saturation coefficient for alkalinity ( H C O 3 " ) K H C O 0.01 MoleHC03"m-3 Temperature dependent parameters APPENDIX A Table A.6. Kinetic rate equations Process Kinetic Rate Equation (r?) Switch function (on / off) Hydrolysis of Particulate Substrate X S ) Aerobic Hydrolysis 1 2 (gCODxs-d"') Anoxic Hydrolysis (gCODxs-d"1) r ° = k „ RNO . K . rh - MNO H X S / ( X H + X P A 0 ) K X + X S / ( X H + X P A o ) X S / ( X H + X P A 0 ) . ' ( X H + X P A o ) 'NO K X + X s / ( X H + X PAo) K N O + S N 0 * ( X H + X P A 0 ) K , K 0 +S 0 Anaerobic Hydrolysis (gCODxs-d') X 5 / ( X H + X P A O ) K X + X S / ( X H + X P A o ) ' ( X H + X P A o ) ^ O • ^ NO K 0 + S 0 K N 0 + SNO c NH • Spo c °HCO K N + S N H K F + S P 0 ^HCO + H^CO Spo °HCO K-N + ^NH K p + S P 0 KpjCO + H^CO Heterotrophic Microorganisms X H 4 Aerobic growth on SF(gCODxHd'1) 5 Aerobic Growth on SA(gCODxH-d"1) 6 Anoxic Growth on SF(gCODXH-d"1) 7 Anoxic Growth on SA(gCODxH-d-1) 8 Fermentation of Sp (gCODsF-d1) 9 Heterotrophic lysis (gCODxH-d'1) RSF M-H RSA — M^H S A + Sp Kp +Sp KQ + S 0 •x. S A + S F K A + S A K 0 +S 0 -•xt R N O . . 'SF MNO HH 'NO S A + S F K F + S F K N 0 + S xt NO rSF — I^NO M-H 'NO S A +S F K A + S A s. ^NO "*"SN0 - ' X T r A N =a K f e + S F rHL = b H ' X H * X T K-o • S N H . SF ' H C O ^o + S 0 K N + S N ? I Kp+Spo K H C 0 + S H C 0 K o • S^ 'PO HCO ^o + S 0 K N + S N H Kp+Sp 0 K H C 0 + S H C 0 K r K , 'HCO K-o"*"S0 K N O + S N O K p I C O + S H C O APPENDIX A Table A.6. Kinetic rate equations Process Kinetic Rate Equation (r?) Switch function (on / off) Phosphorus Accumulating Organisms XpAo> 10 Anaerobic storage of S A (gCODsA-d1) 11 Anaerobic Maintenance (gP-d-1) 12 Anoxic storage of SA (gCODsA-d"1) 13 Anoxic PHA consumption (gCODxPHA-d-1) 14 Anoxic storage of PP (gP-d"1) 15 Anoxic Glycogen formation (gCODxGLVd'1) r A N = a * 'SA MAC K A + S A - * X PAO 'M — '"AN'^PAO NO NO 'SA — MAC 'NO rPHA=n + S A KNO + S N o Y / Y ^PHA'^-PAO • NO , v PHA K , X p H A ^ P A O K N 0 + S N 0 .NO X r P P =n PAO 'PO 'NO NO P^P X c K P O + S p o g P P « K N 0 + S N 0 .NO = W k xt 'NO NO GLY •^GLY ^NO + ^NO x PAO X PAO KPAO K , K NO X GLY xc K 0 + S n K O i v N O ' NO ^ GLY K N O ^ G L Y K p p + X p p X pp K 0 + S 0 K N 0 + S N 0 pp pp K , X t K 0 + S 0 K p P + X p p K r K 0 +S 0 K - N + S n h Kp+Sp 0 K H C 0 + S H C 0 K r X c r- max r y I p p A.pp/A. p A 0 K K , •"O PHA + ^  PHA K f p p + ( f p p X p p / X p A 0 ) X PHA c max y GLY G L Y P A O "*"^ 0 K p H A + X p H A K fQLY + (fGLY ^ G L Y ^ P A o ) 16 Anoxic Maintenance (gCODxPAo-d-1) 17 Aerobic PHA consumption (gCODxPHA-d1) 18 Aerobic storage of PP (gP-d"1) K-NO + ^No -*x PAO r u -1c *PHA — PHA PHA ^  PAO KfPHA + ^ P H A ^ P A O KQ+SQ -*x PAO r ° - k ' 'pp — "^ PP X PAO • 'PO X pp K PO + SPO g p p ' K 0 + S 0 -'Xt K 0 + S 0 NH 'PO 'HCO K N + S N H K P + S P 0 K H C 0 + S H C 0 X PHA • I p p — A p p / A . p A 0 ^ P H A + ^ P H A K f p p + ( f p p — X p p / X p A 0 ) O APPENDIX A 19 Aerobic glycogen formation (gCODxcLY-d-1) 20 Aerobic maintenance (gCODxPAO-d"') 'GLY K-GLY Y rr Q ^PAO •^GLY "^O + JGLY ~ AGLY / V VPAO K PHA + X PHA K (GLY max GLY Y AY 1 GLY PAO J r N 0 = m * 'M 1 1 1 NO ! o _ . x K 0 + S 0 PAO Autotrophic Nitrifying organisms X A 21 Autotrophic growth (gCODxAd-1) 22 Autotrophic Lysis (gCODxA-d1) KNH + S N H K 0 + S 0 X , K PO K P + S P 0 TAL _ bA ' X A APPENDIX A Table A . 7. Changed values of parameter due to the calibration when applied to U B C pilot plant (AI-Atar, 2006) Parameter Units Meijer (2004) Calibrated Parameters (AI-Atar, 2006) life 0.2 0.1 r| N03 0.8 0.65 1] N03 0.8 0.2 n P M N03 0.8 0.8 fxSin gCOD/ gCOD 0.439 0.85 gPP 0.22 0.22 •NXI gN/gCOD 0.03 0.035 'NXS gN/gCOD 0.03 0.04 ipxi gP/gCOD 0.01 0.005 !PXS gP/gCOD 0.01 0.005 K G L Y T gCOD/(gCOD d ) . 