Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigation of electrical conductivity as a control parameter for enhanced biological phosphorus removal… Wylie, Andrew Colin 2009

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

Item Metadata

Download

Media
24-ubc_2009_spring_wylie_andrew.pdf [ 1.83MB ]
Metadata
JSON: 24-1.0067103.json
JSON-LD: 24-1.0067103-ld.json
RDF/XML (Pretty): 24-1.0067103-rdf.xml
RDF/JSON: 24-1.0067103-rdf.json
Turtle: 24-1.0067103-turtle.txt
N-Triples: 24-1.0067103-rdf-ntriples.txt
Original Record: 24-1.0067103-source.json
Full Text
24-1.0067103-fulltext.txt
Citation
24-1.0067103.ris

Full Text

INVESTIGATION OF ELECTRICAL CONDUCTIVITY AS A CONTROL PARAMETER FOR ENHANCED BIOLOGICAL PHOSPHORUS REMOVAL IN A PILOT SCALE SEQUENCING BATCH REACTOR  by  ANDREW COLIN WYLIE B.Sc., Honours Microbiology, University of Guelph, 2001 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Plant Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2009  © Andrew Colin Wylie, 2009  i  Abstract  The relationship between phosphorus (P) concentration and process control parameters was investigated in a pilot scale enhanced biological phosphorus removal (EBPR) sequencing batch reactor (SBR). Datasets were examined to determine whether process control data could be used to detect a (P) removal failure, and P and nitrogen sensors were installed to improve the resolution of their respective measurements of these types of molecules. Track studies were performed to determine whether the electrical conductivity (EC):P relationship that had previously been demonstrated in the lab holds in larger SBR systems: in this relationship EC increases as P is released by polyphosphate accumulating organisms, and decreases as inorganic P is taken up, possibly due to the corresponding transport of the cations magnesium and potassium. The relationship was confirmed, showing that EC could be used as a real-time control parameter to optimize the length of anaerobic and aerobic phases of the SBR in order to maximize energy savings and to decrease the possibility of eutrophication caused by EBPR failure.  ii  Table of Contents ABSTRACT ..................................................................................................................................... IIii iii TABLE OF CONTENTS ................................................................................................................. III vi LIST OF TABLES........................................................................................................................... VI vii LIST OF FIGURES ........................................................................................................................ VII ix LIST OF ABBREVIATIONS AND ACRONYMS ............................................................................ IX xi LIST OF UNITS .............................................................................................................................. XI xii ACKNOWLEDGEMENTS ............................................................................................................. XII 1 INTRODUCTION .......................................................................................................................... 1 1.1 THE SOLUTION TO POLLUTION IS NOT DILUTION: THE BOD 5 PARADIGM VS. WATER RECYCLING ...... 1 1.2 PHOSPHORUS BALANCE IN NATURAL SYSTEMS IS CHALLENGED BY HUMAN ACTIVITY ....................... 2 1.3 W ASTEWATER CAN BE TREATED USING ENHANCED BIOLOGICAL PHOSPHORUS REMOVAL (EBPR) 4 1.3.1 History ....................................................................................................................................... 4 1.3.2 Overview of the EBPR process .......................................................................................... 5 1.3.3 Interferences with EBPR ...................................................................................................... 8 1.3.4 The role of cations in EBPR ................................................................................................ 9 1.4 MICROBIOLOGY OF EBPR PROCESSES ............................................................................................11 1.4.1 Culture-based methods and Acinetobacter ...................................................................11 1.4.2 Molecular methods for identifying PAOs ........................................................................14 1.4.3 Accumulibacter and Rhodocyclus in EBPR ...................................................................18 1.4.4 Other PAOs in EBPR ............................................................................................................21 1.4.5 The GAO question is unresolved for EBPR ...................................................................21 1.5 SBR SYSTEMS IN EBPR PROCESSES ...............................................................................................23 1.6 PROCESS CONTROL OF EBPR PROCESSES .....................................................................................25 1.6.1 ORP – Nitrate knee indicates the end of denitrification ..............................................25 1.6.2 ORP indicates P concentration .........................................................................................26  iii  1.6.3 DO Elbow indicates end of nitrification ...........................................................................26 1.6.4 pH indicates various biological processes ....................................................................26 1.6.5 pH control, and pH effect on P removal ..........................................................................28 1.6.6 EC is related to P release in EBPR ...................................................................................28 1.7 W HY STUDY A PILOT-SCALE PLANT? ..................................................................................................31 2 MATERIALS AND METHODS ................................................................................................... 32 2.1 PRELIMINARY DATA ANALYSIS ............................................................................................................32 2.2 PROTOTYPE ON-LINE P AND N ANALYZER SETUP..............................................................................32 2.3 TRACK STUDIES .................................................................................................................................38 2.3.1 Reactor Setup and Operation ............................................................................................38 2.3.2 Data Collection ......................................................................................................................39 2.3.3 On-line data collection .........................................................................................................39 2.3.4 Track study sampling...........................................................................................................42 2.3.5 Cation Analysis......................................................................................................................44 2.3.6 COD analysis ..........................................................................................................................44 2.3.7 Statistical analyses ...............................................................................................................45 3 RESULTS AND DISCUSSION................................................................................................... 45 3.1 PRELIMINARY RESEARCH ..................................................................................................................45 3.1.1 Data Analysis .........................................................................................................................45 3.1.2 The need for on-line analysis of P and N Removal ......................................................52 3.1.3 The need for track studies ..................................................................................................53 3.2 PROTOTYPE ANALYZER SETUP .........................................................................................................54 3.3 DAILY REACTOR PERFORMANCE .......................................................................................................56 3.4 TRACK STUDIES .................................................................................................................................57 3.4.1 Pseudo-steady states: .........................................................................................................57 3.4.2 Phosphate concentration versus EC ...............................................................................59 3.4.3 Phosphate concentration versus cation concentration ..............................................62  iv  3.4.4 EC versus cation concentration ........................................................................................67 3.4.5 P concentration versus ORP ..............................................................................................68 3.5 APPLICATIONS ....................................................................................................................................70 3.5.1 Chemical precipitant dose alarm for poor P removal ..................................................71 3.5.2 Nitrate knee C dosing to favour PAOs .............................................................................72 3.5.3 Improved algorithms for maximum energy savings ....................................................73 3.6 IMPROVEMENTS..................................................................................................................................73 3.7 PROCESS MODELING AND CONTROL.................................................................................................74 4 CONCLUSION............................................................................................................................ 75 REFERENCES .............................................................................................................................. 76 APPENDIX A: SOLUTIONS ......................................................................................................... 88 APPENDIX B: SCHEDULE ........................................................................................................... 88  v  List of Tables TABLE 3.1  P REMOVAL FAILURES IN THE PILOT SCALE SBR.....................................................................47  TABLE 3.2  DETAILS OF GREATER THAN 50% FAILURE OF P REMOVAL EVENTS .........................................48  TABLE 3.3  DETAILS OF OF LESS THAN 50% FAILURE OF P REMOVAL EVENTS ...........................................49  TABLE 3.4  COEFFICIENTS OF CORRELATION BETWEEN EC (US) AND P CONCENTRATION (MG/L) FOR SEVEN TRACK STUDIES OF NONCONSECUTIVE SEQUENCING BATCH REACTOR CYCLES. ......................59  TABLE 3.5  VALUES FOR THE GRAM TO GRAM RATIO OF MG++ TO P AND K+ TO P OBSERVED IN THIS STUDY AND REPORTED IN SELECTED LITERATURE (ADAPTED FROM BARAT ET AL., 2005* AND KIM ET AL., 2007). ...............................................................................................................................66  TABLE A.1  TRACK STUDY SCHEDULE AND AVAILABILITY OF PLC DATA..................................................88  vi  List of Figures FIGURE 1.1  IN ALTERNATING ANAEROBIC/AEROBIC CONDITIONS, POLYPHOSPHATE ACCUMULATING ORGANISMS (PAOS) WILL TAKE UP CARBON COMPOUNDS AND RELEASE NEGATIVELY CHARGED P AND CATIONS IN THE ANAEROBIC ENVIRONMENT, AND WILL TAKE UP FREE P IN SOLUTION AND CATIONS IN THE AEROBIC ENVIRONMENT. ..................................................  6  FIGURE 1.2  SBR OPERATION CYCLE FOR EBPR. ...................................................................................24  FIGURE 2.1  SETUP OF AUTOMATIC P AND N SAMPLING SYSTEM FOR THE PILOT SCALE SBR. DETAILS OF VARIOUS COMPONENTS IN FIGURES 2.2-2.5 .......................................................................34  FIGURE 2.2  THE TWO TANK SBR (LEFT AND RIGHT) WITH SAMPLE FILTRATION MEMBRANE UNIT (CENTER) ...........................................................................................................................................35  FIGURE 2.3  TOP: EC SENSOR AND CHART RECORDER (LATER REPLACED BY PLC) BOTTOM: SAMPLE PRESSURE REGULATOR ASSEMBLY. .....................................................................................36  FIGURE 2.4  ON-LINE ANALYZER BANK - DR LANGE PROZEΒHOTOMETER-NITRAT (TOP), AND HACH SERIES 5000 PHOSPHATE ANALYZER (BOTTOM). ................................................................37  FIGURE 2.5  EC PROBE WITH BUBBLE SHIELD SHOWING BUILDUP OF COLLOIDAL MATERIAL WITHIN EC CELL: THE BUBBLE SHIELD WAS LATER REPLACED WITH THE INSERTION ASSEMBLY SEEN IN  FIGURE 2.6 WHEN THE CHART RECORDER WAS REPLACED BY A CONNECTION TO THE PLC.38 FIGURE 2.6  ORION CONDUCTIVITY CELL RETROFITTED FOR SUBMERSIBLE OPERATION. ...........................41  FIGURE 2.7  SAMPLE PREPARATION AND ANALYSIS SCHEME FOR TRACK STUDIES PERFORMED ON THE PILOT SBR. .........................................................................................................................43  FIGURE 3.1  SOME TYPICAL DAILY ANALYTICAL DATA SHOWING THE P REMOVAL ((P INF -P EFF )/P INF X 100%) FOR A SEQUENCING BATCH REACTOR INDICATING FAILURE OF P REMOVAL, BUT UNABLE TO RESOLVE THE BOTTOM OF THE FAILURE TROUGH. ............................................51  FIGURE 3.2  PERFORMANCE OF TWIN PILOT-SCALE SBR FED WITH DOMESTIC WASTEWATER FOR THE PERIOD MAY AND JUNE, 2006. ARROWS INDICATE NEGATIVE % REMOVAL EVENTS. ..........56  FIGURE 3.3  TYPICAL PROFILES FOR ORP (BOTTOM LINE), PH (MIDDLE LINE), AND EC (UPPER LINE) IN THE PILOT SCALE SBR. ILLUSTRATED FOR ONE CYCLE AX/AN INDICATES ANOXIC/ANAEROBIC CONDITIONS, AO INDICATES AEROBIC. THE ARROWS INDICATE THE ORP “NITRATE KNEE”  vii  THAT SIGNALS THE SWITCH BETWEEN ANOXIC AND ANAEROBIC CONDITIONS FOLLOWING THE DEPLETION OF INITIAL NITRATE. ..................................................................................57  FIGURE 3.4  EC AND ORP PROFILES LEADING UP TO AND INCLUDING TRACK STUDY 4: EC SENSOR BECOMES FOULED PRIOR TO THE STUDY. .............................................................................58  FIGURE 3.5  EC AND ORP PROFILES LEADING UP TO AND INCLUDING TRACK STUDY 6: EC SENSOR IS FOULED PRIOR TO THE STUDY. ............................................................................................58  FIGURE 3.6  SCATTERPLOTS OF [P] VS. EC IN SEVEN NONCONSECUTIVE SEQUENCING BATCH REACTOR CYCLES IDENTIFIED AS TS 3 THROUGH TS 10. ....................................................................60  FIGURE 3.7  CONDUCTIVITY, PO 4 3-, AND CATION PROFILES FOR K+, MG++, AND CA++ OVER THE COURSE OF A TRACK STUDY PERFORMED ON A PILOT-SCALE SEQUENCING BATCH REACTOR: “TRACK  STUDY 6”. THE CONDUCTIVITY WAS MEASURED MANUALLY WITH AN ORION EC METER, AND LOGGED TO A PLC FROM THE SAME METER.................................................................62  FIGURE 3.8  P CONCENTRATION VS. K CONCENTRATION OVER THE COURSE OF A SINGLE TRACK STUDY: “TRACK STUDY 6”...............................................................................................................64  FIGURE 3.9  P CONCENTRATION VS. MG CONCENTRATION OVER THE COURSE OF A SINGLE TRACK STUDY: “TRACK STUDY 6”...............................................................................................................64  FIGURE 3.10  P CONCENTRATION VS. CA CONCENTRATION OVER THE COURSE OF A SINGLE TRACK STUDY: “TRACK STUDY 6”...............................................................................................................65  FIGURE 3.11 CORRELATIONS BETWEEN P CONCENTRATIONS IN MG/L AND CATION CONCENTRATIONS IN MG/L OVER THE COURSE OF THREE NONCONSECUTIVE SBR CYCLES. ..................................65  FIGURE 3.12 CORRELATION BETWEEN P AND ORP DURING A TRACK STUDY ON A PILOT SCALE SEQUENCING BATCH REACTOR: “TRACK STUDY 4”. .................................................................................69  FIGURE 3.13  CORRELATION BETWEEN P AND ORP DURING A TRACK STUDY ON A PILOT-SCALE SEQUENCING BATCH REACTOR: “TRACK STUDY 6”. ............................................................70  viii  List of Abbreviations and Acronyms A/O  Anaerobic/aerobic (proprietary)  A2O  Anaerobic/anoxic/aerobic (proprietary)  bsCOD  Biodegradable soluble COD  COD  Total chemical oxygen demand  CFB  Cytophaga-Flavobacterium-Bacteroides group of bacteria  DGGE  Denaturing gradient gel electrophoresis  DO  Dissolved oxygen  DPAO  Denitrifying phosphorus accumulating organism  EBPR  Enhanced biological phosphorous removal  EC  Electrical conductivity  FISH  Fluorescence in situ hybridization  GAO  Glycogen accumulating organism  MAR  Microautoradiography  O-PO 4  Orthophosphate  ORP  Oxidation-reduction potential  PAO  Polyphosphate accumulating organism  PHA  Poly-β-hydroxyalkanoate  PHB  Poly-β-hydroxybutyrate  PLC  Programmable logic controller  Poly-P  Polyphosphate  Q-8, Q-9, Q-10  Ubiquinones ix  rbCOD  Readily biodegradable COD  SBR  Sequencing batch reactor  SND  Simultaneous nitrification/denitrification  SRT  Solids retention time  TSS  Total suspended solids  UCT  University of Cape Town process  VFA  Volatile fatty acid  VIP  Virginia Initiative Plant process  x  List of Units degrees Celsius (0C) litres (l) micrometres (μm) microsiemens (µS) (uS in chart legends) milligrams (mg) millilitres (mL) millivolts (mV) nanometres (nm) siemens (S)  xi  Acknowledgements  I would first like to thank my committee members Dr. Sietan Chieng, Dr. Kwang Victor Lo, and Dr. William Ramey for guiding me through this work. Fred Koch, UBC Environmental Engineering, and Bud Fraser and Peter Doig, Vision Envirotech International helped with project design and equipment setup, as did Ping Huang Liao, UBC. Further thanks go to Paula Parkinson and Susan Harper, UBC Environmental Engineering for assistance with laboratory analyses. UBC Civil Engineering graduate student Iqbal Hossain provided help with the EC equation and its temperature coefficient for UBC’s wastewater, UBC Civil Engineering graduate student Parvez Fattah helped with Visio. This project was funded through the National Sciences and Engineering Research Council’s Industrial Postgraduate Scholarship, and I would like to thank Dr. William Cheuk and Vision Envirotech International for their involvement in the project. Finally, my family’s continued patience and support were invaluable throughout the project.  