UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigation of the methanol biological phosphorus removal phenomenon Fugère, Raphaël 2007

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

Item Metadata

Download

Media
831-ubc_2007-0404.pdf [ 15.47MB ]
Metadata
JSON: 831-1.0100876.json
JSON-LD: 831-1.0100876-ld.json
RDF/XML (Pretty): 831-1.0100876-rdf.xml
RDF/JSON: 831-1.0100876-rdf.json
Turtle: 831-1.0100876-turtle.txt
N-Triples: 831-1.0100876-rdf-ntriples.txt
Original Record: 831-1.0100876-source.json
Full Text
831-1.0100876-fulltext.txt
Citation
831-1.0100876.ris

Full Text

INVESTIGATION OF THE METHANOL BIOLOGICAL PHOSPHORUS REMOVAL PHENOMENON by RAPHAEL FUGERE B.Sc. Universite du Quebec a Montreal, 1999 M.ASc. Universite de Montreal, 2002 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA August 2007 © Raphael Fugere, 2007 ABSTRACT The removal of nutrients from a wastewater stream is a well-known process. The typical nutrients in wastewaters are carbon, nitrogen and phosphorus. Since phosphorus is usually the limiting agent for growth, it usually exhibits the most stringent discharge guidelines. In order to meet these discharge guidelines, municipalities and the industry have installed sophisticated treatment processes. Two main types of nutrient removal processes can be used; chemical nutrient removal and biological nutrient removal, with the latter gaining more popularity. While carrying out denitrification in anoxic tanks with the addition of methanol, some researchers have reported "additional" phosphorus removal. This study was aimed at replicating the operating conditions that might yield this "additional phosphorus removal", performing a mass balance and investigating thoroughly the possible physical or chemical mechanisms responsible. Using a laboratory-scale bioreactor fed with synthetic sewage in which the only exogenous source of carbon was methanol, enhanced phosphorus removal was successfully demonstrated. Significant phosphorus removal was observed with this methanol addition, even without any addition of VFAs, often considered to be essential to trigger the phosphorus uptake and release phenomenon. Low oxic conditions seemed to favor good phosphorus removal in these experiments. It was also found that pH and alkalinity had very little impact on the phosphorus removal phenomenon. However, the addition of ethanol to a biomass, already acclimated to methanol, seemed to boost, significantly, the phosphorus release mechanisms, while having little effect on the phosphorus uptake phenomenon. The replacement of methanol by acetate on .a carbon-based stoichiometric ratio was unsuccessful. While the phosphorus release seemed enhanced, acetate totally destabilized the biomass and resulted in a "reactor crash". A possible metabolic pathway was identified, by which the methanol could be converted to acetate and PHB, thus explaining this methanol-induced biological phosphorus removal phenomenon. However, additional research will have to be undertaken to confirm this metabolic pathway, as well as identifying the various organisms involved in the enhanced phosphorus removal process. If the methanol biological phosphorus removal process could be controlled operationally, it could provide a relatively cheap and efficient way of removing both nitrogen and phosphorus within the same biotreatment sequence. o ii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS Hi LIST OF TABLES v LIST OF FIGURES vi ACKNOWLEDGEMENTS viii C H A P T E R I - I N T R O D U C T I O N 1 1.1 OBJECTIVES 1 1.2 RESEARCH HYPOTHESIS 1 1.3 THESIS STRUCTURE 2 C H A P T E R II - L I T E R A T U R E R E V I E W 3 2.1 PROBLEMS WITH PHOSPHORUS 3 2.2 TREATMENT PROCESSES ; 5 2.3 CHEMICAL PHOSPHORUS REMOVAL 5 2.3.1 Natural Chemical Phosphorus Removal 5 2.3.2 Man-made Chemical Phosphorus Removal 5 2.4 BIOLOGICAL PHOSPHORUS REMOVAL 6 2.4.1 Biological Phosphorus Removal - Assimilation 6 2.4.2 Enhanced Biological Phosphorus Removal 6 2.5 MECHANISMS OF PHOSPHORUS REMOVAL 8 2.6 METHANOL USED IN EBPR 12 2.6.1 Methanol as a Denitrification Substrate 12 2.6.2 Methanol-Consuming Organisms (Methylotrophs) 12 2.6.3 Stoichiometry of EBPR 17 2.6.4 Previous Studies on Biological Phosphorus Removal with Methanol 17 C H A P T E R III - M A T E R I A L S A N D M E T H O D S 19 3.1 SEQUENCING BATCH REACTOR (SBR) 19 3.1.1 Reactor Setup/Operation 19 3.1.2 Synthetic Sewage Composition 21 3.1.3 Secondary Feed Composition 22 3.2 BATCH TESTS 22 3.3 ANALYTICAL METHODS 23 3.3.1 General Analytical Methods 23 3.3.2 Chemical Analyses by Gas Chromatography 25 3.3.3 Polyphosphates - Microscopic Examination 26 3.3.4 PHB -Microscopic Examinations 27 3.3.5 Glycogen Method 27 3.3.6 Formaldehyde Method 28 3.4 SAMPLING QA/QC 28 C H A P T E R I V - R E S U L T S A N D D I S C U S S I O N 29 4.1 BACKGROUND INFORMATION .29 iii 4.2 FIRST PHASE 29 4.2.1 Initializing Conditions for Nitrification 29 4.3 EXCESS AERATION EXPERIMENTS - GETTING NITRIFICATION TO WORK 33 4.4 SETTING UP ANAEROBIC CONDITIONS - GETTING DENITRIFICATION TO WORK 38 4.5 EXPERIMENTS WITH LOW DISSOLVED OXYGEN LEVELS 42 4.6 TOTAL PHOSPHORUS/TOTAL NITROGEN BALANCE 46 4.7 BATCH TESTS 47 4.7.1 Batch Test #1 47 4.7.2 Batch Test #2 49 4.8 STAINING AND MICROSCOPIC EXAMINATIONS : 51 4.8.1 Staining of Methanol Bio-P Sludge 51 4.8.2 Staining of UBC Pilot Plant's Sludge 55 4.9 INVESTIGATION OF THE EFFECT OF PH ON BIO-P 56 4.10 EFFECT OF NITRIFICATION AND DENITRIFICATION ON P REMOVAL 62 4.11 ETHANOL AS A REPLACEMENT SUBSTRATE FOR METHANOL 66 4.12 ACETATE AS A REPLACEMENT SUBSTRATE FOR METHANOL 70 4.13 UNIFORMITY OF THE BIOMASS OF THE TWO REACTORS: 75 4.14 CHARACTERIZATION OF ORGANISMS PRESENT IN THE BIO-REACTORS 75 4.15 MOLECULAR MECHANISMS OF METHANOL CONVERSION 77 C H A P T E R V - S U M M A R Y A N D C O N C L U S I O N S 81 5.1 FOR FUTURE INVESTIGATION 83 R E F E R E N C E S 84 iv LIST OF TABLES Table 1: Macro and micronutrient composition of a typical 60 L synthetic sewage batch 22 Table 2 : Analytical Methods 24 Table 3: TKN and TP concentration in main reactor biomass; samples 1-3 were unfiltered from the anoxic phase, 4-6 were filtrate from the anaerobic phase, while 7-9 were unfiltered at the end of the aerobic phase 46 v LIST OF FIGURES Figure 1: The phosphorus cycle 3 Figure 2: Relationship between the substrate metabolism and the phosphorus uptake in anaerobic and aerobic zones of a biological phosphorus removal reactor 7 Figure 3: Nitrogen, phosphorus and carbon profiles measured in a typical cycle of a SBR 10 Figure 4: Pathways of the metabolism of single-carbon compounds by methanotrophic bacteria 14 Figure 5: Genotypes associated with methylotrophs 15 Figure 6: Metabolic degradation of methanol 16 Figure 7: Flow scheme for a continuous bench scale pilot plant feeding with methanol and acetate 18 Figure 8: Schematic of the sequencing batch reactor used for most of the experiments in this study 20 Figure 9: Photo of the two empty SBRs and surrounding setup 21 Figure 10: Batch test apparatus 23 Figure 11: Typical pH and DO profiles over a 5 h period for the SBR reactor 31 Figure 12: Typical ORP profile over a 5 h period for the SBR reactor . 