0.93 0.93 1^  NH4 gN/m 3 .1 0.1 1^ 02 g0 2 /m 3 0.2 0.2 kpHA gCOD/(gCOD d) 5.51 5.51 k T gP/(gCOD d) 0.1 0.1 „ T qfe gCOD/(gCOD d) 3 1 „ max T gCOD/(gCOD d ) 8 8 Y H gCOD/gCOD 0.63 0.63 Yp04 gCOD/gCOD 0.35 0.35 K P A * gCOD S A /m3 4 1 111 APPENDIX B APPENDIX B: INSTRUMENTAL OPERATIONAL PARAMETERS 112 APPENDIX B Table B.l. Instrument operational parameters for the flame atomic absorption spectrophotometer Element analyzed Magnesium Concentration Units mg/L Instrument Mode Absorbance Sampling Mode Auto normal Calibration Mode Concentration Measurement Mode Integrate Lamp current 4.0Ma Replicates Standard 3 Replicates Sample 3 Wavelength 202.6 Range 0-100mg/L Flame Type N 2 0 / C 2 H 2 Calibration Algorithm New Rational Table B.2. Instrument operational parameters for the LaChat QuickChem flow injection analysis instrument Ion Analyzed P O 4 - P N H 3 - N Concentration Units mg/L mg/L Range 0-100mg/L 0-100mg/L Temperature 63UC 63°C Method Ammonium Molybdate Phenate Reference 1 2 1: From LaChat Instuments Methods Manual for the QuickChem r Automated Ion Analyzer (1990) QuickChem method number 10-115-01-1Z. 2: A P H A , A W W A and WPCF (1995) Part 4500-NH3-F .Phenate method. In Standard Methods for the examination of Water and Wastewater, 19 th edition. American Public Health Association, Washington,D.C. 113 APPENDIX C APPENDIX C: DAILY MONITORING RECORDS (IN CD-ROM) MEBR, CEPBR AND SIDESTREAM PROCESSES (In CD-ROM) 114 APPENDIX D APPENDIX D: MASS BALANCE EQUATIONS 115 APPENDIX D Mass balance equations for P O 4 - P release in anaerobic zone and P O 4 - P uptake in anoxic and aerobic zones. 1 Cp2- Qanx rec _l Anaerobic Zone w Cpi. Qin w Cp3- Qana Anaerobic P release = Cpi Q j n + C P 2 Q a Cp3 Q a C P 2 . Q anx ree* Cp3- Q a Cp4. Q a Anoxic Zone Cp5- Q a Anoxic P uptake = C P 3 Q a n a - C P 2 Q a n x rec + C P 4 Q a e r rec - C P 5 Q a Cp4. Q aer rec C p 5 . Q a Anaerobic Zone — • Cp6- Qeff Aerobic P uptake = C P 5 Q a n x - C P 4 Qaer rec - C P 6 Q e f f Where, Cpi = Concentration of PO4-P in the influent Cp2 = Concentration of P O 4 - P in the anoxic recycle Cp3 = Concentration of PO4-P in the flow from anaerobic to anoxic zone C P 4 = Concentration of P O 4 - P in the aerobic recycle\ Cps = Concentration of PO4-P in the flow from anoxic to aerobic zone Cp6 = Concentration of P O 4 - P in the effluent Q i n = Influent flow Qanx rec = Anoxic recycle flow to the anaerobic zone Q a n a = Flow from anaerobic to anoxic zone Qaer r e c = Aerobic recycle flow to the anoxic zone Qanx = Flow from anoxic to aerobic zone Qeff = permeate flow (All concentrations are in mg/L and flows are in m3/d) 116 APPENDIX D APPENDIX D: MASS BALANCE EQUATIONS 115 APPENDIX D Mass balance equations for PO4-P release in anaerobic zone and PO4-P uptake in anoxic and aerobic zones. Cp2. Q a Cpi. Qu Anaerobic Zone Cp3- Qa Anaerobic P release = Cpi Qin + Cp2 Q a Cp3 Q a C p 2 . Q anxrec* Cp3- Qa Cp4. Q a Anoxic Zone Cp5. Q a Anoxic P Uptake = C P 3 Qana - Cp2 Qanxrec + C P 4 Qaerrec - C P 5 Qa Cp5- Qa Cp4. Q a Anaerobic Zone Cp6- Qeff Aerobic P uptake = CPS Q a n x - Cp4 Q a Cp6 Qeff Where, Cpi = Concentration of P O 4 - P in the influent Cp2 = Concentration of P O 4 - P in the anoxic recycle Cp3 = Concentration of P O 4 - P in the flow from anaerobic to anoxic zone Cp4 = Concentration of P O 4 - P in the aerobic recycle\ CP5 = Concentration of P O 4 - P in the flow from anoxic to aerobic zone CP6 = Concentration of P O 4 - P in the effluent Qin = Influent flow Vanx rec = Anoxic recycle flow to the anaerobic zone Qana = Flow from anaerobic to anoxic zone Qaerrec = Aerobic recycle flow to the anoxic zone Qanx = Flow from anoxic to aerobic zone Qeff = permeate flow (All concentrations are in mg/L and flows are in m3/d) 116 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0063252/manifest

Comment

Related Items