xii  1 Introduction  1.1 The solution to pollution is not dilution: the BOD5 paradigm vs. water recycling  Wastewater engineering’s most emblematic laboratory test is the five-day biochemical oxygen demand (BOD5) test. It remains the most widely used measure of organic pollution (Tchobanoglous et al., 2003), since its inception shortly before the turn of the last century. The rationale for the test is based on the observation that in polluted water there is an oxygen demand caused by microbial breakdown of organic material and ammonia: by mixing wastewater with aerated water in a sealed bottle, after five days at 18.3oC a major fraction of the pollutants are broken down resulting in a quantifiable drop in dissolved oxygen (Henze, 1997).  The five-day aspect of the test is no accident: as the story goes, the test was invented in Britain, where no river takes more than five days to empty out into the ocean, and the mean summer water temperature is 18.3oC (Gray, 2004). Thus, as long as wastewater is low enough in nutrients to not cause eutrophication within five days, it will not have an adverse effect on any part of the island. This simple and effective test belies a certain attitude: as long as pollutants can make it out to the ocean without harming freshwater along the way, these pollutants are then gone from the sphere of human concern. In other words, “The solution to  1  pollution is dilution”, as the ocean represents a vast dilution factor compared to Britain’s rivers and streams.  The old attitude persists, but there are signs of change. For instance, the cardinal text “Wastewater Engineering” produced by the firm Metcalf and Eddy, Inc. has gone through some telling name changes over the last three decades. The 1972 edition was subtitled “Collection, Treatment, Disposal”. The subtitle was changed in 1979 to “Treatment, Disposal, Reuse”, which lasted through the 1990 edition. The 2003 edition sports the subtitle “Treatment and Reuse”, representative of the evolving focus of the industry towards meeting ever stricter environmental requirements through processes such as biological nutrient removal (Tchobanaglous et al., 2003), which is the focus of this thesis.  1.2 Phosphorus balance in natural systems is challenged by human activity  Phosphorus is a key element in biological systems: organisms use it to build energy compounds, nucleic acids, and phospholipids. Because most phosphorus in the biosphere is in insoluble forms, it is a limiting nutrient in most freshwater ecosystems (Tchobanoglous et al., 2003, p. 524). An influx of soluble phosphorus from such sources as agricultural runoff of fertilizers, domestic wastewater, or atmospheric pollution into a receiving body that is low in iron, calcium, and aluminum creates a favourable environment for the nitrogen-fixing  2  cyanobacteria. Whereas in soil systems deposited phosphorus is quickly bound as insoluble salts with the aforementioned metals, in freshwater the soluble phosphorus causes harmful eutrophication (Schlegel, 1993). Limits on treatment plant effluents thus depend on the potential impact of the effluent on the receiving body, and have traditionally fallen between 0.1 and 2.0 mg·L-1 of phosphorus (Tchobangalous et al., 2003, p.624).  Around 75-85% of the phosphorus discharged in the developed world comes from domestic wastewater, originating as excreta or detergents (Rybicki, 1997). The proportion contributed by detergents appears to have declined (along with a corresponding overall decline) since the 1970s due to the replacement of phosphate detergents with phosphate free alternatives (Rybicki, 1997). An estimated 30-40% reduction of total phosphorus in municipal wastewater could result from the abandonment of phosphorus-containing detergents, although concerns over the environmental impact of the alternatives remain, especially in the case of nitriloacetic acid which cannot be removed through biological treatment (Rybicki, 1997). A recent report assumes 0.36kg from urine, 0.18kg from faeces, and 0.18kg from detergent per capita (Balmér, 2004). This results in a standard domestic wastewater P concentration of 5-12 mg·L-1 (Gray, 2004 p.1065). Thus, a robust biological phosphorus removal system is needed in order to meet the challenges of human impact on natural systems, while being affordable enough that municipalities will be able to meet ever more stringent phosphorus effluent regulations.  3  1.3  Wastewater  can  be  treated  using  Enhanced  Biological  Phosphorus Removal (EBPR)  1.3.1 History  Phosphorus has traditionally been removed from wastewater by chemical means, most often treatment using alum or iron salts. Chemical removal is an expensive process, and biological removal reduces the costs associated with the purchase of the chemicals required, as well as producing less sludge (Tchobanoglous et al., 2003). Early work on biological phosphate removal was reported by Sawyer in 1944, Rudolfs in 1947, and Greenburg et al. in 1955(Rybicki, 1997). An early report showed that phosphorus could be taken up and released by biomass, termed “luxury uptake” (Levin and Shapiro, 1965). Ensuing work by Menar and Jenkins (1970), Weichers and van Vuuren (1979), and Simpkins (1979), assumed that the mechanism was one of chemical precipitation and adsorption onto sludge, and until Fuhs and Chen (1975) reported that the phosphate removal occurred mostly in the form of intracellular polyphosphates, the technology was viewed with skepticism (Rybicki, 1997). During this period, it was debated whether phosphorus removal was a biological rather than a chemical effect. The two proposed chemical causes were precipitation by naturally occurring precipitants in the wastewater and adsorption to clay and mineral particles (Arvin, 1985). It has recently been shown that less than 10% of the phosphate found in P-accumulating microbial granules is related to chemical precipitation (Liu et al., 2005).  4  Barnard (1974) determined that wastewater exposed to first an anaerobic zone, and then an aerobic zone could undergo biological phosphorus removal, and stressed that the solids retention time (SRT-the average amount of time that sludge remains in the system (Tchobanoglous et al., 2003)) must be kept short. Barnard (1975) developed a process that uses an alternating anaerobic/aerobic process for EBPR. Termed Phoredox, it is commercially known as A/OTM. The initial process has been modified in several ways. It has been combined with various biological nitrogen removal strategies, which required inclusion of an anoxic phase, known as an A2OTM process, for anaerobic/anoxic/aerobic, or UCT for University of Cape Town process, both of which use different recycle mechanisms to add denitrification to the EBPR process (Tchobanoglous et al., 2003). Another system that is commonly used is known commercially as the PhoStripTM process, a “sidestream” process, where some of the return sludge is diverted from the main circuit, and run through a chemical phosphorus removal circuit before being added to influent (Tchobanoglous et al., 2003). The origins of this phosphorus-stripping strategy lie with the early work by Shapiro (Rybicki, 1997).  1.3.2 Overview of the EBPR process  Enhanced  biological  phosphorus  removal  involves  accumulating  phosphorus from influent wastewater into biomass, so that it can be removed. In  5  order to do this, conditions in the reactor must provide a competitive advantage to organisms that are capable of storing intracellular phosphate, termed polyphosphate accumulating organisms (PAOs). These conditions are provided by an alternating anaerobic-aerobic cycle, where the PAOs utilize energy previously stored as polyphosphates to accumulate carbon during the anaerobic phase, and then accumulate up to 10% of the entire sludge mass as polyphosphates (Serafim et al., 2002) using this stored carbon as an energy source in the nutrient limited aerobic phase. The stored polyphosphate is then removed from the system through sludge wasting.  Figure 1.1 In alternating anaerobic/aerobic conditions, polyphosphate accumulating organisms (PAOs) will take up carbon compounds and release negatively charged P and cations in the anaerobic environment, and will take up free P in solution and cations in the aerobic environment.  The process details are as follows. Influent wastewater enters the anaerobic phase, containing biodegradable soluble carbon (bsCOD). This  6  influent wastewater carbon begins to be fermented by chemoorganic fermentative bacteria to volatile fatty acids (VFAs): bsCOD is the most easily converted, but some undissolved (particulate and colloidal) COD may ferment as well (Tchobanoglous et al., 2003). The lower molecular-weight fermentation products (acetate is the most studied of these in EBPR) are assimilated by PAOs, using stored polyphosphates as an energy source, an advantage over other heterotrophs lacking this stored energy resource, which are starved under these conditions. In experimental systems, 7-10 mg of acetate can yield about 1.0 mg of P removal (Wentzel et al., 1989, Wentzel et al. 1990). Acetate is converted to polyhydroxybutyrate (PHB) storage products, and other PHAs can be generated from other substrates (Pijuan et al., 2004). Stored glycogen serves as an electron source for PHA synthesis in these reactions (Pereira et al., 1996) This conversion depletes the amount of polyphosphate stored in the PAO cells, as well as consuming some glycogen, releasing phosphate into the mixed liquor as orthophosphate (O-PO4), and Mg2+, Ca2+, and K+ cations (Tchobanoglous et al., 2003). The orthophosphate is taken up once the system is aerated, degrading the stored PHA reserves as an energy source for P accumulation, growth, and glycogen production. (Pereira et al., 1996). These polyphosphates are either maintained in the return sludge in the form of poly-P and glycogen rich PAOs that will use them for energy during the next cycle, or wasted to produce a net reduction of P from the system.  7  1.3.3 Interferences with EBPR  A lower P removal efficiency is found when O-PO4 is released at the wrong time, termed secondary release (Barnard, 1984). This occurs when O-PO4 is released without a corresponding generation of PHB from acetate due to either waste sludge being held under anaerobic conditions, or an extended anaerobic phase, for example, greater than three hours (Stephens and Stensel, 1998). In either of these conditions, the PAOs use their stored polyphosphates for energy, without storing the PHAs that are needed for O-PO4 reuptake, so the O-PO4 remains in solution rather than in the sludge biomass (Tchobanoglous et al., 2003).  Other interfering factors can include contamination of the anaerobic zone with significant amounts of dissolved oxygen or nitrate (rendering it aerobic or anoxic, respectively) wherein volatile fatty acids will be utilized by other organisms before the PAOs have a chance, and phosphate removal will be negatively affected. For this reason, denitrification must also accompany nitrification in these systems if the sludge is to be returned to the anaerobic zone or phase. This denitrification can ensure that levels of nitrate are not high enough in the anaerobic zone to prohibit EBPR (Tchobanoglous et al., 2003). As mentioned earlier, the processes A2O and UCT have a recycle mechanism that favours denitrification in an intermediate anoxic zone, as do other related processes such as Modified 5-stage Bardenpho, VIP, and Johannesburg (Tchobanoglous et al., 2003). For example, the A2O process recycles nitrified  8  aerobic effluent into the anoxic zone, minimizing the amount of nitrate in the return sludge. The UCT process recycles sludge to an intermediate anoxic zone, where denitrification occurs, and return sludge is taken from the anoxic zone, further limiting the amount of nitrate that enters the anaerobic zone. Additionally, SRTs longer than 4-5 days can cause enough methanogenesis to consume the VFAs. (Tchobanoglous et al., 2003).  A final interfering factor is the uneven flow of biodegradable material into the reactor, as periods of starvation change relative intracellular concentrations of glycogen, polyphosphate, and PHAs, which causes process failure. This uneven flow of rbCOD can be caused by cold, wet weather, and has been shown to decrease BPR performance (Stephens and Stensel, 1998): carbon source has been shown to have an effect on the process efficiency (Lemos et al., 1998). Prefermentation of sludge, or continuous acetate addition can aid the EBPR process (Tchobanoglous et al., 2003). The former strategy has been tested with success in Kelowna, BC (Oldham and Stevens, 1985) Although it should be noted that Beer et al. (2006) and He et al. (2008) showed that prefermentation is not positively correlated to the populations of bacteria held to be responsible for the EBPR process.  1.3.4 The role of cations in EBPR  Cations must be present in wastewater for EBPR to function: they stabilize the polyphosphate granules (Schönborn et al., 2001), and balance the solution’s  9  electroneutrality because they are co-transported with polyphosphate (Kim et al., 2007, from Comeau et al., 1987). Early work showed that Mg, K, and Ca were necessary to stabilize the PAOs’ polyphosphate granules, in the ratio Mg:K:Ca of 0.71:0.50:0.25 (Wentzel et al., 1989), and that they participate in phosphate storage in the molar ratios 0.28:0.26:0.09 per mole of phosphate. Domestic wastewater provides these conditions, but industrial wastewater may not. In particular, influent Mg is important for the long-term stability of EBPR (Schönborn et al., 2001, Seviour et al., 2003).  