31 Figure 13: Nitrate, nitrite, orthophosphate and ammonia profiles over one cycle of the biological reactor 32 Figure 14: pH, DO and ORP profiles over one cycle of the methanol-fed anoxic-aerobic SBR. 36 Figure 15: NO x, orthophosphates, ammonia and COD concentration profiles over one cycle of the methanol-fed anoxic-aerobic SBR 37 Figure 16: pH, DO and ORP profiles over one cycle of the methanol-fed anoxic-aerobic SBR. 40 Figure 17: NO x, orthophosphates, ammonia, COD and TOC concentration profiles over one cycle of the methanol-fed anoxic-aerobic SBR 41 Figure 18: pH, DO and ORP profiles over one cycle of the methanol-fed anoxic-aerobic SBR. 44 Figure 19: NO x, orthophosphates, ammonia concentration profiles as well as the methanol addition profile over one cycle of the methanol-fed anoxic-aerobic SBR 45 Figure 20: NO x concentrations vs time for a batch test examining the impact of yeast extract addition 48 Figure 21: Orthophosphate concentrations vs time for a batch test examining the impact of yeast extract addition 48 Figure 22: NO x concentration vs time for different yeast extract concentrations in a batch test. 49 Figure 23: Orthophosphate concentration vs time for different yeast extract concentrations for a batch test 50 Figure 24: Polyphosphate stain of a sludge sample prior to a prolonged anaerobic cycle 51 Figure 25: Polyphosphate stain of a sludge sample after a prolonged anaerobic cycle 52 Figure 26: Polyphosphate stain of a sludge sample after a prolonged anaerobic cycle 52 Figure 27: Polyphosphate stain of a sludge sample after a prolonged anaerobic cycle (lOOOx magnification) 53 Figure 28: Micrometer viewed at a lOOOx magnification 54 Figure 29: Polyphosphate stain of UBC Pilot Sewage Treatment Plant sludge under a 1000 x microscopic magnification 55 Figure 30: pH, DO and ORP profiles over tree cycles of the methanol-fed anoxic-aerobic SBR. : ...: 59 vi Figure 31: N0 X , orthophosphates, ammonia, VFA and methanol concentration profiles as well as the cumulative methanol addition profile, formaldehyde concentration over three cycles of the methanol-fed anoxic-aerobic SBR 60 Figure 32: Formaldehyde, o-P04 and glycogen profiles over three cycles of the methanol-fed anoxic-aerobic SBR 61 Figure 33: pH, DO and ORP profiles over three cycles of the methanol-fed anoxic-aerobic SBR. 64 Figure 34: Formaldehyde, methanol, ammonia, nitrate and 0-P04 profiles over three cycles of the methanol-fed anoxic-aerobic SBR 65 Figure 35: pH, DO and ORP profiles over three cycles of an ethanol-fed anoxic-aerobic SBR.. 68 Figure 36: Formaldehyde, 0-P04 profiles over three cycles of the ethanol-fed anoxic-aerobic SBR 69 Figure 37: The two laboratory-scale sequential batch reactors that were used in this research project 72 Figure 38: pH, DO and ORP profiles over thee cycles of the acetate-fed anoxic-aerobic SBR. . 73 Figure 39: Acetate, NOx, o-P04 and methanol profiles over three cycles of the acetate-fed anoxic-aerobic SBR 74 Figure 40: Various organisms that were observed under optical microscopy throughout this study. All images taken at a magnification of lOOOx 75 Figure 41: Biochemical pathways : 1- Phosphotransferase system, 2- Pyruvate kinase, 3-Pyruvate dehydrogenase, 4- phosphoenol pyruvate carboxylase, 5- pyruvate oxidase, 6-Lactate dehydrogenase, 7- Acetyl CoA synthase, 8- Phosphotransacetylase, 9- acetate kinase, 10- isocitrate lyase, 11- Malate synthase 78 Figure 42: Metabolic pathway between methanol, formaldehyde, acetate and PHB 79 vii ACKNOWLEDGEMENTS This research project would never have taken place if it wasn't for the support I received from a number of people. I would like to express my sincere appreciation to everyone who helped me out through this academic journey; Donald S. Mavinic for giving me this interesting, yet challenging, research opportunity, William Ramey for his wealth of knowledge and his advice, Paula Parkinson for her analytical, experimental and personal support, Susan Harper for setting up an excellent working environment, Patricia Keen for her conceptual, analytical and personal support, Nandini Sabrina for her help in the laboratory and for just being curious, Dean Shiskowski for his insights into bacterial metabolism, Robert Simm, who helped shape my academic life and career, Frederic Koch for his endless enthusiasm for research projects and scientific advice, Jim Atwater for guiding me through the maze of UBC and APEGBC administration and to all of the silent, too often unrecognized contributors. NSERC funding is also gratefully acknowledged. viii CHAPTER I - Introduction 1.1 Objectives The purpose of this study is to investigate unusual phosphorus uptake by an intermittently-aerated, activated sludge process, fed with methanol. This phenomenon has been observed by Zhao (1997), and later by wastewater treatment plant operators in Kent, British Columbia. The observed phenomenon appeared to be luxury uptake of phosphorus in a sequencing batch reactor (SBR), operated under simultaneous nitrification and denitrification (SND) conditions. The established literature attributes phosphorus removed in the activated sludge process to the enhanced biological phosphorus removal (EBPR) if the phosphorus content of the waste activated sludge is increased beyond the typical 1.5 to 2.5 %. EBPR is defined as a process that will increase phosphorus uptake beyond what is needed for microbial growth. This process typically requires specific conditions and reaction zones: anaerobic zone, anoxic zone and aerobic zone. The anaerobic zone is a zone without aeration, in which the bacteria will hydrolyze stored polyphosphates as an energy source for uptake of volatile fatty acids (VFA). Those VFAs are energy-rich source and will be processed and stored in the form of poly-B-hydroxybutyrate (PHB) or polyhydroxyvalerate (PHV). The anoxic zone is typically used for denitrification and phosphorus uptake (Zhao, 1997). The phosphorus uptake in this anoxic zone will be proportional to the amount of external substrate present. In the aerobic zone, the phosphorus-accumulating organisms (PAO) are believed to oxidize their stored carbon reserves, take up phosphorus and store it in the form of volutin (polyphosphate granules) (Zhao, 1997; Tchobanoglous et ai, 2003). 1.2 Research Hypothesis The principle hypothesis of this project was that the phenomenon that had been observed sporadically in denitrifying bioreactors fed with methanol is a biological process. Under low dissolved oxygen concentrations, some organisms would use methanol as a sole carbon source to achieve biological phosphorus removal. 1 It was also hypothesized that the phenomenon could be replicated using a laboratory-scale, bioreactor, fed with synthetic sewage and methanol as the sole substrate. 1.3 Thesis Structure Chapter one will present the context as well as research objectives and hypothesis. The second chapter will summarize the literature on traditional biological phosphorus removal processes, as well as literature related to methanol biological phosphorus removal. The third chapter will describe the analytical and sampling methods as well as the equipment used in the laboratory experiments. Chapter 4 will summarize and analyze the results that were obtained in the experiments. Finally, Chapter 5 will synthesize the various points, review the original objectives and propose improvements to the system, in addition to presenting avenues for future research. 2 CHAPTER II - Literature Review 2.1 Problems With Phosphorus Phosphorus, the element with the atomic number 15, is the 11th most abundant element on the planet and composes approximately 0.1 % of the earth's crust (Morton et al, 2005). With time, the phosphorus is eroded from the crust, becomes soluble and thus available for biological organisms, but the turnover is very slow - more than a billion years (Morton et al, 2005). A schematic of the phosphorus cycle is presented in Figure 1. Figure 1: The phosphorus cycle (Khan et al., 2005) Since this turnover rate is very slow, phosphorus is considered to be a limiting agent, meaning that the concentration of phosphorus in water will often dictate the occurrence of algae growth in a water body - especially in freshwater environments. When the limit of phosphorus availability is overcome, for example by accidental spill or human pollution, it is known to cause excessive plant growth - or eutrophication (Morton et al, 2005, Khan et al, 2005). This eutrophication phenomenon can be described as "...the promotion of the growth of plants, animals, and microorganisms in lakes and rivers, has been a very slow, natural process. If this is allowed to occur uninterrupted, it results in an excessive deficiency of oxygen in the water. Thus organisms that thrive under anaerobic conditions are favored more and more at the expense of aerobic organisms" (Khan et al., 2005). In surface waters, is has been reported that phosphorus concentrations as low as 0.05 mg/L could cause eutrophication (Khan et ah, 2005). Eutrophication often causes shifts in occurrence, distribution and diversity of the biotic community (Khan et al, 2005, Bennion, 1996). Phosphorus is the corner stone of all cell life; unicellular and multicellular alike. In fact, phosphorus is essential to microorganisms for energy transfer and for the construction of such cell components as phospholipids adenosine triphosphate, adenosine diphosphate, nicotiamide adenosine dinuclotide (NAD), flavin adenine dinucleotide (FAD), as well as nucleic acids such as DNA and RNA (Khan et al., 2005, USEPA, 1976). Phosphorus accounts for approximately 10-12 % of RNA or DNA mass (USEPA, 1976). Phosphorus possesses a unique characteristic compared to the other major nutrients (carbon, nitrogen, oxygen); it is often assumed to lack a gaseous phase. However, recent research by Glindemann et al. (1995, 1996a, 1996b) found that volatile phosphine gas (PH3) could be detected in the earth's atmosphere at minute levels. For decades, the phosphorus has been assumed to be non-volatile, and all atmospheric transport of phosphorus was assumed to be via phosphate dust (Morton, 2005). The properties of phosphorus have an impact on phosphate fluxes. Morse et al. (1993) reported that in Europe, phosphorus inputs into watersheds were due to the following: 10 % from background sources, 17 % from fertilizers, 7 % from the industry, 32 % from livestock, 11 % from detergents and 23 % from human waste origins. Humans are a major source of phosphorus in surface waters in the form of sewage and agricultural runoff. In order to mitigate the negative impacts of the practices, most industrialized countries have installed waste treatment plants, to remove nutrients such as carbon, nitrogen and phosphorus from the wastewater. Several nutrient removal technologies - biological, physical and chemical - are available. In this study, we will focus only on biological phosphorus removal processes. 4 2.2 Treatment Processes Chemical phosphorus removal typically relies on the precipitation of metallic salts with phosphorus, leading to the formation of insoluble compounds that can subsequently be removed by settling. Chemical-Physical processes can remove the soluble fraction of phosphates using crystallization (forming metal phosphates), ion exchange, adsorption and reverse osmosis. In order to tackle the particulate fraction, coagulation-flocculation, with a filtration process, can also be used (Morse et al., 1998). Biological phosphorus removal relies on the fact that phosphorus can be removed by incorporation of phosphate into the organic matter of cells that can subsequently be removed from wastewater by settling. Enhanced biological phosphorus removal is a particular form of biological phosphorus removal since it is based on the fact that, with the help of an aerobic/anaerobic cycle, micro-organisms will accumulate phosphorus in the form of intracellular polyphosphates that will be disposed of in the sludge wasted from the process. 2.3 Chemical Phosphorus Removal 2.3.1 Natural Chemical Phosphorus Removal Phosphorus removal can also be the result of natural occurring chemical precipitation. In fact, if ions of Al , Ca, Fe, Mg and Zn are present in significant concentrations, up to 6 mg P/L can be removed, assuming a typical wastewater composition (Arvin et ah, 1983). This suggests that chemical precipitation can play a significant role in phosphorus removal. Some phosphorus precipitates take complex forms such as tricalciumphosphate and hydroxyapatite (Arvin et al., 1983). 2.3.2 Man-made Chemical Phosphorus Removal Chemical phosphorus removal is typically achieved by adding metal salts such as iron, aluminum or lime to wastewater. The salts used are alum (Al2(S04)3*nH20), sodium aluminate, ferric chloride (FeCls), ferrous chloride (FeC^), ferrous sulphate (FeS04) (Tchobanoglous and Burton, 1991; USEPA, 1976). The salts react with phosphates to form iron or aluminum precipitates. The compounds are typically injected upstream of either the primary or the 5 secondary clarifier (USEPA, 1976). In a system exhibiting chemical precipitation, total phosphorus removal can reach 80-95%. In order to achieve a high level of phosphorus removal, the metal ions must be present in excess of the stoichoiometric ratio. 2.4 Biological Phosphorus Removal Phosphorus requires conversion to a solid form to be removed from wastewater. Soluble phosphorus is typically found in the form of orthophosphates, which have to be converted to organic phosphorus, polyphosphates or other compounds in a wastewater treatment plant. This section will explore phosphorus removal techniques with an emphasis on biological phosphorus removal mechanisms. 2.4.1 Biological Phosphorus Removal - Assimilation The primary treatment of wastewater removes approximately 5-10 % of the phosphorus -mostly the phosphorus associated with a solid matrix while conventional secondary treatment removes approximately 10-25 % of the phosphorus (Tchobanoglous et al., 1991). It was found that normal biological phosphorus assimilation was significant; Bitton et al. (1994) reported that the cell requirement of normal bacteria was 1-3 % of a cell's dry weight. This finding is very close to the value reported of 2.5 % for the average phosphorus content of bacterial cells by Tchobanoglous et al. (1991), the range between 1.5 to 3 % reported by Tetreault et al. (1986) and the range between 1.5 and 2 % reported by the U.S. EPA, (1976). 2.4.2 Enhanced Biological Phosphorus Removal Enhanced biological phosphorus removal relies on the fact that, when the mixed liquor is exposed to an anaerobic/aerobic sequence, it favors microorganisms that accumulate higher levels of intracellular phosphorus than other organisms. A schematic of the process is presented in Figure 2. 