The roles of the individual cations have been investigated: Rickard and McClintock (1992) found that EBPR requires both of Mg and K, and Barat et al. (2005) found that K and Mg are cotransported with phosphate, but Ca levels remain constant throughout the EBPR process. These authors raise the issue that in order to balance the charge of polyphosphate, the sum of positive charges for a mole of P uptake must be balanced, and assuming along with activated sludge models that the ratios are (KaMgbPO3)n where a=b=0.33, most of the literature values for Mg/K cotransport ratios do not achieve this charge balance. The authors suggest that biologically induced precipitation occurs as calcium or magnesium phosphates, or struvite (MgNH4PO4·6H2O) under anaerobic conditions, but they rule out struvite precipitation based on the MINTEQA2 model, and calcium phosphate precipitation based on the findings that calcium concentrations remain constant throughout their experiments.  10  1.4 Microbiology of EBPR processes 1.4.1 Culture-based methods and Acinetobacter  Although at this point at least one PAO, Candidatus Accumulibacter phosphatis, has been consistently recovered from EBPR systems, the history of the search for such an organism is an interesting one. The first microorganism reported to be responsible for biological phosphorus removal was Acinetobacter (Fuhs and Chen, 1975), which still crops up in the literature from time to time. The short (1.5µm) Gram-negative γ-Proteobacterium is found in pairs, short chains, and clusters, and can be isolated from sludge performing EBPR. Other researchers followed in this vein, determining that Acinetobacter junii, Acinetobacter lwoffii, and Acinetobacter johnsonii were the most commonly found species of the Acinetobacter genus, at 30, 30, and 20%, respectively, and that the structure of the bacteria differed slightly depending on the aerobic, anoxic, or anaerobic conditions (Cloete et al., 1988a, Beacham et al., 1990, Beacham et al., 1992, reviewed by Rybicki, 1997). Other groups that isolated Acinetobacter from EBPR sludges included Deinema et al. (1985) and Streichan et al. (1990), and so during this period it was thought that Acinetobacter was the PAO responsible for performing biological phosphorus removal. Investigators attempted somewhat unsuccessfully to culture the clustered cells from EBPR systems: only a few of the observed cells grew in culture, and those that did were not always Acinetobacter (Duncan et al., 1988, Beacham et al., 1990). The nonAcinetobacter isolates were not identified, because of the belief that  11  Acinetobacter cells were the PAO in EBPR systems (Seviour et al., 2003). Much work ensued, identifying further Acinetobacter spp. From activated sludge (Carr et al., 2001, Carr et al., 2003), and especially examining the physiology and biochemistry of A. johnsonii strain 210A (van Groenestijn, 1989, Bonting et al., 1992, van Veen et al., 1993, and Kortstee et al., 2000). Although many isolates were found to release P anaerobically, none have been found able to assimilate acetate, synthesize PHAs, and release P according to the biochemical models. These inconsistencies point toward the possibility that Acinetobacter did not behave the same in pure culture as in EBPR, or that Acinetobacter was not the major PAO in EBPR (Tandoi et al., 1998, Seviour et al., 2003). Blackall (2002) claimed that the evidence provided by Cloete and Steyn (1987), Hiraishi and Morishima (1990), Auling et al. (1991), Wagner et al., (1994), Bond et al. (1995), and Kämpfer et al. (1996) is strong enough to rule that Acinetobacter species are not PAOs.  In the wake of a seeming inability to culture and identify the PAO, chemical markers were used in an attempt to sort out the culture-based results. Fluorescent antibodies prepared against Acinetobacter indicated that only 3% of the cells present in an EBPR system were of this genus (Cloete and Steyn, 1988b), an amount that could have a sizeable impact on the EBPR process (Hesselmann et al, 1999), but one that is held to be an unconvincing report of Acinetobacter importance (Seviour et al., 2003). Chemotaxonomic markers were also used to determine EBPR sludge community composition, a relatively  12  unbiased culture-independent approach when compared to rRNA-based approaches (Hiraishi et al., 1998). Using diaminopropane as a biomarker indicative of Acinetobacter, the genus was found to be present only at low amounts in plants performing efficient P removal, higher in plants with less efficient removal (Auling et al., 1991). Similarly, since the quinine Q-9 is representative of the γ-Proteobacteria, which contains Acinetobacter: a plant performing EBPR would be expected to have higher levels of Q-9 than one that was not. Instead, it was found that Q-8, representative of the β-Proteobacteria was the most abundant (Hiraishi et al., 1998), followed by the Q-10 containing Actinobacteria and α-Proteobacteria, a group that contains the putative PAO Microlunatus phosphovorus. Only 3% of the Actinobacteria, however, were of this species. Studies consistently found that Q-8 and Q-10 were dominant in both EBPR and non-EBPR sludge (Hiraishi et al., 1998, Sudiana et al., 1998, Lin et al., 2000, and Liu et al., 2000), suggesting again that only a small proportion of the overall EBPR sludge population was responsible for P removal if the active organism was Acinetobacter (Seviour et al., 2003). In the Lin et al. (2000) study, it was found that although Q-8 and Q-10 were the dominant ubiquinones, increased P removal was correlated with increased Q-9 and decreased Q-8, suggesting that a  γ-Proteobacteria may be involved in EBPR, but cannot be a  major PAO.  Acinetobacter is, of course, not alone in its ability to store polyphosphate. Other organisms have been cultured from EBPR systems that exhibit this trait,  13  including Tetrasphaera species, Microthrix parvicella, Nostocoida limicola II, Microlunatus phosphovorus, and an isolate that has the morphology of a Lampropedia species. Of these isolates, the bulking and foaming Microthrix and Nostocoida both stain for poly-P, but there is too little information to provide evidence that they behave according to the biochemical model of EBPR (Seviour et al., 2003). Microlunatus phosphovorus comes closer to the model by assimilating high levels of P aerobically as poly-P, and releasing P anaerobically, but this organism does not assimilate acetate anaerobically, nor do molecular studies indicate that this is a dominant microbe in EBPR systems (Seviour et al., 2003). Finally, an organism identified as Lampropedia based on its morphology (which is different than associated with EBPR PAOs) assimilates acetate anaerobically to synthesize PHAs and simultaneously releases P, albeit with a low acetate uptake to P release ratio (Seviour et al., 2003).  1.4.2 Molecular methods for identifying PAOs  Molecular techniques became readily accessible to researchers in the last decade. Relatively established methods had been used to examine community structure in N removal systems (Schramm and Amann, 1999, Loy et al., 2002), and so applying them to P removal was a natural progression. The techniques remained prone to bias due to inherent problems with DNA extraction and PCR from activated sludge, discussed in Wintzingerode (1997), Schramm and Amann (1999), Theron and Cloete (2000), and Loy et al. (2002). This suggests that a  14  combination of DNA extraction techniques, followed by a pooling of sampled DNA would be a good solution, but because of potential biases in preparation, fingerprinting methods such as denaturing gradient gel electrophoresis (Muyzer and Smalla, 1998), single strand conformational polymorphism (Schweiger and Tebbe, 1998), or T-restriction fragment length polymorphism (Osborn et al., 2000) are not appropriate (Seviour et al., 2003) without the proper controls included.  An alternative that can circumvent DNA extraction is, of course, to use an in situ technique such as fluorescent in situ hybridization (FISH). Although the RNA probes required for FISH are now simple to design, the technique is not without its own set of difficulties. The probes themselves are limited by the sequence databases used to design them, which must exist beforehand from previously cultured organisms or cloned DNA. Similarly, the technique can only find organisms that it is targeting from the outset. In addition, although FISH can be very revealing in terms of the community composition of a sample, it cannot provide any data as to the function of members of the population on its own. A technique that can be combined with FISH to add this data is microradiography (MAR), as described in Lee et al. (1999) and Gray and Head (2001). The technique involves feeding radioactive substrates to a sample, which can then be tracked to a population identified using FISH (Seviour et al., 2003).  15  Functional gene analysis can also be used to determine whether a sample is likely to contain microbes capable of performing certain metabolic functions. Seviour et al. (2003) point out that this technique is sensitive to problems with DNA extraction from activated sludge, as well as being prone to false positives resulting from a particular gene being present but not being expressed.  Molecular studies have only recently studied full-scale plants (Pijuan et al., 2008), opting rather to use controlled, laboratory scale plants and artificial sewage. As such none of the full-scale molecular studies concern EBPR systems, and all such studies have ignored the possibility that Archaea make a contribution to P removal aside from Gray et al. (2002), where methanogens were detected in conventional sludge processes. The studies performed prior to 2003 are summarized in chart form in Seviour et al. (2003). These studies have shown that EBPR communities (as well as non-EBPR communities) are very diverse phylogenetically, and that although molecular 16s rRNA clone library studies do generate similar clones, these cloned sequences do not match up with data from either FISH studies of the same communities, or culture-based studies. Using DGGE it was found that community structure became less diverse as EBPR improved in a laboratory-scale SBR (Onuki et al., 2002), and using FISH, it was found that the diversity of the EBPR community structure was dependent on influent concentration: a complex influent composition resulted in the laboratory scale SBR exhibiting a community structure more akin to full-scale EBPR systems than a parallel acetate-fed SBR (Hesselmann et al., 1999).  16  Whereas the acetate-fed reactor was dominated by β-Proteobacteria, the complex influent-fed reactor was less so, and contained more Actinobacteria and CFB division bacteria. Further FISH investigation confirmed that Acinetobacter represented only a small part of the EBPR population, and the β-Proteobacteria and Actinobacteria reigned supreme in terms of numbers (Kawaharasaki et al., 1999, Mudaly et al., 2001, Kong et al., 2002, Wagner et al., 2002). Furthermore, when two Acinetobacter species were sequenced from activated sludge, their sequences did not match those from which the FISH probes had been developed (Snaidr et al., 1997). Later work did, however show that out of a range of Acinetobacter species, only a few responded to probes based on the new sequences, and all responded to the original probes (Carr et al., 2001, Carr et al., 2003, reviewed in Seviour, 2003). The authors of these studies are now focusing on Fourier transform infrared spectroscopy, an organism fingerprinting method, to bolster the ability to study Acinetobacter environmental isolates (Winder et al., 2004).  All of the FISH and quinone-based studies agree that Acinetobacter represent only a small fraction of the EBPR bacteria, the β-Proteobacteria are relatively ubiquitous in activated sludge, regardless of EBPR ability, and Actinobacteria ranges from dominance in a lab-scale reactor to very small numbers in an SBR, even when the same DNA extraction methods are used (Bond et al., 1995, Majone et al, 1999). An important observation from the Bond et al. (1995) study, was that the Rhodocyclus group from the β-Proteobacteria  17  was very highly represented in their 16s rDNA clone libraries. The same group later (Bond et al., 1999) used FISH to determine that more than half of the bacteria in efficient EBPR sludge were from the β-Proteobacteria-2 subgroup. Hesselmann et al. (1999) determined that the PAO from this group was closely related to the Rhodocyclus genus, and they dubbed the unculturable PAO ‘Candidatus Accumulibacter phosphatis’. Further work using FISH based on EBPR SBR clone libraries that contained Rhodocyclus relatives showed that clustered cells (a hallmark of EBPR) did indeed hybridize with the FISH probes (Hessleman et al., 1999). Importantly, the cells also stained with methylene blue, indicating the presence of intracellular poly-P, and the sludge P content was found to be highly correlated with PAO FISH probe binding (Crocetti et al., 2000). Using single-strand conformational polymorphism (SSCP), Dabert et al. (2001) noted that an increased number of Accumulibacter were present in functioning EBPR over deteriorated EBPR.  1.4.3 Accumulibacter and Rhodocyclus in EBPR  The PAO status of Accumulibacter gained strength from this point forwards: By separating out PAOs using differential centrifugation, Zilles et al. (2002) showed that Rhodocyclus relatives were a significant part of EBPR sudge, but not of non-EBPR sludge. Onuki et al. (2002), used DAPI staining for poly-P combined with FISH probes for the Rhodocyclus relative to show that the poly-P containing cells were morphologically similar to Accumulibacter, but were not  18  able to stain any particular clusters using both techniques to be certain that the same bacteria were Accumulibacter and accumulating Phosphate. Kawaharasaki et al. (2002) were also unable to stain individual clusters with both DAPI and FISH probes, rather much of the DAPI bound to CFB and α-Proteobacteria. It was generally maintained that it was not the sole contributor to EBPR, but rather that Accumulibacter was only a part of a group of PAO bacteria that remained to be identified (Kawaharasaki et al., 2002, Lee et al., 2002). (He et al., 2006) found that Accumulibacter-related bacteria comprised more than 80% of a lab-scale SBR, but only 9-24% of full-scale plants studied. Furthermore, Accumulibacter comprised 40-69% of the PAO population in full scale plants, and these PAOs demonstrated EBPR activity (He et al., 2008).  Accumulibacter gained widespread notice when Martin et al. (2006) published a metagenomic analysis of EBPR communities dominated by Accumulibacter, setting Accumulibacter up as the model organism (He et al., 2008) for the study of genetic and biochemical aspects of EBPR processes. Some members of the same consortium that performed the metagenomic analysis then went on to finer-scale investigation of Accumulibacter populations to determine whether there was a geographic effect: because the highly conserved 16s rRNA determination does not offer high enough resolution to differentiate between different Accumulibacter populations (He et al. 2006), the faster evolving 16S-23S internally transcribed spacer (ITS) in the rrn operon (He et al., 2006) which codes for ribosomal RNA were used instead to resolve five  19  clades within Accumulibacter from various lab and full scale WWTPs, and polyphosphate kinase 1 (ppk1) genes (He et al., 2007a), which code for the polyP synthesis were also used to resolve five clades. Closely related Accumulibacter genes are found at wide ranging geographies (He et al., 2007a), and a wider variety of Accumulibacter are found in full-scale over lab-scale plants, and the dominant clades appear to vary, which may be attributed to operational conditions, periodic clade selection, or viral invasion (He et al., 2007a). The authors suggest the use of the reinvented ecotype concept proposed by Cohan (2002) to describe the five Accumulibacter clades.  ppk1 is of particular experimental interest, as it is a single copy gene and can thus be used by quantitiative PCR to estimate cell abundance (assuming one genome copy per cell), whereas cells possess multiple copies of the rrn operon per genome (He et al., 2007a). ppk still cannot be used as a functional indicator of EBPR, however, as it has not been confirmed as the sole gene for polyP (He et al., 2007a), and with the hindsight provided by the culture-dependent Acinetobacter era, investigators in the EBPR field are cautious to accept a single PAO as responsible for EBPR. Non-Accumulibacter PAOs are found frequently, especially in full-scale WWTPs (He et al. 2006).  20  1.4.4 Other PAOs in EBPR  Contenders for PAO found by Melasniemi and Hernesmaa (2000) were large cells present in EBPR, thought to be yeasts, but doubt was shed on this conclusion by Chua et al. (2004), who found the large cells to be βProteobacteria. Actinobacteria have been found in some studies to be the dominant PAO in some WWTPs (Kong et al., 2005, Beer et al., 2006). Further investigation potential is found in this area , as the identities of PAOs in EBPR have begun to be unraveled, but work remains before the identities of the PAO community and how it operates can be considered solid enough to build process models on top of: an account of the generally good fit between biochemical models of EBPR based on enriched PAO cultures and full scale WWTP EBPR is found in Pijuan et al., (2008).  1.4.5 The GAO question is unresolved for EBPR  EBPR processes often behave unpredictably, and much of the work to figure out why they do so has revolved around the “G bacteria”. These phylogenetically diverse tetrad or cluster-forming 2-3 µm bacteria have been found in EBPR reactors since the early 1990s (Levantesi et al., 2002). The central hypothesis runs that these ‘G-bacteria’ (or GAOs, in this case) outcompete PAOs in anaerobic/aerobic systems by anaerobically assimilating carbon substrates more efficiently than the PAOs, eventually converting those to  21  PHA, and metabolizing the PHA under aerobic conditions but then not performing the task of aerobic poly-P synthesis like their PAO compatriots. Simply put, PAOs use glycogen as an electron source and polyphosphate as an energy source during the anaerobic period, GAOs use glycogen as both an electron and an energy source (Zheng et al., 2005).  Under certain conditions, particularly low pH (Oehmen et al., 2005, Zhang et al., 2005) the GAO populations began to dominate the EBPR system, after which it no longer performed phosphate removal well. This hypothesis has been tested, and evidence points towards its robustness (Liu et al., 1996, Liu et al., 1997). The search to find the GAO organism is well reviewed in Blackall et al. (2002), Seviour et al. (2003) and in Oehmen et al. (2007), where it is divided into sections of isolated GAO candidates, Gammaproteobacteria GAOs, and Alphaproteobacteria GAOs. The identities of organisms showing the GAO phenotype are still under investigation, and the competitive model is the subject of recent work such as Beer et al. (2004), Dabert et al. (2005), Oehemen et al. (2005), and Zhang et al. (2005). One bacterium exhibiting the GAO phenotype has been enriched and dubbed Candidatus Competibacter phosphatidis (Crocetti et al., 2002). It has been found in both lab scale and full-scale systems, but is not alone in expressing the GAO phenotype: a group of GAOs from the Actinobacteria have also been found, and these also have the tetrad morphology of G Bacteria (Oehmen et al., 2005). The terminology has evolved so that “Gbacteria” has come to mean the tetrad cocci in sludges, whereas ‘GAO’ refers to  22  a bacterium that exhibits the competitive phenotype (Blackall et al., 2002), and genotype in the case of Competibacter.  In contrast to the widespread acceptance of the GAO model, a group led by Zhiguo Yuan has found that in their analysis of Australian WWTPs, no GAOs are present in the sludge (Pijuan et al., 2008), and in the lab, PAOs can function as (although not necessarily compete with) GAOs under P limited conditions (Zhou et al., 2008). Further surveys need to be conducted to ascertain whether this is the case worldwide, and whether GAOs are laboratory artifacts or widespread occupants of WWTPs.  1.5 SBR systems in EBPR processes  Tchobanoglous et al. (2003) list the following positive and negative aspects of the SBR’s suitability for biological phosphorous removal: both N and P removal are possible, the process is easy to operate, Mixed liquor solids cannot be washed out by hydraulic surges, lower effluent TSS may result from quiescent settling, and SBRs have flexible operation. The process can also be easily adjusted to optimize plants to increase removal efficiency and decrease cost (Akin and Ugurlu, 2005). Negative aspects include more complex operation, a larger SBR container for P removal than N removal, a reliable decanting facility required for high effluent quality, a more complex design, a requirement for  23  skilled maintenance, and a suitability for smaller flowrates. (Tchobanaglous et al., 2003)  Figure 1.2 SBR operation cycle for EBPR.  In order for EBPR to be successful in an SBR, the sludge that remains in the container between decant and fill stages must be low in nitrate, so that the fill and anaerobic react periods can remain anaerobic. As mentioned above, if nitrate remains in the fill and react stage, nitrate-reducing bacteria can compete with PAOs for rbCOD, decreasing P removal (Tchobanoglous et al., 2003). Simultaneous nitrification/denitrification provides a convenient method for limiting  24  the amount of nitrate in the return sludge, providing its own nutrient removal, as well as aiding EBPR. Some PAOs are also capable of denitrification (KerrnJespersen and Henze, 1993), but may not perform EBPR as efficiently while denitrifying, indicated by an inflection point in SBR conductivity profile inflection point (Serralta et al., 2004). A more complete review of the process factors that affect EBPR is found in Mulkerrins et al. (2004).  1.6 Process control of EBPR Processes  On-line controls using direct measurement of wastewater treatment parameters such as BOD, COD, NO3-, NH4+, PO43- are difficult or costly to measure directly on-line (Lee et al., 2001), but these wastewater characteristics are related in various ways to pH, DO, and ORP. As such the latter three parameters are the most widely used parameters for on-line monitoring and control of biological nutrient removal processes (Akin and Ugurlu, 2005).  1.6.1 ORP – Nitrate knee indicates the end of denitrification  ORP tracks the oxidation state of aqueous solutions, and correlates with biological oxidation and reduction reactions such as those involved in nitrogen removal (Akin and Ugurlu, 2005) The end of denitrification causes a sharp decline in the ORP curve during the anoxic phase referred to as the “nitrate knee” which signals the beginning of anaerobic conditions, important to  25  phosphate release (Akin and Ugurlu, 2005). ORP is linearly correlated with the log DO concentration, so increases when aeration is started (Kishida et al., 2003).  1.6.2 ORP indicates P concentration  Kim et al. (2007) found a strong inverse correlation between ORP values and phosphorus concentration, and recommend its use for monitoring P concentration rather than or in addition to EC due to problems using EC under anoxic conditions.  1.6.3 DO Elbow indicates end of nitrification  DO indicates the depletion of ammonium substrate, indicating the end of nitrification where DO increases which can serve as a break point. It is not always detectable as there are complications with heterotrophic versus autotrophic nitrifiers (Akin and Ugurlu, 2005)  1.6.4 pH indicates various biological processes  pH  also  correlates  with  biological  reactions,  increasing  during  ammonification and denitrification, and decreasing during nitrification (Chang and Hao, 1996). pH can be used to determine the end of the aerobic phase of an  26  SBR by looking for an “ammonia valley” that corresponds with the end of nitrification (NH4  NO3, NO2), visually apparent on pH versus t plots, but detected by control systems as a switch from negative to positive value in the dpH/dt versus time plot (Kim et al., 2004), followed by a later plateau that corresponds with the end of phosphorus uptake (Akin and Ugurlu, 2005). Akin and Ugurlu report that the maximum phosphate uptake rate is reached only after a significant amount of nitrification. pH is not an effective measurement for controlling anoxic reactions as it does not clearly reflect the end of nitrification (Akin and Ugurlu, 2005)  Chang and Hao (1996) studied combined use of pH and ORP for combined N and P removal processes, work continued by Spagni et al. (2001), who show a “strict” (r2=0.5059) relationship between pH decrease and P-release once nitrate is depleted, and Spagni et al. (2001) suggest that because nitrification and P-uptake occur at the same time, but with different endpoints, pH can be used to monitor P-uptake if it finishes before ammonia oxidation is complete, but not if P uptake goes longer because the pH plateau reached after P uptake could be due to CO2 stripping, which raises the pH by decreasing the amount of carbonic acid in the system. Spagni et al. (2001) claim that conductivity measurement is similarly affected by carbon dioxide stripping.  27  1.6.5 pH control, and pH effect on P removal  Serafim et al. (2002) implemented pH control of their reactor, and found that pH flux increased PAO advantage over GAOs at high pHs. They used a combination of acetate, butyrate, and propionate in this study. This observation is supported by another study by Zhang et al., (2005), who found that in a pH controlled reactor, even a slight change in the mixed liquor pH from 7 to 6.5 resulted in a decrease of removal efficiency from an initial steady state removing 99.9% P to a new steady state removing 17%, along with a drastic shift in microbial population and intracellular P storage.  1.6.6 EC is related to P release in EBPR  Electrical Conductivity is the measure of the ability of a solution to conduct an electric current. Since current is carried by ions in solution, it serves as a measure of the concentration of ions (Tchobanoglous et al., 2003, p. 56) It can be used to monitor denitrification and P release, but carbon dioxide stripping makes the EC/P uptake ratio difficult to correlate under aerobic conditions (Spagni et al., 2001), and initial denitrification complicates its measurement under anoxic conditions. In the literature it is held to be a useful measurement only under strictly anaerobic conditions (Kim et al., 2007).  28  Maurer and Gujer (1995) showed that EC can be used to monitor Prelease after initial denitrification has occurred in batch experiments. Serralta et al. (2004) showed that inflection points in the conductivity increase curve corresponding to P release could be used to monitor the P release rate, and to detect whether PAOs were involved in initial denitrification.  Aguado et al. (2006) showed that the relationship held true throughout SBR cycles. Kim et al. (2007) maintained that the P-release:EC correlation held only under strictly anaerobic conditions, and was negligible during initial nitrification. (Arvin and Krisensen, 1985; Röske et al., 1995). Others showed that anaerobic phosphorus release was related to the metal cations K+, Mg2+, and Ca2+ by the following Equation:  [  Error! Bookmark not defined. M 1k−+k PO3−  ]  n  + nH 2 O  → nH 2 PO4− +  n k+ M k  where M represents the metal cations K+, Mg2+, and Ca2+, n is the number of mols of each species, and k is the charge on the metal cation. Following Aguado et al. (2006), Kim et al. (2007) hypothesize that according to the EC equation EC in siemens per metre = ∑ Λ+c ,i ⋅ z i+ ⋅ ci+ + ∑ Λ−c ,i ⋅ z i− ⋅ ci− (S·m-1) i  i  where Λc equals the molar conductivity of a given ion in S·m2·mol-1, z equals the charge of that ion, and c equals the concentration in mol·m-3, all of the ions that are present in most wastewater treatment plants such as HCO3-, HPO42-, H2PO4-, PO43-, NO3-, NH4+, CH3COO-, K+, Mg2+, Ca2+ , not to mention the complication of these ions changing species in response to pH fluctuations, have a complex  29  effect on mixed liquor conductivity such that EC alone is not sufficient to describe changes in P, despite the strong correlation found between EC and P. Aguado et al. (2006) derive the following equation ΔEC in siemens per metre = f a ∑ Λ o ,i ⋅ z i ⋅ ∆ci (S·m-1) i  where Λo equals the equivalent molar conductivity of an ion in S·m2·mol-1 which includes a concentration factor fa to take account of the behaviour of ions in the SBR: fa =  Λc ≤1 Λ0  Maurer and Gujer (1995) added this concentration factor to the EC equation which produced an equation for the theoretical increase in conductivity per concentration of released phosphate (Λp): ΔΛP in S ⋅ m 2 ⋅ g P−1 =  fa ⋅ ∑ Λ o ,i ⋅ | z i | ⋅∆ci 30.974 i  ( S ⋅ m 2 ⋅ g P−1 )  where i represents the various ion species listed above.  EC measurement is not without some complications: it provides a low signal to noise ratio in saline wastewaters (Maurer and Gujer, 1995), and must be combined with pH in order to estimate anaerobic P release. With this in mind, however, conductivity is simple to measure (as long as care is taken with signal isolation), and has been shown to be a good indicator of aspects of the EBPR process at the lab scale such as the beginning of P release, and the decrease in P release rate (Maurer and Gujer, 1995).  30  1.7 Why study a pilot-scale plant?  Biofilters occur when a biofilm attaches to a solid surface, and reacts by diffusion with the surrounding liquid phase. Effects caused by the reactor acting as a biofilter must be avoided in order to simplify scale-up, but lab-scale reactors have an inherent tendency to act as biofilters more than full-scale plants due to differences in the ratio of surface to volume. A biofilter could form on the inside surface of the reactor, and the surface of any plumbing or probes inside the container. The flow rates inside the major plumbing should be high enough to mitigate this effect, but the biofilter effect should not be entirely ruled out. The pilot scale reactor used in this investigation was on the order of one hundred times larger than a lab-scale reactor, and thus reflected an intermediate between the lab scale reactors of other investigators, and a full scale WWTP. In fact, the reactor used here could serve as a small functional WWTP for a resort or small community, and thus gives a much more realistic picture of an actual activated sludge process free of biofilter effects.  In addition, this study used actual domestic wastewater from the University of British Columbia, which is a more complex substrate than the typically acetate-based artificial wastewaters used in laboratory studies, and was thus more akin to domestic wastewater fed to full scale WWTPs.  31  2 Materials and Methods  2.1 Preliminary data analysis The SBR reactors used in this study were automatically controlled, and the control system generated a large amount of process data. Analysis was later carried out on samples of effluent and influent. We were interested in the relationship between certain control parameters and the results on the analytical data. Because the process control and analytical result datasets were not compatible, charts that combined the analytical and process data were created, and analytical data were manually entered into the process data spreadsheets. Comparison tests were performed on the combined data.  2.2 Prototype on-line P and N analyzer setup  In order to generate larger datasets, a sampling apparatus and analyzer bank allowing on-line analysis of nitrate and phosphate of reactor effluent, or near-continuous analysis of mixed liquor conditions was constructed. Samples were drawn from inside the SBR, filtered using a membrane module, and pumped to the analyzers (Figures 2.1-2.3). Nitrate was measured using a Dr. Lange Prozeβhotometer-nitrat (Hach Company, Loveland CO), and phosphate using a Hach Series 5000 Phosphate Analyzer (Hach Company, Loveland Colorado)(Figure 2.4). EC was simultaneously measured using a retrofitted  32  probe for an Orion model 180 EC sensor (Orion now Thermo Scientific, Beverly, MA) (Figure 2.5), which first recorded to a chart recorder, later to a PLC.  33  Air pump Pressure gauge  Sample pump  Air spurge line  Membrane module Overflow 1 NO3 Analyzer  Conductivity Chart meter recorder  To drain Overflow 2 Ortho- Phosphate analyzer P Membrane vacuum  To drain  Diaphragm valve  Permeate pump P Sample pressure  To drain  Pressure regulator  To drain  Holding tank Manual samples  Conductivity cell  SBR “ A”  Figure 2.1 Setup of automatic P and N sampling system for the pilot scale SBR. Details of various components in Figures 2.2-2.5  34  Figure 2.2 The two tank SBR (left and right) with sample filtration membrane unit (center)  35  Figure 2.3 Top: EC sensor and chart recorder (later replaced by PLC) Bottom: sample pressure regulator assembly.  