6 Anaerobic Aerofcic Substrate Bacteria ProAictUm PAOs Figure 2: Relationship between the substrate metabolism and the phosphorus uptake in anaerobic and aerobic zones of a biological phosphorus removal reactor (Adapted from Grady et al, 1999) Greenburg et al. (1955) was credited with the first results demonstrating that activated sludge could take up phosphorus at higher levels than growth requirements. In 1965, an enhanced biological phosphorus removal was reported in a District of Columbia activated sludge plant. Phosphorus removal exceeding 80 % was observed when the sludge was vigorously aerated. The high phosphorus removal phenomenon was termed "luxury uptake" (U.S. EPA, 1976) The scientists observed the formation of volutin granules inside the microorganisms. These granules were identified as containing polyphosphates (USEPA, 1976). Shapiro et al. (1967) later observed high phosphorus uptake at the Baltimore sewage treatment plant and noted that some phosphorus was being released under conditions of low dissolved oxygen (DO). High levels of phosphorus removal (85-95 %) were observed throughout the United States and phosphorus content of the waste sludge up to 2.7-3.0 % on a dry weight basis was reported. All these plants were under the plug-flow configuration and diffused aeration (USEPA, 1976). The operators of these plants realized that the plants needed 1) a DO concentration of 2.0 mg/L or higher from the middle to the end of the aeration basins, 2) a way to prevent the recycling of phosphorus back to the activated sludge basin via sludge recycle streams, 3) to maintain aerobic conditions in the secondary clarifier, to prevent the release of phosphorus in the effluent (USEPA, 1976). 7 Several variations on the biological phosphorus removal process were developed, such as: Phostrip process - Modified Bardenpho process A/O process UCT process Sequencing Batch Reactor process The Phostrip process incorporates an anaerobic zone in the sludge recycle system. It takes a sidestream from the return activated sludge and subjects it to anaerobic conditions in a tank, before returning the sludge to an aeration basin. The released phosphorus from the anaerobic tank is then precipitated with lime (U.S. EPA, 1976) The modified Bardenpho process and the UCT process are both designed to remove phosphorus, as well as nitrogen. The A/O process achieves biological phosphorus removal by modifying an existing plug-flow activated sludge reactor. The modifications consist of shutting down the aerators at the head of the reactor and by boosting the aeration at the tail of the reactor, thus subjecting the wastewater to anoxic and oxic conditions (U.S. EPA, 1976). The SBR system has the main advantage of requiring only one tank for the oxic and anoxic phases, thus limiting the size of the installations. 2.5 Mechanisms of Phosphorus Removal The mechanisms of enhanced uptake of phosphorus by micro-organisms have been the subject of numerous studies. It was demonstrated that, as a general rule for phosphorus removal to occur, the biomass first needed to be subjected to an anaerobic phase, i.e. an environment devoid of oxygen or nitrate. Following the anaerobic phase, the biomass needed to be subjected to an environment where an electron acceptor is present, such as nitrate (anoxic phase) or oxygen (aerobic phase). A schematic of the process is presented in Figure 3. As it can be observed from Figure 3, a quick release of phosphorus occurs in the anaerobic phase, followed by a slow uptake in the aerobic phase. The profile of poly-(3-bydroxybutyrate (PHB) concentrations is very similar to the profile of orthophosphates and seems to be inversely proportional to the concentration of acetate. 8 In fact, when sewage enters the anaerobic phase, bacteria called polyphosphate accumulating organisms (PAO) assimilate the available carbon sources (such as acetate) and store it as polyhydroxyalkanoates (PHA). PHAs typically take two forms: poly-p-bydroxybutyrate and poly-(3-bydroxyvalerate (PHV). The energy required to synthesize and store these PHAs is extracted from glycogen, as well as an energy-storage compound; polyphosphates (Poly-P). The Poly-P is broken down into smaller units called ortho-phosphates, conserving energy in the process. Ortho-phosphate molecules are soluble, so their release will increase the phosphate concentration in the anaerobic phase. Since the polyphosphates are typically stored in the form of volutin granules, the ortho-phosphate release is accompanied by a release of magnesium, potassium and calcium cations (Tchobanoglous et al., 2003). In typical activated-sludge heterotrophic bacteria, the phosphorus composition is 1.5 to 2.0 percent of the cell dry weight. However, several bacteria have been reported to store phosphorus in their cells in the form of polyphosphates, resulting in a cell content as high as 20 to 30 percent by dry weight (Tchobanoglous et al., 2003). The phosphorus plays a vital role in cellular energy transfer mechanisms via adenosine triphosphate and polyphosphates. As energy is stored, adenosine diphosphate (ADP) will be converted to adenosine triphosphate (ATP) with 7.4 kcal/mole of energy captured by the phosphate bond. When the cell requires energy, it converts ATP to ADP, releasing phosphorus (Tchobanoglous et al., 2003). 9 Anaerobic Aerobic 0 2 3 Time (h* Figure 3: Nitrogen, phosphorus and carbon profiles measured in a typical cycle of a SBR. Source: (Zeng et al., 2003) It was assumed for a long time that the anaerobic phase provided an exclusive environment for the PAOs, allowing them to store enough carbon for themselves without threat of competition from other organisms. It was demonstrated that another type of organism called Glycogen-accumulating organisms (GAOs) were able to store carbon and could successfully compete with PAOs (Tchobanoglous et al., 2003). The anaerobic phase needs to be followed by a nitrate or oxygen-rich phase, causing the consumption of stored PHB, providing energy for growth and reproduction. During this phase of "feast", glycogen supplies are replenished and orthophosphate is accumulated to reform polyphosphates (Tchobanoglous et al, 2003). This results in a decrease in the concentration of phosphorus in the anoxic/aerobic phase. Moreover, since the polyphosphate accumulating 10 organisms can store a large fraction of their dry weight in polyphosphates, wasting of sludge will result in a net removal of phosphorus from the wastewater (Yeoman, 1988a, Tchobanoglous et al, 2003). Previous studies have reported that the presence of acetate was essential to the formation of PHB under anaerobic conditions, providing a competitive advantage for the PAO. This anaerobic zone, typically at the head of a treatment process is often called a "selector" because it provides conditions favorable to the proliferation of PAOs. PAOs typically prefer low-molecular-weight fermentation product substrates, thus they require an anaerobic zone in which the fermentation of the readily biodegradable COD is converted to acetate (Tchobanoglous et al., 2003). The anaerobic zone is also required to prevent the utilization of VFAs with 0 2 or NO x. The competitive advantage of PAOs in the anaerobic zone resides in their polyphosphate storage ability, thus giving them energy to assimilate and metabolize acetate. Other aerobic heterotrophic bacteria do not have such a mechanism for acetate uptake and become starved, while the PAOs assimilate COD in the anaerobic zone. It was also noted that the PAOs form very dense and good settling floes in the activated sludge, providing an additional benefit (Tchobanoglous et al, 2003). The stored intracellular polyphosphate in polyphosphate-accumulating organisms can be classified in two categories; the "volutin" granules and the fraction loosely bound to the cell membrane. Volutin (polyphosphate) granules can easily be observed under traditional microscopy or by staining with many basic dyes, while the bound fraction is more difficult to measure. PHA was also found to be stored in the form of granules (Russell et al., 2006). In fact, it was reported that the PHA granules originating from the EBPR process were composed of random co-polymers, with various monomer compositions. The physical characteristics of these polymers can vary greatly but several have plastic-like consistencies. PHA granules possess an affinity for fat soluble dyes such as Sudan Black and Sudan Blue. These staining characteristics are useful for observing the granules using visible light microscopy. 11 2.6 Methanol Used in EBPR 2.6.1 Methanol as a Denitrification Substrate Denitrification is one of multiple pathways of the nitrogen cycle. This particular pathway is a biochemical reaction mediated by microorganisms that transform nitrates to molecular nitrogen. This reaction is the result of the coupling of the electron transport by the respiratory chain and the energy production via oxidative phosphorylation (Naidoo, 1999). The reaction is presented in equation 1 2N03- (+5)-» 2N02" (+3) 2NO ( g ) (+2 ) N 2 0 ( g ) (+1 ) -> N 2 ( g ) (0) (1) Where [+5] denotes the oxidation state of the nitrogen atom in each compound The biological denitrification processes are typically used after secondary treatment nitrification processes, in order to reduce the nitrate/nitrite in the effluent. An exogenous carbon source is typically provided to serve as an electron donor that can be oxidized with nitrate or nitrite (Tchobanoglous et al., 2003). Methanol is typically the substrate of choice, because of its resulting low cell yield, efficiency, and cost. A study conducted in 1995 published in Chemical Marketing Report found that methanol was the cheapest source of commercially-based carbon. About USD $0,011 of methanol are required for the denitrification of each m3 of water containing 30 mg/L of NO3-N/L (Bilanovic et al., 1999). This substrate would thus be very attractive if it could also be used to remove phosphorus. According to Tchobanoglous et al. (2003), the stoichiometry of biological denitrification in the presence of methanol is the following: 5CH3OH + 6 N0 3" -> 3N 2 + 5C0 2 + 7H 20 + 6OH" (2) 2.6.2 Methanol-Consuming Organisms (Methylotrophs) In order to identify the potential organisms that could grow in the biological reactor, it is a good strategy to define the various microbiological groups and their characteristics: • An autotroph is an organism that derives its cell carbon from atmospheric C 0 2 by means of fixation and reduction. This is an energy-consuming process that 12 typically uses sunlight or reduced inorganic compounds as an energy source (Crowther, 2004). • A heterotroph is an organism that obtains its cell carbon by incorporating directly some form of reduced carbon substrate • A methylotroph is an organism that derives its energy, and often its cell carbon from reduced carbon molecules that have no C-C bonds (these molecules are also called Ci compounds) (Marco, 2004). • Methanotrophs are methane oxidizing bacteria; they are a unique group of methylotrophs that use methane as their sole substrate source for energy and carbon. Some methanotrophs can use methanol as a carbon source, but none of them can grow on multi-carbon compounds. True methylotrophs assimilate the carbon at the level of formaldehyde. This group of true methylotrophs is further divided in two groups; the methanotrophs and the non-methane methylotrophs. Non-methane methylotrophs use mostly methanol, some use methylated compounds as their carbon source. Some of them are denitrifiers (Crowther, 2004). (Murrell et al., 1998 ; Crowther, 2004). This study focused primarily on investigation of non-methane methylotrophs. While these compounds require Q compounds for growth and energy, they all tend to use formaldehyde as their first metabolic intermediate (Crowther, 2004). The pathways of Ci assimilation by non-methane methylotrophs are presented in Figure 4. We notice that, after the Ci compound has been metabolized as formaldehyde, three pathways can be used to integrate the carbon into the biomass. 13 CHd (Methane) j - 0 2 + 2 CHgOH (Methanol) I' • - 2 e-CHgO (Formaldehyde) Figure 4: Pathways of the metabolism of single-carbon compounds by methanotrophic bacteria Source: Chistoserdova, 2004 Note: RuMP stands for ribulose monophosphate, TCA stands for tricarboxylic acid and CBB stands for Calvin-Benson-Bassham cycle. As mentioned before, the methylotrophs are restricted to a few genera, as observed in Figure 5. 14 Figure 5: Genotypes associated with methylotrophs (Crowther, 2004). While Figure 4 illustrates a fairly reductionist view of the metabolic pathways involved in the assimilation of a C\ compound into biomass, the complete cycle is significantly more complicated. It was assumed from the beginning that the cycle that would mediate most of the carbon assimilation in the experiments would be Serine/TCA cycle. The complete Serine/TCA cycle is presented in Figure 6. 15 Methanol Methylamine : Methanol I Methanol . 1 dehydrogenase V Formaldehyde production Formaldehyde < — Methylamine Methylamine dehydrogenase NAD(P) NAD iP iH , - NAD N A D H - - r FotmaldPhyde - r For male CO, [Energy \ r . I Metabolism Foimaldohydo Foimato | oxidation oxidation | Methylene H4F Assimilation F A D H 2 FAD*. BIOSYNTHESIS Hydroxypyruvate reductase •. 2 PGA A T P * Glycerate N A D 4 '4 N A D H f Hydroxypyruvate Serine MCO2 Methacrylyl-C6A\{jJVtjso"b'u^ ' phydroxyisobutyryl CoA 2 J^tls** Mi>tiwlsuccinyl:CoA cm PropionylCoA CO?-AA \ / N A D H X ^ N A D Succinyl-CoAJ F A D * . * — • F A D H 2 - Malate CO. N A D P . N A D P H Ethvlmalonyl-CoA Glyoxylate JButyr yrcoA^ regeneration cycle Crotonyi-GoA'j Methylene H4F Serine cycle Malyl-OoA . Glyoxylate f Glycine flalyl- COA "* t " • ' fo-Hydroxybutyryl-CoA _ L , N A D P : Acetyl-CoA N A D P H Acetdac'ety^CgA^ BIOSYNTHESIS Figure 6: Metabolic degradation of methanol (Crowther, 2004) 2.6.3 Stoichiometrv of EBPR Based on the known phosphorus removal mechanisms, acetate uptake in the anaerobic zone is critical in determining the amount of PHAs that can be produced and thus, the amount of phosphorus that can be removed by this pathway. This amount of phosphorus being removed, by biological storage, can be estimated from the amounts of readily biodegradable COD that are available in the wastewater influent; most of this COD will be converted to acetate in the short anaerobic detention time. These assumptions are used to evaluate the stoiochiometry of biological phosphorus removal: 1- 1.06 g of acetate / g of readily biodegradable COD will be produced. Most of the COD fermented will be converted to VFAs due to low cell yield. 