36  Figure 2.4 On-line analyzer bank - Dr Lange Prozeβhotometer-nitrat (top), and Hach Series 5000 Phosphate Analyzer (bottom).  37  Figure 2.5 EC probe with bubble shield showing buildup of colloidal material within EC cell: the bubble shield was later replaced with the insertion assembly seen in Figure 2.6 when the chart recorder was replaced by a connection to the PLC.  2.3 Track Studies  2.3.1 Reactor Setup and Operation  The reactor was set up and operated as above, with samples drawn manually through the decant line (Figure 2.1).  38  2.3.2 Data Collection  Data collection was threefold: on-line data were collected and logged to a PLC over the duration of the study, daily influent and effluent samples were collected to monitor overall reactor performance, and during the track study samples were collected for analysis.  2.3.3 On-line data collection  DO, ORP, pH, and EC data were collected once a minute and logged to PLC with the following sensors: DO - Point Four Systems OxyGuard DO probe (Point Four Systems, Coquitlam BC) ORP - Sensorex flat surface, self cleaning submersible process ORP probe (part # 970212, Sensorex, Garden Grove CA) pH - Sensorex flat surface, self cleaning submersible process pH probe (part # 970211, Sensorex, Garden Grove CA) EC – Orion model 180 (Orion now Thermo Scientific, Beverly, MA) with a conductivity cell, cell constant 1.0 cm-1, retrofitted for submersible operation (Figures 2.5 and 2.6). This included an insertion assembly: this cap allowed the probe to be inserted past the foamy top layer of the SBR without becoming fouled with colloidal material as it is in Figure 2.5. Unlike the bubble shield in Figure 2.5, this assembly was removed manually as soon as the probe was  39  inserted. Bubble noise was a problem when using the analog chart recorder because it caused the stylus to move up and down quite rapidly, resulting in a solid smear of markings. This was not a problem when using the PLC which gathered discrete data points from the EC sensor, so a bubble shield was not necessary during the track studies. Automatic temperature compensation was not possible, and linear temperature compensation was performed manually according to the formula ECt = EC25[1+a(t-25)] Where ECt is the EC in μS at the recorded temperature, EC25 is the sensor reading in μS, t is the recorded temperature in oC, and a temperature coefficient a=0.021 was determined to be appropriate for our wastewater.  40  Figure 2.6 Orion conductivity cell retrofitted for submersible operation.  41  2.3.4 Track study sampling  Daily influent and effluent samples were collected and analyzed for COD, O-PO4, NOx, and NH4 (see Figure 2.7) Samples to be prepared as per Figure 2.7 were taken through the decant system of the SBR (see Figure 2.1): 6L was flushed through the line, and a 1L sample was collected. This sample was stirred, measured for temperature, and divided into two 50 mL Falcon tubes. Two mL of this sample was aliquoted into a COD vial, and the remainder was centrifuged for 1 minute to settle out solids. The pellet was discarded and the supernatant was filtered through a 1.2μm glass filter, and distributed into 5mL tubes for O-PO4, NOx, NH3 analysis, with two drops of phenylmercuric acetate preservative added to PO4, NOx samples, and two drops 5% H2SO4 added to NH3 samples. Two mL of filtered sample were preserved with 2% H3PO4 for VFA analysis. One hundred mL of filtered sample was acidified to less than pH 2 by adding 0.5mL of concentrated HNO3 for cation analysis. O-PO4, NOx, NH3, and VFA analysis was performed by Paula Parkinson, UBC Environmental Engineering lab (graduate students are not permitted to perform these analyses).  42  Sample  Unfiltered to COD tube  Centrifuge, filter super’t  Filtrate  Preserve with phenylmercuric acetate  O-PO4, NOx analyses  Preserve with H2SO4  NH3 analysis  Preserve with H3PO4  VFA analysis  Preserve with HNO3  Digest in HNO3, filter  Mg2+, Ca2+, K+ analyses  Figure 2.7 Sample preparation and analysis scheme for track studies performed on the pilot SBR.  43  2.3.5 Cation Analysis  Acidified cation samples were digested in concentrated HNO3 until clear, refiltered at 25µm with a Whatman 541 HA filter, and diluted back to their original volume. The samples were then analyzed by atomic absorption for Mg2+, Ca2+, and K+ as per Standard Methods (1998).  2.3.6 COD analysis  Sample vials were prepared as per Standard Methods (1998) for closed reflux colorimetric analysis, with low range 0-150mg O2/L for effluent testing, and high range 0-1500mg O2/L for influent testing.  Two mL of unfiltered sample were pipetted into a COD tube and mixed with the contents for storage. Samples were digested as per Standard Methods (1998) and absorbance was read at 600nm and 420nm for high range and low range, respectively using a Hach DR2800 spectrophotometer. Standard potassium hydrogen phthalate solutions (Standard Methods, 1998) were used to produce a curve used to translate the observed absorbance values into COD.  44  2.3.7 Statistical analyses  Correlations were made using the equations (Fan, 2006):  where  x=  1 n  ∑x i  i  and  y=  1 n  ∑y i  i  3 Results and Discussion  3.1 Preliminary Research  3.1.1 Data Analysis  Some of the comparisons between analytical data sets appeared to show a relationship: for instance influent and effluent ammonia were correlated at r2=0.71, and influent COD was correlated to effluent COD at r2=0.56: not particularly useful information vis-à-vis EBPR.  It was concluded that larger  datasets were needed, and the data mining would have to be more specific to certain areas of the SBR cycle and would have to address process failure.  45  Phosphorus removal failures present a problem from a regulatory perspective because in some cases so much P is released, the receiving body would be better off without the process at all during the failure period. These are thus the most important failures to address. This study hypothesized that these failures would fit into a definite “type”, whereas poor removal in general could be caused by various factors. An attempt was made to characterize the “type” of a P removal failure.  The datasets were scoured for examples of the P removal failure phenomenon, and the data around these events were analyzed more closely:  Sub optimal P removal values (<75% removal) were extracted from the analytical data from our reactor for the following dates:  2004 on 5 Apr, 14 Apr, 27 Jun, 9 July, 8 Nov, 12 Nov, 23/24 Nov, 14 Dec  2005 on 11 Mar, 15 Mar, 22 Mar, 21 Apr, 3/4 May, 10 May, 17/18/19 May, 4 July  The analytical data for these dates were graphed (see Figure 3.1). The graphs were qualitatively assessed to determine whether each of these failures fit a pattern: near perfect P removal for a sustained period, followed by a large release of P in the effluent (Table 3.1).  46  Table 3.1 P removal failures in the pilot scale SBR Date 5 14 27 9 8 12 Apr Apr Jun Jul Nov Nov % removal  23/2 14 4 Dec Nov  11 Mar  7.7* 58.5 32.0 72.7 57.9 52.4 17.4 26.3 -14.0  22 Mar  21 Apr  3/4 10 May May  37.7 52.9 -44  1719 May  4 Jul  60.7 46.5 -23.3  * assuming that P removal was good prior to the failure: no influent data were available but effluent levels were consistently low.  The individual dates were then examined to determine whether there were specific changes occurring around the failures that could be used to describe the failures, with a focus on nitrate concentration at the time of failure and speed of recovery (Table 3.2, Table 3.3). Failures with insufficient data points (9 Jul 2004, 10 May 2005, 17 May 2005) are not included.  47  Table 3.2 Details of greater than 50% failure of P removal events % date Nitrate Depletion Recovery removal (d) -44 3/4 May No data 1.5 2005 -23.2  4 Jul 2005  -14  11 Mar 2005 Lower NO3 EFF No NO3 INF data 5 Apr 2004 No data  <4  Low CODINF 1 at time of failure NO3 EFF dips with failure PINF dips at failure, amplifying effect CODINF lowers at time of failure, but continues to drop after recovery NH3 EFF, NO3 EFF differ between reactors, PEFF does not. NO3 EFF dips with failure  1  May be associated with a PINF increase  23/24 Nov Low NO3 EFF 2004 14 Dec 2004 NO3 EFF declines following failure 27 Jun 2004 Slightly lower NO3 EFF  No data  Low PINF Low NO3 EFF CODINF spike NH3 EFF increases Failure likely occurred over weekend  22 Mar 2005 NO3 EFF increases at failure  No data  7.7 17.4 26.3 32.0 37.7  1  Other Changes  Slightly elevated NO3 EFF NO3 INF steady  1.5  1 1?  Table Legend: CODINF : Influent COD; NO3 EFF: Effluent NO3; PINF: Influent P; PEFF: Effluent P; NO3 INF: Influent NO3; NH3 EFF: Effluent NH3  48  Table 3.3 Details of less than 50% failure of P removal events % date Nitrate Depletion removal 52.4 12 Nov NO3 INF dips 2d prior, then maybe again over 2004 weekend, NO3 EFF decreases in subsequent week 52.9 21 Apr NO3 INF increases at failure 2005 57.9 58.5  8 Nov 2004 14 Apr 2004  NO3 INF low in previous week NO3 EFF high on this day NO3 INF steady No NO3 EFF data  Recovery Other Changes >7 <4  >12 1-2  Process not working well prior to failure Apparent failure largely due to lower PINF not an actual failure Low PINF amplifies effect: poor removal spread out over 2 weeks PINF spike: possibly an overload  49  No consistent “type” was found for P removal failures: nitrate depletion was not universally seen, nor was COD a consistently definitive factor as data were missing in many instances. Recovery from failure was consistently rapid (11.5d), although many recoveries occurred over the weekend making it impossible to determine the exact time for recovery. Relapses into poor failure were seen to occur despite consistent rapid recovery.  A lack of data points on weekends hampered analysis due to the relatively long time intervals they represent: numerous P removal failures occurred at the beginning or end of the week, and thus no failure trough (or PEFF peak) could be resolved (Figure 3.1) due to the lack of data points. Plots that connect data points between days with no samples are commonly used but are misleading, particularly if the lines are smoothed, as they indicate to the reader that a trough or peak has been found. It is clear from Figure 3.1, however, that the maximum release of PEFF could have occurred at any point over the weekend, and was not necessarily found by testing on the following Monday. Based on the existing plots, in order to resolve the P failure pattern a sampling rate of 2 points/day including weekends could be adequate in order to compare analytical data sets. The sampling rate would have to be increased if the data were to be correlated with PLC data, ideally in the form of a track study over the course of a normal cycle, and then over the course of a cycle where P release occurs without full reuptake. An online P analyzer could aid in data collection for this purpose,  50  particularly if it is used to sample mixed liquor in conjunction with a sample multiplexer and P level alarm, which would allow for the collection of samples prior to a P removal failure.  120  100  % P removal  80  60  40  20  0 17-Jun08  19-Jun08  21-Jun08  23-Jun08  25-Jun08  27-Jun08  29-Jun- 01-Jul-08 03-Jul-08 05-Jul-08 08  Date  Figure 3.1 Some typical daily analytical data showing the P removal ((PINF-PEFF)/PINF x 100%) for a sequencing batch reactor indicating failure of P removal, but unable to resolve the bottom of the failure trough.  The Municipal Sewage Regulations (MSR) for British Columbia indicate that O-PO4 for effluent going into all receiving bodies excluding open marine must be below 0.5 mg/L for the class of daily flows up to 2.0 times ADWF, so in a case such as this, even a drop to 73% removal such as 9 Jul 2004 produced effluent with twice the acceptable levels of O-PO4. Ground infiltration guidelines are generally less stringent, but some plants aim for even higher levels of 51  treatment. For example, Banff National Park is aiming for <0.15 mg/L total P, in which case it would be very important that the process operate at high efficiency all of the time.  Although the release of an effluent P spike would average out to be very small over time, if the effluent were sampled during a process failure, the process would not meet requirements. Furthermore, it is the intermittent release of large quantities of P that can contribute to deleterious eutrophication, more so than the gradual release of low levels, which would cause a slower, more balanced eutrophication. Thus, representing the effluent P spikes as part of an average hides their actual potential to cause environmental damage.  The proposed remedy for the lack of resolution situation was a combination of track studies and on-line effluent P and N monitoring.  3.1.2 The need for on-line analysis of P and N Removal  The literature has only recently begun to include long-term studies at a greater than bench-scale that monitor the P removal process and its associated parameters (Pijuan et al., 2008). This limitation is of particular importance for an understanding of the P removal failure phenomenon, which occurs unexpectedly, and has not been satisfactorily studied in situ. It has been argued in the literature that the long-term stability of EBPR is negatively affected by the shortsightedness  52  of process models, which treat the reactor system as a chemical process rather than a biological process (Yuan and Blackall, 2002).  The daily sampling and analysis protocol used to monitor the pilot scale SBR influent and effluent P and N did not have sufficient data to resolve the peaks of what may be failure events, as illustrated above. In order to obtain a more detailed dataset, on-line N and P analyzers capable of taking hourly samples, or at least a single sample per SBR cycle are a possible solution.  Although constant monitoring of N and P levels is prohibitively expensive for smaller WWTPs (Aguado et al., 2006), the goal of our investigation was to determine whether a P removal failure can be linked to specific, more easily and cost-effectively monitored process parameters such as pH, EC, or ORP, which would provide a useful tool for WWTP operators.  In particular, one of the objectives of conductivity control application was the prevention of process failure events: it was hypothesized that a failing P removal system would possess a different EC pattern than a working one.  3.1.3 The need for track studies  Since the type of direct measurement of EBPR outlined above is not costeffective for smaller WWTPs (Aguado et al., 2006), and an affordable proxy  53  measurement is sought, it must be demonstrated that the relationship between the EBPR process and EC exists at larger scales, and also that it is associated with changes in the concentration of the ions that are correlated to the EBPR process. The literature lacks the necessary statistical analysis of the relationships between the various participants in EBPR and EC. Detailed track studies of the process could provide this information.  Furthermore, Serralta et al. (2004) described timely addition of VFA at the end of the anoxic period, so that these VFAs were only available for PAOs, not for denitrifying heterotrophs: conductivity measurement would allow an operator to tell whether initial denitrification were caused by PAOs or heterotrophs, as PAO involvement would cause an inflection point in the conductivity curve, another detail that could be resolved by track studies of our process with a view to the implementation of an EC based EBPR control algorithm.  3.2 Prototype Analyzer Setup  The prototype analyzer setup did not yield any useful data: the configuration was unsuitable for the continuous monitoring of our process, but with some modifications similar setups could be used in future process monitoring.  