2- a cell yield of 0.30 g VSS/g acetate 3- phosphorus content of 0.3 g P/g VSS According to these assumptions, 10 g of readily biodegradable COD is required to remove 1 g of phosphorus, by phosphorus storage in cells. Normal cell synthesis will increase that phosphorus removal efficiency (Tchobanoglous et al, 2003). When initiating this research project, no data regarding the ratio of methanol to phosphorus that might be required for phosphorus removal was available. The first experiments with methanol were thus based on a 1-to-l carbon ratio, between methanol and acetate. 2.6.4 Previous Studies on Biological Phosphorus Removal with Methanol A study done by Cho et al. (2004) used sequential methanol and acetate feed on a denitrifying phosphorus removal bacterial culture. They showed a significant improvement in denitrification, as well as phosphorus removal, compared to using either substrate alone. They noticed that the amount of PHB stored by the phosphorus-accumulating organisms could improve aerobic denitrification. A schematic of their process is presented in Figure 7. 17 1)MeOH 2) HAc 3) MeOH HAc 1 J Influent Effluent AX1 AX2 Owe Return sludge Sludge waste Figure 7: Flow scheme for a continuous bench scale pilot plant feeding with methanol and acetate Source: Cho et ah, 2004. The most interesting finding of this study is that methanol did not affect phosphorus release, phosphorus uptake or even secondary release in the system. Only the denitrifiers seemed to benefit from the dual feed system (Cho et al., 2004). The study conducted by Louzeiro et al. (2002) was aimed at determining the potential for denitrification and phosphorus removal in a full-scale sequencing batch reactor, with and without methanol as an external carbon source. Two denitrification rates were observed in the experimental SBR; an initial fast rate and a slower second rate. It was hypothesized that the initial fast denitrification rate was proportional to methanol concentration, until a maximum denitrification rate of 19 mg NOx-N/g MLVSS/day was reached. It was reported that after depletion of methanol, denitrification continued using available natural carbon in the influent. The authors hypothesized that methanol was probably not used as a carbon source for the enhanced biological phosphorus removal process, but instead played a critical role by depleting the nitrates and allowing EBPR to take place (Louzeiro et al., 2002). In summary, a lot of information was found on phosphorus removal processes, conventional biological phosphorus removal, as well as organisms involved in bio-P. Literature was also available on methanol degradation, assimilation and use in denitrification applications. However, the literature was found to be scarce on methanol-induced biological phosphorus removal. From this point, it was realized that laboratory experiments would be required to investigate this further. 18 CHAPTER III - Materials and Methods This chapter will describe the methodologies and the various methods used in this study. The first section of the chapter will describe the sequencing batch reactor (SBR) setup as well as feeding regimes; the second section will describe the jar-test setup that was used, while the last section will describe the sampling and analysis methods. 3.1 Sequencing Batch Reactor (SBR) 3.1.1 Reactor Setup/Operation The sequencing batch reactors that were used in this study were two, 10-liter Plexiglas containers, fitted with mixers. The top of the reactors was partially open to the atmosphere, due to the multiple holes required for probe and mixer insertion. Six access ports to the reactor facilitated the control of the process: one was used for sampling, one as an overflow port, one as a decant port, one as the primary feed port, one as a sludge wasting port and one to connect an air line to the diffuser stone at the bottom of the reactor. A schematic of the sequencing batch reactor is presented in Figure 8. The reactor was fully automated and all controls were performed with the help of electronic timers. The reactor was operated on a 24 h basis, with 4 h cycles, for a total of 6 cycles per day. The reactor was operated for approximately 500 days, from the first reactor setup to the end of the experiments. The target SRT/HRT was 60. Throughout the operation, the hydraulic retention time was fixed at 0.5 days, while the solids retention time was typically 30 days. The target TSS concentration was 3 g/L. Refrigerated sludge from former nitrification reactors, combined with some sludge from UBC's pilot sewage treatment plant, provided the initial reactor seed. The typical SBR operation started with the peristaltic feed pump pumping approximately 3.3 L from a 60 L container of primary feed into the reactor. At the end of the feed cycle, the reactor was simply mixed with no aeration, creating an anoxic/anaerobic environment. After the anoxic/anaerobic cycle, the air was turned on to initiate an aerobic cycle. Right after the aerobic cycle, the air and mixing was turned off for the settling and decant phase. The secondary feed 19 was typically activated right after the feeding phase and ended with the aeration phase. Sludge was wasted towards the end of the aerobic phase from the completely-mixed reactor with a peristaltic pump. Each cycle was composed of the following phases: • Feed phase: 2 minutes: Mixing, no aeration • Anoxic/anaerobic cycle: 48 minutes - Mixing, no aeration • Aerobic phase: 162 minutes - Mixing, aeration • Settling/decant phase: 28 minutes - No mixing, no aeration Data Logging: pH, D O , O R P Air • Overf low Decant port (3.3 L / Cyc le ) > Sampl ing port Pr imary feed (3.3 L / Cyc le ) Synthetic feed S ludge wasting Figure 8: Schematic of the sequencing batch reactor used for most of the experiments in this study The sequencing batch reactor, as well as the surrounding setup, is also illustrated in Figure 9. The picture was taken while the reactor, including the controls were being tested with tap water. 20 Figure 9: Photo of the two empty SBRs and surrounding setup A computer was used to continuously monitor three physico-chemical parameters: the pH, the dissolved oxygen (DO) and the oxidation-reduction potential (ORP). The pH, DO and ORP measurements were taken using simple electrolytic probes. The pH probe was an Orion Ag/AgCl combination electrode. The DO probe was an Oxyguard Model II Silver cathode/lead anode. The ORP probe was a Broadley-James model 51. 3.1.2 Synthetic Sewage Composition Synthetic sewage was used in this study, in order to reduce the number of unknowns. The synthetic sewage aimed at providing every macro and micronutrient necessary for microbial growth. In order to do this, the recipe was based on previous work by Grady et al. (1999) and Shiskowski (2004). The main change to these recipes was the source of carbon. Instead of feeding the reactor with volatile fatty acids (mainly acetate), it was fed methanol as the sole carbon substrate. The synthetic feed typically used tap water. The pH of the synthetic feed was 21 typically 9.0 - 10.0, an increase caused by the addition of Na2C03. The recipe is presented in Table 1. Table 1: Macro and micronutrient composition of a typical 60 L synthetic sewage batch I Nutrient CaCi2 Quantity 62.5 mg/L Final Concentration in Reactor; > 20.8 mg/L MgSCU 29.7 mg/L 9.9 mg/L NH4C1 114.5 mg/L 38.2 mg/L Yeast Extract 20 mg/L 6.7 mg/L K2HPO4 33.7 mg/L 11.2 mg/L Methanol 0.1 mL/L 0.03 mL/L Na 2C0 3 468 mg/L 156 mg/L CuS0 4 12.5 ug/L 4.2 mg/L ZnS04 220 ug/L 73ug/L MnS0 4 82.5 ug/L 27.5 ug/L Na2Mo4 2.5 ug/L 0.83 pg/L C0CI2 0.405 |ag/L 0.135 ug/L NiCl 2 0.405 ug/L 0.135 ug/L FeS04 1.9 mg/L 0.6 pg/L 3.1.3 Secondary Feed Composition It was discovered early in the experiments that the methanol feed had to be separated into a small volume of primary feed and a continuous secondary feed. This had the purpose of reducing the "shock load" on the system. In response to this phenomenon, a secondary feed was installed and pumped dilute methanol using a peristaltic pump at a rate of 0.8 mL/min for the anoxic/anaerobic and aerobic periods (210 minutes per cycle) at a concentration of 0.6 % v/v to give a total addition equivalent to an addition of 1 mL of undiluted methanol per cycle. 3.2 Batch Tests Batch tests were performed using a multi-place magnetic stirrer. Six 500 mL Erlenmeyer flasks were filled with 400 mL of mixed liquor suspended solids from the bench scale SBR and were stirred using magnetic bars. Oxygen was bubbled into the Erlenmeyers with a compressed 22 air line and a plastic manifold, to distribute pressure equally to all ports. In order to create anoxic/anaerobic conditions, the air supply was terminated and nitrogen was bubbled into the Erlenmeyer flasks. A 50 mL syringe was used to remove samples from the Erlenmeyer flask The setup is presented in Figure 10. Figure 10: Batch test apparatus 3.3 Analytical Methods 3.3.1 General Analytical Methods The following parameters were measured or examined during the course of this study: • Orthophosphate-P • Ammonia-N • Nitrate/Nitrite-N • Total Suspended Solids • Formaldehyde • Methanol • Volatile Fatty Acids (VFA) • Glycogen • Polyphosphates • PHB (Poly-hydroxybutyrate) The various methods are presented in Table 2. 23 Table 2 : Analytical Methods Analyses Analysis method . Container Preservation technique Sample . Volume Needed Total Suspended Solids Standard Methods # 2540 D Plastic bottle Filter the stirred sample 20 ~ 50 ml COD Standard Methods # 5220 D, Closed Reflux, Colorimetric Method Culture tube 25x150 mm Preserve sample by acidification to pH < 2 using H 2S0 4 , if the sample is not analyzed in the same day. 10 mL Filt-COD Standard Methods # 5220 D Closed Reflux, Colorimetric Method Culture tube 25x150 mm Filtered, straight into tube. Preserve sample by acidification to pH < 2 using H2SO4, if the sample is not analyzed in the same day. 10 mL Total-P Standard Methods #4500-P H. Flow Injection Analysis Glass bottle Preserve samples by acidification to pH $2 (with H2SO4) and storing at 4°C ~ 10 mL Ortho-P Standard Methods #4500 Ortho-P G. Flow Injection Analysis Glass bottle Preserved with phenylmercuric acetate ~ 15 mL TKN Standard Methods #4500-NorgD. Flow Injection Analysis Plastic bottle Preserve samples by acidification to pH < 2 (with H2SO4) and storing at 4°C ~ 10 mL Ammonium-N Standard Methods # 4500-NH3 H. Flow Injection Analysis Vial For preservation for up to 28 days, freeze at -20°C unacidified, or preserve samples by acidification to pH <2(withH2S04) and storing at 4°C ~ 15 mL Nitrate-N Standard Methods #4500-NO3 B. Ultraviolet Spectrophotometric Tube For preservation, add 1 drop of HCL 1 N. ~ 15 mL Nitrite-N Standard Methods # 4500-NO2 B Tube For preservation, add 1 drop of HCL IN ~ 15 mL VFA GC packed column method GC Vial 1 drop of 2 %H 3 P0 4 into GC vial with ~2mL 24 sample Methanol GC packed column method GC Vial 1 drop of2%H 3 P0 4 into GC vial with sample - 2 mL Formaldehyde Modified NIOSH method - See section 3.3.5 Test tube Analyzed right away ~ 15 mL Glycogen See Section 3.3.4 Immediate treatment Samples are dried ~ 5 mL PHB Microscopic examinations with dye addition - See section 3.3.3 Dried sample on an examination slide Smudge sample evenly on the slide ~ 1 drop Polyphosphates Microscopic examinations with dye addition - see section 3.3.2 Dried sample on an examination slide Smudge sample evenly on the slide ~ 1 drop «fote: the Standard Methods edition was : Tom Cleresci et al., 1998. 3.3.2 Chemical Analyses by Gas Chromatography A gas chromatography (GC) system model HP5880 Series 2 (Hewlett Packard) was used for all volatile fatty acid (VFA) and methanol analyses. Carrier gas was helium delivered at a flow rate of 36 cm/s and the instrument detector was a flame ionization detector (FID) used for all analyses. All samples were filtered prior to analyses and stored in the dark at 4oC. Volatile Fatty Acids - Chromatographic Conditions The standard method used for VFA analyses was based on Hewlett Packard Application Note 228-398 (HP January 1998). The chromatographic column used was HP-FFAP 25 m (length) x 0.32 mm (diameter) x 0.5 um (film thickness). All aqueous samples for VFA analyses were preserved with one drop of 2% phosphoric acid (H3P04) and the same solution (2% H3P04) was used to ensure pH was less than 3.0 prior to injection for GC analyses. Injection volume of 1.0 ul was used throughout. Methanol - Chromatographic Conditions The standard method used for methanol analyses was based on Hewlett Packard Application Note 228-398 (HP January 1998) with minor modifications outlined as follows. The chromatographic column used was DB-WAX 15 m (length) x 0.25 mm (diameter) x 0.25 um (film thickness). Aqueous injection of 1.0 ul of neat sample (unpreserved and without pH 25 adjustment) was used for methanol analyses. Oven temperature was held at 30oC for 2 min and then ramped to lOOoC in 15o/min increments. The methanol peak for each analyses was clearly evident between 2 and 3 minutes post injection. 3.3.3 Polyphosphates - Microscopic Examination. Polyphosphate concentrations were qualitatively estimated from stained slides, examined under optical microscopy. This allowed the screening of vast amounts of samples without the tedious analytical procedures of polyphosphate analysis. Furthermore, this method allowed us to determine which types of organisms were responsible for the bulk of polyphosphate accumulation. The selected staining method for polyphosphate determination was the Neisser stain. The method was taken from Eikelboom (2000). The method is as follows: -Prepare Solution 1 and Solution 2. The Solution 1 is made of two parts; Part A and Part B. Solution 1, Part A: • 0.1 g of methylene blue • 5 mL of ethanol 95 % • 5 mL of glacial acetic acid and • 100 mL of distilled water. Solution 1, Part B • 3.3 mL of crystal violet (10 % w/v in 95 % ethanol) • 6.7 mL of ethanol 95 % and . • 100 mL of distilled water. Solution 2: • 33 mL of Bismarck Brown (1 % w/v aqueous) • 67 mL of distilled water Procedure • Mix solutions A and B in a 2:1 ratio respectively, this will be the final Solution 1 • Lightly heat-fix the slides with a Bunsen burner • Stain 30 seconds with Solution 1 • Rinse gently with water • Stain 1 minute with solution 2 • Rinse well and blot dry 26 • Examine under the microscope. Blue-violet dots are positive while yellow-brown is negative to polyphosphate 3.3.4 PHB -Microscopic Examinations The PHB staining method was also taken from Eikelboom, 2000. The method is the following: Prepare Solutions 1 and 2. Solution 1: • 0.15 g Sudan Black B(IV) • 50 mL of ethanol 60% Solution 2: • 0.25 g Safranin O • 50 mL of distilled water Procedure: • Heat-fix the slides