54  The major shortcomings of the setup were threefold. The sample volume required for the N and P analyzer setup was too high without recycle, and recycle would constantly filter the mixed liquor resulting in a large sampling effect. Figure 2.2 shows that the membrane unit is relatively large when compared to the SBR tanks. This problem would not occur when running the sampling setup at a fullscale rather than pilot-scale plant, where the sampling volume would be negligible compared to the overall reactor volume. For a pilot or lab scale study, the sampling setup would need to be reworked to avoid using a large volume sample. Furthermore, the nitrate analyzer was not appropriate for our pilot plant due to high interior humidity. This resulted in a requirement for constant maintenance of the desiccator unit on the analyzer. A field-type analyzer could be used if shielding the analyzer from the elements proved impossible. Finally, it took a long time for a sample spike to reach the analyzers. This would make it difficult to match up the various readings of the on-line analysis of pH, EC, and ORP with results from the N and P analyzers. There was no simple remedy to this problem: the mixed liquor samples must be filtered before they reach the analyzers. Also, sparging the membrane filters with air made it difficult to accurately measure anaerobic samples without adding oxygen to them. Constructing a composite sampler that was able to prepare and store samples (according to Figure 2.7) quickly is one solution, but was not possible in the limited time for this study.  55  3.3 Daily Reactor Performance  As shown in Figure 3.2 (May-June 2006), the SBRs performed as expected. Ammonia removal was very consistent, and phosphorus removal was often very close to 100% (particularly in reactor B) despite random failures. After some of these failures the concentration of P was greater than the concentration of the influent, suggesting the release of biologically stored phosphorus. Sensor data demonstrated the expected patterns for ORP and pH (Figure 3.3). 150 100  % Removal  50 B P % removal  0  A NH3 % removal B NH3 % removal  -50  A P % removal  -100 -150 -200 06-05-15  06-05-25  06-06-04  06-06-14  06-06-24  06-07-04  Date  Figure 3.2 Performance of twin pilot-scale SBR fed with domestic wastewater for the period May and June, 2006. Arrows indicate negative % removal events.  56  Figure 3.3 Typical Profiles for ORP (bottom line), pH (middle line), and EC (upper line) in the pilot scale SBR. Illustrated for one cycle Ax/An indicates Anoxic/anaerobic conditions, AO indicates aerobic (aerated). The arrows indicate the ORP “Nitrate Knee” that signals the switch between anoxic and anaerobic conditions following the depletion of initial nitrate.  3.4 Track Studies  3.4.1 Pseudo-steady states:  Process control data were constantly logged to a PLC throughout the periods surrounding the track studies. Figures 3.3-3.5 show the EC and ORP profiles: the shapes of the ORP profiles indicated that the process was operating as expected. The conductivity profiles also showed a consistent shape when the  57  sensor was not fouled. Both track studies 4 and 6 began with a cleaning of the sensor, visible on the charts as a return to the periodic EC profile (Figures 3.4  300  150  250  50  200  -50  150  -150  100  -250  50  -350  ORP (mV)  EC (uS)  and 3.5.  EC ORP  0 -450 22-Jul-06 23-Jul-06 23-Jul-06 24-Jul-06 24-Jul-06 25-Jul-06 25-Jul-06 26-Jul-06 26-Jul-06  350  0  300  -50  250  -100  200  -150  150  -200  100  -250  ORP (mV)  EC (uS)  Figure 3.4 EC and ORP profiles leading up to and including track study 4: EC sensor becomes fouled prior to the study.  EC ORP  50  -300  0  -350  202121222223232424Aug-06 Aug-06 Aug-06 Aug-06 Aug-06 Aug-06 Aug-06 Aug-06 Aug-06  Figure 3.5 EC and ORP profiles leading up to and including track study 6: EC sensor is fouled prior to the study.  58  3.4.2 Phosphate concentration versus EC  If EC is to be used as a proxy indicator of phosphate concentration, it follows that the two must be highly correlated under the conditions where it is to be used as such.  Table 3.4 Coefficients of correlation between EC (uS) and P concentration (mg/l) for seven track studies of nonconsecutive sequencing batch reactor cycles. Track Study 3 4 6 7 8 9 10  Coefficient of Correlation (r) 0.93 0.94 0.95 0.92 0.90 0.87 0.94  Table 3.4 shows that P and EC were indeed highly correlated, but these correlations of course only give part of the story. From scatterplots of the data (Figure 3.6), it was clear that the slope of the best fit line was not consistent for all track studies, and the data were translated along the X axis between track studies. In order for these best fit lines to be congruent, first the overall change in P concentration would have to be uniform for all cycles. Because this study used domestic wastewater with varying influent P concentration this was impossible. It was expected that data in Figure 3.6 would differ with respect to the Y axis for this reason.  59  9  TS3  8 TS4  7  P (mg/l)  6  TS6  5 TS7  4  3  TS8  2 TS9  1  0 150  170  190  210  230  250  270  290  TS10  EC (uS)  Figure 3.6 Scatterplots of [P] vs. EC in seven nonconsecutive sequencing batch reactor cycles identified as TS 3 through TS 10.  60  From Figures 3.3-3.5 it is clear that EC values tend to fluctuate between SBR cycles. This pattern was consistent throughout the data, and could be explained by a gradual change in the concentration of an ion uninvolved in EBPR, or by sensor drift. This fluctuation explains the translation along the X (EC) axis between cycles in Figure 3.6. More important (from a process control standpoint) than these gradual fluctuations was the relative change in conductivity due to P removal over the course of a cycle.  The ratio of P/EC correlation was not as strong during the fill and decant stages as it was during the intervening stages. A major P removal failure (negative % removal, see Figure 3.2) should, however, result in a high EC after settling. This is because following a P removal failure, the released P is of PAO origin, and has associated ions that will increase the EC of the reactor supernatant. This differs from the influent P, in that the P:EC ratio has not been established in the influent (with the exception of studies that use synthetic wastewater with predetermined P:cation ratios). Once the P is stored as poly-P in the PAOs, it is co-transported with cations when released, raising the mixed liquor EC. So whereas influent EC could not be used to determine P concentration in influent, EC could be used to signal a failed re-uptake of PAOderived P. Unfortunately this cannot be shown from the data in this study, as the collected EC data were unreliable on their own due to constant fouling of the sensor, and none of the track studies demonstrated P removal failure.  61  3.4.3 Phosphate concentration versus cation concentration  The relationship between P and the EBPR associated cations has been observed at the lab scale (Table 3.5), and this study investigated whether the established ion/P ratios held at the pilot scale. Figure 3.7 gives typical ion profiles for a track study, along with a typically shaped conductivity profile.  10  300  9 8  Ion concentration (mg/l)  Conductivity (uS/cm)  250  200  150  100  7  Orion EC  6  PLC EC  5  PO₄ ³⁻ mgP/l  4  K⁺ mg/l  3  Mg⁺ ⁺ mg/l  2  50  Ca⁺ ⁺ mg/l  1 0 16:48  0 17:16  17:45  18:14  18:43  19:12  19:40  20:09  Time  Figure 3.7 Conductivity, PO43-, and cation profiles for K+, Mg++, and Ca++ over the course of a track study performed on a pilot-scale sequencing batch reactor: “Track Study 6”. The conductivity was measured manually with an Orion EC meter, and logged to a PLC from the same meter.  62  The ionic data did not clearly follow the conductivity profile, which is consistent with the literature. With a few exceptions, notably Maurer and Gujer (1995) and Kim et al. (2007), the literature tendency is to state that profiles with similar shapes are correlated, without statistical analysis. Comparison tests were performed in this study (Figures 3.8-3.10), which bear out the relationship r(Mg:P)>r(K:P)>r(Ca:P) in the three track studies for which ion data is available (Figure 3.11). There was a strong correlation to magnesium, a moderate correlation to potassium, and a highly variable weak correlation to calcium.  Many lab studies have recorded the Mg/P and K/P co-transport ratios (Table 3.5), with a view to being able to understand the stoichiometry of the EBPR process, as well as provide values at which these ions may limit the process. The goal of these determinations was to improve the ASM2d model (Barat et al., 2005), which would aid in the design of better plants, as well as better process control parameters.  63  9 8  K ion concentration (mg/L)  7 6 5 4 3  y = 0.2155x + 5.397 R2 = 0.3893  2 1 0 0  1  2  3  4  5  6  7  8  Phosphorus concentration (mg/L)  Figure 3.8 P concentration vs. K concentration over the course of a single track study: “Track study 6”.  Mg ion concentration (mg/L)  1.4 1.2 1 0.8 0.6  y = 0.068x + 0.6211 R 2 = 0.7575  0.4 0.2 0 0  1  2  3  4  5  6  7  8  Phosphorus concentration (mg/L)  Figure 3.9 P concentration vs. Mg concentration over the course of a single track study: “Track study 6”.  64  4  Calcium ion concentration (mg/L)  3.5  3  2.5  2  1.5  y = 0.0434x + 2.8275 R2 = 0.1322  1  0.5  0 0  1  2  3  4  5  6  7  8  Phosphorus concentration (mg/L)  Figure 3.10 P concentration vs. Ca concentration over the course of a single track study: “Track study 6”.  1  Correlation Coefficient  0.8 0.6 0.4 corr[P]:[K] 0.2  corr[P]:[Mg] corr[P]:[Ca]  0 -0.2 -0.4 -0.6  Figure 3.11 Correlations between P concentrations in mg/l and cation concentrations in mg/l over the course of three nonconsecutive SBR cycles. 65  Table 3.5 Values for the gram to gram ratio of Mg++ to P and K+ to P observed in this study and reported in selected literature (adapted from Barat et al., 2005* and Kim et al., 2007). Source Mg/P Rickard et al. 1992 0.24 Jardin and Pöpel 1996 0.27 Wild et al. 1997 0.25 Brdjanovic et al. 1996 0.26 Comeau et al. 1987 Miyamoto-Mills et al. 1983 0.20 Arvin and Kristensen 1985 0.25 Imai et al. 1988 0.28 Maurer and Gujer 1995 0.25 Aguado et al. 2006 0.28 Kim et al. 2007 0.18 This study 0.09 *originally expressed in molar concentrations rather than grams  K/P 0.27 0.33 0.42 0.42 0.34 0.29 0.37 0.35 0.46 0.40  Table 3.5 shows that although within the range of 0.27-0.46 the observed ratio of K to P was consistent with the literature, the observed ratio of Mg to P was less than half the ratios that have been shown at lab scale. Taking the Kim et al. (2007) study as an example, however, the absolute concentration of Mg observed in this study was about 10x lower overall, whereas the concentration of K was 1/5 the concentration that they reported. The concentrations of P were similar.  The lab-scale studies were fed high amounts of cations in artificial sewage, where the soft Vancouver water used in the feed for our reactor should have lower cation concentrations. These two factors could explain why the Mg value was lower than reported in the literature, if the ion concentration of the 66  influent has an effect on how P and cations are co-transported. This study does indicate that the EBPR process can function with lower concentrations of Mg than the values reported in previous studies.  Although Ca is an important constituent of the polyphosphate granules, only Mg and K are co-transported with phosphate, leading some investigators to speculate that there exist “reactive” Mg and K associated polyphosphates, and “inert” Ca associated phosphates (Schönborn et al., 2001, Liu et al., 2005). Thus the negative correlation between Ca and P in the single track study where it was found may have been caused by some effect external to the EBPR process. These external factors might have also contributed to the large variation in the weak correlations between Ca and P (Figure 3.10, 3.11).  3.4.4 EC versus cation concentration  The correlation of cation concentrations to P concentrations that were observed in this study occurred because the cations are important to the EBPR process as explained above. Since the ions are the factor that directly affects EC, rather than the larger phosphate ion, they must be correlated to P concentration at scales larger than lab scale if the models for EBPR are to apply in these larger systems. Figure 3.7 shows that the pattern of ion concentrations over a SBR cycle was similar to the pattern of EC over the same cycle, and according to the literature, they should affect EC according to the equation  67  ΔEC = f a ∑ Λ o ,i z i ∆ci (Sm-1) i  Where  fa =  Λc ≤1 Λ0  Mg concentration was the most highly correlated to P concentration as demonstrated above, but it was found in the lowest concentrations. In order for EC to serve as a proxy indicator of P concentration, the conductivity resulting from the most highly correlated ions must overshadow that from interfering ions such as acetate, bicarbonate, and even calcium. Maurer and Gujer (1995) found a lower fa value than expected in their experimental system, indicating a high level of ionic interactions but they still found P concentration to correlate highly with EC after initial denitrification in batch experiments for P release.  3.4.5 P concentration versus ORP  Kim et al. (2007) suggest that ORP may be a better indicator of P concentration in EBPR than conductivity, because the inverse ratio of ORP/P concentration holds under anoxic as well as anaerobic conditions, whereas EC only works under strictly anaerobic conditions after the disappearance of initial nitrate. This limitation to EC monitoring of P concentration has been widely noted, and is not surprising given that influent P has no reason to obey the P/cation ratios of the EBPR process, and the associated correlation with EC.  Only two track studies had data available for comparison of P concentration and ORP profile: 4 and 6. Track study 4 showed an inverse correlation between P and ORP (r=-0.91, expressed as R2 (equal to r2) in Figure  68  3.12), and although track study 6 initially appeared to show a lower correlation (r=-0.49, expressed as R2 (equal to r2) in Figure 3.13), it appears that there may be two separate interactions, one during the aerobic phase, and one during the anoxic phase, which could be explained by different equations.  Phosphorus concentration (mg/l)  6  5  4  3  2 y = -0.01x + 1.3353 R2 = 0.8239 1  0 -400.00  -300.00  -200.00  -100.00  0.00  ORP (mV)  Figure 3.12 Correlation between P and ORP during a track study on a pilot scale sequencing batch reactor: “Track Study 4”.  69  8  Phosphorus concentration (mg/l)  7 6 5 y = -0.0259x - 2.5508 R2 = 0.2415  4 3 2 1 0 -400  -300  -200  -100  0  ORP (mV)  Figure 3.13 Correlation between P and ORP during a track study on a pilot-scale sequencing batch reactor: “Track study 6”.  These two studies did not provide enough data to confirm that there is consistently a strong inverse correlation between ORP and [P] in our pilot plant, but they did agree with the literature, and they point towards a good avenue for future investigation.  3.5 Applications  P concentration and EC are correlated in this system, and the data is consistent with lab studies. Thus, even using real wastewater, EC can be used  70  as a proxy indicator of EBPR success. Applications include a chemical precipitant dose alarm, nitrate knee carbon dosing to favour PAOs, and improved algorithms to maximize energy savings.  3.5.1 Chemical precipitant dose alarm for poor P removal  In order to tell that P uptake is incomplete, an algorithm would have to be able to detect that the EC curve does not level off (Serralta et al., 2004). Although the inverse of the ORP curve may serve as a good indicator of anoxic, anaerobic, and aerobic P concentration, the EC gives a better picture at the settling phase. At this point, the ORP profile shows a spike, which then levels off during settling. The P concentration does not follow this trend, indicating that ORP measurement would be inadequate for a P release alarm, whereas conductivity should show if a massive amount of phosphorus and its corresponding ions are present in the final supernatant because the supernatant EC would be much higher than the influent EC. Experiments could be performed to determine an acceptable relative effluent EC that would not require precipitant dosing, and any higher EC would trigger a precipitant dose to ensure consistent effluent quality. The important observation from this study was the finding that although the absolute EC values varied between cycles, the EC profile maintained a characteristic shape when EBPR was occurring. This shape could be used to differentiate between a cycle with P re-uptake, and one without.  71  This simple addition to an EBPR process would prevent the potential release of a massive pulse of P, and thus prevent the extensive eutrophication of the receiving body that excessive P release could cause. This should be a rare, but important event, as EBPR SBRs usually operate at high removal efficiencies. As discussed above, however, a P pulse of this nature could be far more deleterious to a freshwater ecosystem than a constant low-level of P from an inefficient removal process, or even no removal at all, and this type of failure would be unlikely to be detected by routine sampling.  3.5.2 Nitrate knee C dosing to favour PAOs  As recommended by Serralta et al. (2004), VFAs could be added at the beginning of the anaerobic phase, so that these VFAs are only available to PAOs. These authors use an inflection point in the conductivity curve as the indicator of the end of denitrification process, which marks the end of the anoxic phase and the beginning of the anaerobic phase. The ORP “nitrate knee” is an alternative marker of this point. If both EC and ORP measurements were used to find this point, VFA could be added even if one of the signals were too weak to be registered by its respective algorithm. This redundancy could potentially make the EBPR process more resistant to failure.  72  3.5.3 Improved algorithms for maximum energy savings  Too much aeration is deleterious to the EBPR process (Serralta et al., 2004), and is an unnecessary expense. Closer EBPR process monitoring by EC should be able to reduce the length of the aeration phase, without compromising P removal. Sin et al., (2006) provide an example of model fitting to this end. Implementing real-time control with EC as a control parameter would also achieve this goal.  3.6 Improvements  Three main physical factors limit the application of EC control to the SBR system: interference, sensor fouling and sensor damage. EC (and other) measurements suffer from interference, particularly when contacting-type sensors are immersed in a liquid that acts as a ground. Proper galvanic isolation is required for reliable operation, but adds to setup complexity and cost. We were able to use a contacting-type sensor for the pilot study, but for an industrial application, a non-contacting toroidal-type sensor would be preferable in order to reduce interference, and reduce the maintenance associated with electrode coating, fouling, and physical damage prone to contacting sensors (Novak, 2003).  73  3.7 Process Modeling and Control  Two different but not altogether disparate approaches exist for the optimization of EBPR processes: that of improved process modeling, and that of real-time control. Kim et al. (2000) perform a comparison between the modelbased approach, and the pH/ORP profile control point approach for ammonia removal. A similar study for P removal would be helpful.  A challenge has been traditionally found in the coupling of process models and on-line control, due to the complexity of the models. The ASM1 model consists of 8 rate equations, 13 state variables, and 17 stoichiometric and kinetic parameters (Kim et al., 2000). The solution has traditionally been to simplify the ASM model, but with increased computational power the model may no longer have to be as simple in order to work in process control.  One issue that must be addressed is that there was no data for the EC profile of a failing EBPR cycle: this is a good candidate for a study on a full scale SBR. The development of an algorithm to detect poor P percent removal is complicated by the fact that the influent P:cation ratio is impossible to measure with sensors inexpensive enough to be used during treatment. If an on-line P and EC measurement system such as was attempted at the outset of this study were installed at a large plant, the collected data could be used to determine the feasibility of this sort of helpful algorithm.  74  This study provides an approach to both monitor wastewater treatment plants using EC in order to more thoroughly characterize the EBPR process, which would lead to better modeling, and also an avenue to improved process control using EC as a proxy measurement of EBPR process success. Further to an operator’s ability to use the model based approach to predict effluent quality, EC monitoring of an EBPR SBR provides a real-time measure of the effluent quality with respect to P removal.  4 Conclusion  This study confirmed that during the EBPR process, P concentration and EC might correlate, at the pilot scale, due to the presence and co-transport of cations. This confirmation shows that the findings of laboratory scale EBPR experiments can be replicated at a larger scale, and are not laboratory artifacts. These findings can be put to use in the development of EC-based process control algorithms for EBPR, most importantly to detect a massive failure of the EBPR process so that an automated SBR could release a dose of chemical precipitant to prevent the potential release of a P pulse into an aquatic (P-limited) environment.  75  References  Aguado, D., Montoya, T., Ferrer, J., and Seco, A. (2006). Relating ions concentration variations to conductivity variations in a sequencing batch reactor operated for enhanced biological phosphorus removal. Environmental Modelling and Software 21, 845-851. Akın, B. S., and Ugurlu, A. (2005). Monitoring and control of biological nutrient removal in a sequencing batch reactor. Process Biochemistry 40, 2873-2878. Arvin, E. (1985). Biological removal of phosphorus from wastewater. CRC Critical Reviews in Environmental Control 15, 25-64. Arvin, E. and Kristensen G.H. (1985). Exchange of organics, phosphate and cations between sludge and water in biological phosphorus and nitrogen removal processes. Water Science and Technology 17, 147-162. Auling, G., Pilz, F., Busse, H.J., Karrasch, S., Streichan, M., and Schon, G. (1991). Analysis of the polyphosphate-accumulating microflora in phosphorus-eliminating, anaerobic-aerobic activated sludge systems by using diaminopropane as a biomarker for rapid estimation of Acinetobacter spp. Applied and Environmental Microbiology 57, 3585-3592. Balmér, P. (2004). Phosphorus recovery--an overview of potentials and possibilities. Water Science and Technology 49, 185-190. Barat, R., Montoya, T., Seco, A., and Ferrer, J. (2005). The role of potassium, magnesium and calcium in the enhanced biological phosphorus removal treatment plants. Environmental Technology 26, 983-992. Barnard, J.L. (1974). Cut P and N without chemicals. Water and Wastes Engineering 11, 41-44. Barnard, J.L. (1975). Biological nutrient removal without the addition of chemicals. Water Research 9, 485-490. Barnard, J. (1984) Activated primary tanks for phosphate removal. Water SA 10, 121128. Beacham, A.M., Seviour, R.J., Lindrea, K.C. and Livingston, I. (1990). Genospecies diversity of Acinetobacter isolates obtained from a biological nutrient removal pilot plant of a modified UCT configuration. Water Research 24, 23–29.  76  Beacham, A.M., Seviour, R.J. and Lindrea, K.C. (1992). Polyphosphate accumulating abilities of Acinetobacter isolates from a biological nutrient removal pilot plant. Water Research 26, 121-122. Beer, M., Kong, Y.H. and Seviour, R.J. (2004). Are some putative glycogen accumulating organisms (GAO) in anaerobic: aerobic activated sludge systems members of the α-Proteobacteria? Microbiology 150, 2267-2275. Beer, M., H. M. Stratton, P. C. Griffiths and R. J. Seviour (2006). Which are the polyphosphate accumulating organisms in full-scale activated sludge enhanced biological phosphate removal systems in Australia? Journal of Applied Microbiology 100, 233-43. Blackall, L.L., Crocetti, G.R., Saunders, A.M. and Bond, P.L. (2002). A review and update of the microbiology of enhanced biological phosphorus removal in wastewater treatment plants. Antonie van Leeuwenhoek 81, 681-691. Bond, P.L., Erhart, R., Wagner, M., Keller, J. and Blackall, L.L. (1999). Identification of some of the major groups of bacteria in efficient and nonefficient biological phosphorus removal activated sludge systems. Applied and Environmental Microbiology 65, 4077-4084. Bond, P.L., Hugenholtz, P., Keller, J. and Blackall, L.L. (1995). Bacterial community structures of phosphate-removing and non-phosphate- removing activated sludges from sequencing batch reactors. Applied and Environmental Microbiology 61, 1910-1916. Bonting, C.F.C., Willemsen, B.M.F., Akkermans-van Vliet, W., Bouvet, P.J.M., Kortstee, G.J.J. and Zehnder, A.J.B. (1992). Additional characteristics of the polyphosphate-accumulating Acinetobacter strain 210A and its identification as Acinetobacter johnsonii. FEMS Microbiology Letters 102, 57-64. Carr, E., Ward, A., Gürtler, V. and Seviour, R.J. (2001). Pyrolysis mass spectrometry (PyMS) and 16S–23S rDNA spacer region fingerprinting suggests the presence of novel Acinetobacters in activated sludge. Systematic and Applied Microbiology 24, 430–442. Carr, E.L., Kämpfer, P., Patel, B.K.C., Gurtler, V. and Seviour, R.J. (2003). Seven novel species of Acinetobacter isolated from activated sludge. International Journal of Systematic and Evolutionary Microbiology 53, 953-963. Chang, C.H. and Hao, O.J. (1996). Sequencing batch reactor system for nutrient removal: ORP and pH profiles. Journal of Chemical Technology and Biotechnology 67, 27–38. Kishida, N., Kim, J.H., Chen, M., Sasaki, H. and Sudo, R. (2003). Effectiveness of oxidation-reduction potential and pH as monitoring and control parameters for nitrogen removal in swine wastewater treatment by sequencing batch reactors, J. Biosci. Bioeng. 96, 285–290.  77  Chua, A.S., Eales, K,. Mino, T., and Seviour, R. (2004). The large PAO cells in fullscale EBPR biomass samples are not yeast spores but possibly novel members of the beta-Proteobacteria. Water Science and Technology 50, 123-130. Cloete, T.E. and Steyn, P.L. (1987). A combined fluorescent antibody-membrane filter technique for enumerating Acinetobacter in activated sludge. In Biological Phosphate Removal from Wastewaters, pp. 335–338. Edited by R. Ramadori. Oxford: Pergamon Press. Cloete, T.E. and Steyn, P.L. (1988a) The role of Acinetobacter as a phosphorus removing agent in activated sludge. Water Research 22, 971-976. Cloete, T.E. and Steyn, P.L. (1988b). A combined membrane filter-immunofluorescent technique for the in situ identification and enumeration of Acinetobacter. Water Research 22, 961-969. Cohan, F. M. (2002). What are bacterial species? Annu Rev Microbiol 56, 457-487. Comeau, Y., Rabinowitz, B., Hall, K.J. and Oldham, K.W. (1987). Phosphate release and uptake in enhanced biological phosphorus removal from wastewater, Journal Water Pollution Control Federation 59, 707–715. Crocetti, G.R., Banfield, J.F., Keller, J., Bond, P.L. and Blackall, L.L. (2002). Glycogen-accumulating organisms in laboratory-scale and full-scale wastewater treatment processes. Microbiology 148, 3353-3364. Crocetti, G. R., Hugenholtz, P., Bond, P. L., Schuler, A., Keller, J., Jenkins, D. and Blackall, L. L. (2000). Identification of polyphosphate-accumulating organisms and design of 16S rRNA-directed probes for their detection and quantitation. Applied and Environmental Microbiology 66, 1175-1182. Dabert, P., Delgenès, J.P. and Godon, J.J. (2005). Monitoring the impact of bioaugmentation on the start up of biological phosphorus removal in a laboratory scale activated sludge ecosystem. Applied Microbiology and Biotechnology 66, 575-588. Dabert, P., Sialve, B., Delgenès, J.P., Moletta, R. and Godon, J.J. (2001). Characterisation of the microbial 16S rDNA diversity of an aerobic phosphorus-removal ecosystem and monitoring of its transition to nitrate respiration. Applied Microbiology and Biotechnology 55, 500-509. Deinema, M.H., Van Loosdrecht, M. and Scholten, A. (1985). Some physiological characteristics of Acinetobacter spp. accumulating large amounts of phosphate. Water Science and Technology 17, 119-125.  78  Duncan, A., Vasiliadis, G.E., Bayly, R.C., May, J.W. and Raper, W.G.C. (1988). Genospecies of Acinetobacter isolated from activated sludge showing enhanced removal of phosphate during pilot-scale treatment of sewage. Biotechnology Letters (Historical Archive) 10, 831-836. Fuhs, G.W. and Chen, M. (1975). Microbiological basis of phosphate removal in the activated sludge process for the treatment of wastewater. Microbial Ecology 2, 119–138. Gray, N. F. (2004). Biology of wastewater treatment, 2nd edn. London: Imperial College Press. Gray, N.D. and Head, I.M. (2001). Linking genetic identity and function in communities of uncultured bacteria. Environmental Microbiology 3, 481-492. Gray, N.D., Miskin, I.P., Kornilova, O., Curtis, T.P. and Head, I.M. (2002). Occurrence and activity of Archaea in aerated activated sludge wastewater treatment plants. Environmental Microbiology 4, 158-168. van Groenestijn, J.W., Bentvelsen, M.M., Deinema, M.H., and Zehnder, A.J. (1989). Polyphosphate-degrading enzymes in Acinetobacter spp. and activated sludge. Applied and Environmental Microbiology 55, 219-223. He, S., Gu, A. Z., and McMahon, K. D. (2006). Fine-scale differences between Accumulibacter-like bacteria in enhanced biological phosphorus removal activated sludge. Water Science and Technology 54, 111-7. He, S., Gall, D. L., and McMahon, K. D. (2007a). "Candidatus Accumulibacter" population structure in enhanced biological phosphorus removal sludges as revealed by polyphosphate kinase genes. Applied and Environmental Microbioogy 73, 5865-5874. He, S., Gu, A. Z., and McMahon, K. D. (2008). Progress toward understanding the distribution of Accumulibacter among full-scale enhanced biological phosphorus removal systems. Microbial Ecology 55, 229-236. Henze, M. (1997) Characterization of Wastewaters and Sludges In Henze, M., Harremoes, P., La Cour Jansen, J., and Arvin, E. (1997). (1997). Wastewater treatment : Biological and chemical processes, 2nd edn. Berlin ; New York: Springer. Hesselmann, R.P.X., Werlen, C., Hahn, D., Van der Meer, J.R. and Zehnder, A.J.B. (1999). Enrichment, phylogenetic analysis and detection of a bacterium that performs enhanced biological phosphate removal in activated sludge. Systematic and Applied Microbiology 22, 454-465. Hiraishi, A., Ueda, Y. and Ishihara, J. (1998). Quinone Profiling of Bacterial Communities in Natural and Synthetic Sewage Activated Sludge for Enhanced Phosphate Removal. Applied and Environmental Microbiology 64, 992-998.  79  Hiraishi, A., and Morishima, Y. (1990). Capacity for polyphosphate accumulation of predominant bacteria in activated sludge showing enhanced phosphate removal. Journal of. Fermentation and Bioengineering. 69, 368-371. Fan, Shuyu, (2006). Report for STAT 551, Statistical Research and Consulting Laboratory, Depatment of Statistics, University of British Columbia, Vancouver BC. Hu, Z.R., Wentzel, M.C. and Ekama, G.A. (2003). Modeling biological nutrient removal activated sludge systems--a review. Water Research 37, 3430-3444. Kämpfer, P., Erhart, R., Beimfohr, C., Bohringer, J., Wagner, M. and Amann, R. (1996). Characterization of bacterial communities from activated sludge: Culturedependent numerical identification versus in situ identification using group- and genusspecific rRNA-targeted oligonucleotide probes. Microbial Ecology 32, 101-121. Kawahar asaki M., Tanaka H., Kanagawa, T. and Nakamur a, K. (1999). In situ identification of polyphosphate-accumulating bacteria in activated sludge by dual staining with rRNA-targeted oligonucleotide probes and′,64 -diamidino-2-phenylindol (DAPI) at a polyphosphate-probing concentration, Water Research. 33, 257–265. Kawaharasaki, M., Manome, A., Kanagawa, T., and Nakamura, K. (2002). Flow cytometric sorting and RFLP analysis of phosphate accumulating bacteria in an enhanced biological phosphorus removal system. Water Science and Technology, 46, 139–144. Kerrn-Jesperson, J.P. and Henze, M. (1993). Biological phosphorus removal from wastewater by aerobic-anoxic sequencing batch reactors, Water Research 27, 617–624. Kim, H., McAvoy, T.J., Anderson, J.S. and Hao, O.J. (2000). Control of an alternating aerobic-anoxic activated sludge system -- Part 2: optimization using a linearized model, Control Engineering Practice 8, 279-289. Kim, J., Chen, M., Kishida, N., and Sudo, R. (2004). Integrated real-time control strategy for nitrogen removal in swine wastewater treatment using sequencing batch reactors. Water Research 38, 3340-3348. Kim, K.-S., Yoo, J .-S., Kim, S. Hee, J .L., Ahn, K.-H. and Kim I.S. (2007). Relationship between the electric conductivity and phosphorus concentration variations in an enhanced biological nutrient removal process. Water Science and Technology 55, 203-208. Kong, Y.H., Beer, M., Seviour, R.J., Lindrea, K.C. and Rees, G.A. (2002). Functional analysis of EBPR microbial communities in sequencing batch reactors with different phosphorus/carbon (P/C) ratios. Microbiology (UK) 148: 2299-2307. Kong, Y.H., Nielsen J.L and Nielsen, P.H. (2005). Identity and ecophysiology of  80  uncultured actinobacterial phosphate-accumulating organisms in full-scale enhanced biological phosphorus removal plants. Applied and Environmental Microbiology, 71: 4076-4085. Kishida N., Kim, J .H., Chen, M., Sasaki, H. and Sudo, R., (2003). Effectiveness of oxidation-reduction potential and pH as monitoring and control parameters for nitrogen removal in swine wastewater treatment by sequencing batch reactors. Journal of Bioscence and Bioengineering 96, pp. 285–290. Kortstee, G.J.J., Appeldoorn, K.J., Bonting, C.F.C., van Niel, E.W.J. and van Veen, H.J. (2000). Ecological aspects of biological phosphorus removal in activated sludge systems. Advances in Microbial Ecology 16, 169-200. Lee, N., Nielsen, P.H., Andreasen, K.H., Juretschko, S., Nielsen, J.L., Schleifer, K.H. and Wagner, M. (1999). Combination of Fluorescent In Situ Hybridization and Microautoradiography---a New Tool for Structure-Function Analyses in Microbial Ecology. Applied and Environmental Microbiology 65, 1289-1297. Lee, D. S., Jeon, C. O., and Park, J. M. (2001). Biological nitrogen removal with enhanced phosphate uptake in a sequencing batch reactor using single sludge system. Water Research 35, 3968-3976. Lee. N, Jansen, J.C., Aspegren, H., Henze, M., Nielsen, P.H., and Wagner, M. (2002). Population dynamics in wastewater treatment plants with enhanced biological phosphorus removal operated with and without nitrogen removal. Water Science and Technology 46:163-70. Lemos, P. C., Viana, C., Salgueiro, E. N., Ramos, A. M., Crespo, J.P. and Reiszcorr, S. G. (1998). Effect of carbon source on the formation of polyhydroxyalkanoates (PHA) by a phosphate-accumulating mixed culture. Enzyme and Microbial Technology 22, 662671. Levantesi, C., Serafim, L.S., Crocetti, G.R., Lemos, P.C., Rossetti, S., Blackall, L.L., Reis, M.A.M. and Tandoi, V. (2002). Analysis of the microbial community structure and function of a laboratory scale enhanced biological phosphorus removal reactor. Environmental Microbiology 4, 559-569. Levin G.V. and Shapiro J. (1965). Metabolic uptake of phosphorus by wastewater organisms. Journal of the Water Pollution Control Federation 37 800-821. Lin, C.-K., Katayama, Y., Hosomi, M., Murakami, A. and Okada, M. (2000). The relationship between isoprenoid quinone and phosphorus removal activity. Water Research 34, 3607-3613.  81  Liu, W.-T., Mino, T, Nakamura, K., and Matsuo, T. (1996). Glycogen accumulating population and its anaerobic substrate uptake abilities in anaerobic-aerobic activated sludge without biological phosphorus removal. Water Research 30, 75–87. Liu, W.T., Nakamura, K., Matsuo, T. and Mino, T. (1997). Internal energy-based competition between polyphosphate- and glycogen-accumulating bacteria in biological phosphorus removal reactors--Effect of P/C feeding ratio. Water Research 31, 14301438. Liu, W.-T., Linning, K.D., Nakamura, K., Mino, T., Matsuo, T. and Forney, L.J. (2000). Microbial community changes in biological phosphate-removal systems on altering sludge phosphorus content. Microbiology 146, 1099-1107. Liu, Y., Lin, Y., and Tay, J. (2005). The elemental compositions of P-accumulating microbial granules developed in sequencing batch reactors. Process Biochemistry 40, 3258-3262. Loy, A., Daims, H. and Wagner, M. (2002). Molecular techniques for determining community composition, p. 26-43. In G. Bitton (ed.), Encyclopedia of environmental microbiology. Wiley, New York, N.Y. Majone, M., Dirks, K. and Beun, J.J. (1999) Aerobic storage under dynamic conditions in activated sludge process. The state of the art. Water Research 39, 61-73. Marsili-Libelli, S. (2006). Control of SBR switching by fuzzy pattern recognition. Water Research 40, 1095-1107. Mar tin, H. G., Ivanova, N., Kunin, V. and other author s (2006). Metagenomic analysis of two enhanced biological phosphorus removal (EBPR) sludge communities. Nature Biotechnology 24, 1263-1269. Maur er , M. and Gujer , W. (1995). Monitoring of microbial phosphorus release in batch experiments using electric conductivity. Water Research 29, 2613–2617. Melasniemi, H. and Hernesmaa, A. (2000). Yeast spores seem to be involved in biological phosphate removal: a microscopic in situ case study. Microbiology 146, 701707. Menar A.B. and Jenkins D. (1970). Fate of phosphorus in waste treatment processes: Enhanced removal of phosphate by activated sludge, Environmental Science and Technology 4, 1115-1121. Mudaly, D.D., Atkinson, B.W., Bux, F. (2001). 16S rRNA in situ probing for the determination of the family level community structure implicated in enhanced biological nutrient removal. Water Science and Technology 43. 91-98.  82  Mulkerrins, D., Dobson, A.D.W. and Colleran, E. (2004). Parameters affecting biological phosphate removal from wastewaters. Environment International 30, 249-259. Muyzer, G. and Smalla, K. (1998). Application of denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie van Leeuwenhoek 73, 127-141. Novak, J. (2003). What is conductivity and how is it measured: a technical handbook for industry. Hach Company. NRC (1996). Use of reclaimed wastewater and sludge in food crop production. Committee on the use of treated municipal wastewater effluents and sludge in the production of crops for human consumption. Water Science and Technology Board. Commission on Geosciences, Environment and Resources. National Research Council. National Academy Press, Washington D.C. Oehmen, A., Teresa Vives, M., Lu, H., Yuan, Z. and Keller, J. (2005). The effect of pH on the competition between polyphosphate-accumulating organisms and glycogenaccumulating organisms. Water Research 15, 3727-3737. Oehmen, A., Lemos, P.C., Carvalho, G., Yuan, Z., Keller, J., Blackall, L.L. and Reis, M.A.M. (2007). Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Research 41, 2271-2300. Oldham, W.K. and Stevens, G.M. (1985). Operating experiences with the Kelowna Pollution Control Center. Proceedings of the Seminar on Biological Phosphorus Removal in Municipal Wastewater Treatment, Penticton, British Columbia. Onuki, M., Satoh, H. and Mino, T. (2002). Analysis of microbial community that performs enhanced biological phosphorus removal in activated sludge fed with acetate. Water Science and Technology 46, 145-153. Osborn, A.M., Moore, E.R.B. and Timmis, K.N. (2000). An evaluation of terminalrestriction fragment length polymorphism (T-RFLP) analysis for the study of microbial community structure and dynamics. Environmental Microbiology 2, 39-50. Per eir a, H., Lemos, P.C., Reis, M.A.M., Cr espo, J .P.S.G., Car r ondo, M.J .T. and Santos, H. (1996). Model for carbon metabolism in biological phosphorus removal processes based on in vivo 13C-NMR labelling experiments Water Research 30, 21282138. Pijuan, M., Baeza, J.A., Casas, C. and Lafuente, J. (2004). Response of an EBPR population developed in an SBR with propionate to different carbon sources. Water Science and Technology 50, 131-138.  83  Pijuan, M., Oehmen, A., Baeza, J A,, Casas, C. and Yuan, Z. (2008). Characterising the biochemical activity of full-scale enhanced biological phosphorus removal systems: a comparison with metabolic models. Biotechnology and Bioengineering 99, 170-179. Qur eshi, A., Lo, K.V., Mavinic, D.S., Liao, P.H., Koch, F. and Kelly, H. (2006). Dairy manure treatment, digestion and nutrient recovery as a phosphate fertilizer. Journal of Environmental Science and Health B 41, 1221-35. Rickar d, L.F. and McClintock, S.A. (1992). Potassium and magnesium requirements for enhanced biological phosphorus removal from wastewater. Water Science and Technology 26, 2203–2206. Röske, I., Schönbor n, C. and Bauer , H.-D., (1995). Influence of the addition of different metals to an activated sludge system on the enhanced biological phosphorus removal. Internationale Revue der gesamten Hydrobiologie und Hydrographie 80, 1-17. Rudolfs W. (1947). Phosphates in sewage treatment: I, Quantities of Phosphates. Sewage Works Journal 19, 43-47. Rybicki S.M. (1997). Advanced Watewater Treatment: Phosphorus removal from Wastewater-A literature review Joint Polish-Swedish Reports. Stockholm. retrieved 10 May, 2005 from www.lwr.kth.se/Forskningsprojekt/Polishproject/JPS3s121.pdf Sawyer C.N. (1944). Biological engineering in sewage treatment. Sewage Works Journal 16, 925-935. Schlegel, H. G., and Zaborosch, C. (1993). General microbiology, 7th edn. Cambridge, Eng.; New York, NY: Cambridge University Press. Schönborn, C., Bauer, H.D. and Roske, I. (2001). Stability of enhanced biological phosphorus removal and composition of polyphosphate granules. Water Research 35, 3190-3196. Schramm, A. and Amann, R. (1999). Nucleic-acid based techniques for analyzing the diversity, structure and dynamics of microbial communities in activated sludge. In: Environmental Processes 1 (Winter, J., Ed.), pp. 85–108. Wiley-VCH, Weinheim. Schwieger, F. and Tebbe, C.C. (1998). A New Approach To Utilize PCR-SingleStrand-Conformation Polymorphism for 16S rRNA Gene-Based Microbial Community Analysis. Applied and Environmental Microbiology 64, 4870-4876. Serafim, L. S., Lemos, P. C., and Reis, M. A. (2002). Effect of pH control on EBPR stability and efficiency. Water Science and Technology 46, 179-84.  84  Serralta, J., Borras, L., Blanco, C., Barat, R., and Seco, A. (2004). Monitoring pH and electric conductivity in an EBPR sequencing batch reactor. Water Science and Technology 50, 145-52. Seviour, R.J., Mino, T. and Onuki, M. (2003). The microbiology of biological phosphorus removal in activated sludge systems. FEMS Microbiology Reviews 27, 99127. Simpkins, M.J. (1979). The present status of research on phosphorus removal in South Africa, with particular reference to biological processes. Proceedings of Technology Transfer Seminar “Nutrient removal from municipal effluents”, Pretoria, South Africa. Sin, G., Govor eanu, R., Boon, N., Schelstr aete, G., and Vanr olleghem, P. A. (2006). Evaluation of the impacts of model-based operation of SBRs on activated sludge microbial community. Water Science and Technology 54, 157-166. Snaidr, J., Amann, R., Huber, I., Ludwig, W. and Schleifer, K.-H. (1997). Phylogenetic analysis and in situ identification of bacteria in activated sludge. Applied and Environmental Microbiology 63, 2884–2896. Spagni, A., Buday, J., Ratini, P., and Bortone, G. (2001). Experimental considerations on monitoring ORP, pH, conductivity and dissolved oxygen in nitrogen and phosphorus biological removal processes. Water Science and Technology 43, 197-204. Standar d Methods for the Examination of Water and WasteWater . (1998). Ed.: L.S. Clesceri, A.E. Greenberg, and A.D. Eaton. American Public Health Association, the American Water Works Association and the Water Environment Federation. 20th edition. Washington, DC. Stephens, H.L. and Stensel, H.D. (1998). Effect of operating conditions on biological phosphorus removal. Water Environment Research. 70, 360-369. Streichan, M., Golecki, J.R. and Schön, G. (1990). Polyphosphate-accumulating bacteria from sewage plants with different processes for biological phosphorus removal.FEMS Microbiology Ecology 73, 113-124. Sudiana, I.M., Mino, T., Satoh, H. and Matsuo, T. (1998). Morphology, In-Situ characterization with rRNA targetted probes and respiratory quinone profiles of enhanced biological phosphorus removal sludge. Water Science and Technology 38, 69–76. Tandoi, V., Majone, M., May, J. and Ramadori, R. (1998). The behaviour of polyphosphate accumulating Acinetobacter isolates in an anaerobic-aerobic chemostat. Water Research 32, 2903-2912. Tchobanoglous, G., Burton, F.L. and Stensel, H.D. (2003). Wastewater Engineering: Treatment and Reuse, Metcalf and Eddy, Inc. 4th Edition, McGraw Hill, Toronto.  85  Theron J, and Cloete T.E. (2000). Molecular techniques for determining microbial diversity and community structure in natural environments. Critical Reviews in Microbiology 26, 37-57. van Veen, H.W., Abee, T., Kortstee, G.J., Konings, W.N. and Zehnder, A.J. (1993). Characterization of two phosphate transport systems in Acinetobacter johnsonii 210A. Journal of Bacteriology 175, 200-206. Wagner, M., Erhart, R., Manz, W., Amann, R., Lemmer, H., Wedi, D. and Schleifer, K.-H. (1994). Development of an rRNA-targeted oligonucleotide probe specific for the genus Acinetobacter and its application for in situ monitoring in activated sludge. Applied and Environmental Microbiology 60, 792–800. Wagner, M., Loy, A., Nogueira, R., Purkhold, U., Lee, N. and Daims, H. (2002). Microbial community composition and function in wastewater treatment plants. Antonie Van Leeuwenhoek 81, 665-680. Wentzel, M.C., Ekama, G.A., R.E. Lowenthal, P.L. Dold, and Marais G.v.R.. (1989). Enhanced polyphosphate organism cultures in activated sludge systems, part II, experimental behavior. Water Science 15, 71-88. Wentzel, M.C., Lötter, L.H., Ekama, G.A., Loewenthal, R.E. and Marais, G.v.R. (1990). Evaluation of biochemical models for biological excess phosphorus removal. Water Science and Technology 23, 567-576. Wiechers H.N.S. and van Vuuren L.R.J (1979). A review of applied technology for nutrient removal from municipal effluents. Proceedings of Technology Transfer Seminar ”Nutrient removal from municipal effluents”, Pretoria, South Africa. Winder, C.L., Carr, E., Goodacre, R. and Seviour, R. (2004). The rapid identification of Acinetobacter species using Fourier transform infrared spectroscopy. Journal of Applied Microbiology 96, 328-339. von Wintzingerode, F., Gobel, U.B. and Stackebrandt, E. (1997). Determination of microbial diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbiology Reviews 21, 213-229. Yuan, Z., and Blackall, L. L. (2002). Sludge population optimisation: A new dimension for the control of biological wastewater treatment systems. Water Research 36, 482-490. Zhang, T., Liu, Y., Fang, H.H.P. (2005). Effect of pH change on the performance and microbial community of enhanced biological phosphate removal process. Biotechnology and Bioengineering 92, 173-82.  86  Zhou, Y., Pijuan, M., Zeng, R. J., Lu, H. and Yuan, Z. (2008). Could polyphosphateaccumulating organisms (PAOs) be glycogen accumulating organisms (GAOs)? Water Research 42, 2361-2368. Zilles, J.L., Peccia, J., Kim, M.-W., Hung, C.-H. and Noguera, D.R. (2002). Involvement of Rhodocyclus-Related Organisms in Phosphorus Removal in Full-Scale Wastewater Treatment Plants. Applied and Environmental Microbiology 68, 2763-2769.  87  Appendix A: Solutions Phenylmercuric acetate preservative (O-PO4/NOx preservative) 0.1g phenylmercuric acetate 20 mL acetone 80 mL dH2O phosphoric acid (VFA preservative) 2% H3PO4 sulfuric acid (NH3 preservative) 5% H2SO4, 1N H2SO4 nitric acid (cation preservative) conc HNO3 Lanthanum stock (5ml added to a 50 ml sample) 15.16 g La2(NO3)3 1.9058 g CsNO3 to 130mL dH2O Appendix B: Schedule Table A.1 Track study schedule and availability of PLC data Track Study 1 2 3 4 5 6 7 8 9 10  Date 07/06/06 15:07 – 17:25 14/06/06 14:32 – 17:29 23/06/06 14:29- 17:24 25/07/06 14:57- 17:42 18/08/06 17:20 – 19:33 23/08/06 17:15 – 19:40 28/08/06 10:27 – 12:03 31/08/06 15:21 – 16:39 07/09/06 16:58 – 19:34 15/9/06 14:30 – 16:20  Comments EC sensor not calibrated EC sensor not calibrated PLC not set up to record EC PLC data recorded Used tank B PLC data recorded PLC data overwritten PLC data overwritten PLC data overwritten PLC data overwritten  88  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0067103/manifest

Comment

Related Items