with a Bunsen burner • Stain 10 minutes with Solution 1 • Rinse gently with water • Stain 10 seconds with solution 2 • Rinse well and blot dry • Examine under the microscope. Blue-black intracellular granules are positive while the rest should be pink or clear 3.3.5 Glycogen Method Glycogen measurements were performed using the anthrone method, as described by Murray (1981), Morris (1948) and Jenkins et al. (1993). • Take 5 mL of sludge • Centrifuge for 5 minutes at 10 000 g • Remove supernatant • Dry aliquot at 103 °C for 24 h • Determine the mass of the solid sample in a plastic centrifuge tube • Digest samples with 1 mL of KOH 30 % (w/v) at 100 °C for 3 hours. • Let the tubes cool • Add 3 mL of water and 8 mL of ice-cold ethanol to precipitate the glycogen 27 • Centrifuge tubes for 15 minutes at 10 000 g • Remove the supernatant • Re-suspend solids in 5 mL of distilled water to resolubilize the glycogen • Take 0.1 mL of the diluted sample and add 0.9 mL of distilled water (1:10 dilution) • Prepare the anthrone reagent: o Add 200 mg of anthrone to 5 mL of ethanol o Add up to 100 mL with 75 % sulfuric acid (v/v) • In an ice-cold bath add 5 mL of anthrone reagent and vortex thoroughly • Transfer tubes to a boiling bath for exactly 10 minutes • Return to the ice bath until the sample cooled • Prepare glucose standards by adding 100 mg of glucose and 150 mg of benzoic acid to 100 mL of distilled water and dilute as appropriate • Measure absorbance at 625 nm 3.3.6 Formaldehyde Method The method used for formaldehyde analysis was the original NIOSH Method 3500 (NIOSH, 1994). This spectrophotometric method uses the chromotropic acid assay technique. 3.4 Sampling QA/QC Whenever possible, every 10th sample was analyzed in duplicate. Three quality control samples were simultaneously taken through the analysis, to confirm the results. Standard quality assurance and quality control procedures were followed throughout the analyses; samples were analyzed several times to determine the variance and an internal standard was injected to check the detector response. 28 CHAPTER IV - Results and Discussion 4.1 Background Information This study on methanol Bio-P removal was undertaken with the aim of investigating an unusual phosphorus uptake by an intermittently-aerated, activated sludge process, fed with methanol. The observed phenomenon was usually a luxury uptake of phosphorus in a sequencing batch reactor (SBR) operated under low dissolved oxygen concentration - potentially under simultaneous nitrification and denitrification (SND) conditions. Given the fact that methanol-induced, biological phosphorus removal had received little attention at the onset of this project, information upon which to base our experiments was limited. The initial protocol was based on previous nitrogen-removal studies, such as that performed by Shiskowski (2005). 4.2 First Phase 4.2.1 Initializing Conditions for Nitrification The reactors were operated for a period of 4 months before the beginning of any sampling. The reactors were fed with methanol from the first day. The methanol was fed all at once in the feeding phase of each cycle. Initially, it was noticed that settling of the sludge was good, although no accompanying increase in the concentration of solids was observed, even without any sludge wasting. Quick field test kits confirmed that nitrification appeared to be incomplete, and that denitrification did not seem to occur. In order to determine the cause of this disconnect, the pH and dissolved oxygen (DO) profiles logged by the computer were examined (Figure 11). The cycles, spanning an entire day, were identical, showing little or no improvement in the treatment process. The pH curve showed a significant drop - according to this data, pH decreased from 8.4 to 6.8. This was surprising, given that nitrification was never complete. Alkalinity consumption was then examined: 29 The alkalinity of a solution can be determined with the following formula (3): Alkalinity (equivalents/m3) = HC0 3 + C 0 3 2 + OH" - H + (3) According to the calculations, there existed approximately 4.416 mmol/L of alkalinity or 8.8 meq/L. A quantity of 8.8 meq/L would thus be expressed as 440 mg/L as CaC0 3 (Tchobanoglous et al., 2003). The amount of alkalinity required to carry out nitrification can be estimated by the following equation (4): N H 4 + + 2HC03" + 20 2 -> N0 3" + 2C0 2 + 3H 20 (4) For each gram of ammonia nitrogen (as N) converted, 7.14 g of alkalinity as CaC0 3 will be required [2 x (50 g CaC03/eq)/14]. In the feed, 114.5 mg/L of NH4C1 was present. Given the following information: • Molecular mass of NH4C1 = 53.45 g/mol • Molecular mass of N = 14 g/mol Thus there existed 14 g/mol / 53.45 g/mol *114.5 mg/L = 30 mg/L-N. • 30 mg/L nitrogen (as N) • Nitrification requirements = 30*7.14 = 214 mg/L. According to these calculations, there was sufficient alkalinity to carry out complete nitrification, as well as microbial growth (214 mg/L alkalinity as CaC0 3 consumed vs 440 mg/L as CaC0 3 in the feed). A close examination of the reactor was undertaken to identify what might be going on with the biological process. Dissolved oxygen, pH (Figure 11), ORP levels were monitored (Figure 12), as well as nitrate, nitrite, ammonia and orthophosphates (Figure 13). 30 12 10 • pH DO 20Q00 Seconds Figure 11: Typical pH and DO profiles over a 5 h period for the SBR reactor 140 120 100 ? 80 a. O 60 40 20 • 1 1 1 1 > 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 Seconds Figure 12: Typical ORP profile over a 5 h period for the SBR reactor Phases 3 D. to O O 100 150 Cycle time (min) E Figure 13: Nitrate, nitrite, orthophosphate and ammonia profiles over one cycle of the biological reactor. Results presented in Figure 11 shows that the process was fairly stable over a 24 h period. The pH drop (close to 1.5 pH units) suggests a potentially significant nitrification process It was thus surprising to see such a pH drop, when only half the alkalinity should theoretically have been consumed. One possible explanation could be an accumulation of formic acid in the reactor. However, the data from the dissolved oxygen (D.O.) profiles indicate that carbon oxidation was probably complete after two hours, after which the nitrification started. The data suggested that nitrification was not complete, given that the rapid rise in reactor DO, characteristic of ammonia disappearance, was not observed. Results from Figure 12 confirmed these suspicions. Oxygen reduction potential values remained fairly grouped, with as little as 25 mV variation between the anoxic and the aerobic phase. This suggests that the nitrification was not complete, and that oxygen was being taken up by the heterotrophic biomass; otherwise the ORP would have spiked significantly. Data from the Figure 13 strengthened these suspicions. From Figure 13, it was seen that nitrate levels dropped from about 11.0 mg N/L to about 6.0 mg N/L. This suggests incomplete denitrification. During the aerobic phase, ammonia was oxidized to nitrate at a rate of about 10 mg N/L per cycle, which seems to indicate incomplete nitrification. Orthophosphate concentrations were relatively stable throughout the cycle, suggesting that no enhanced phosphorus removal had occurred. 32 After examination of the data base in this stage, it was believed that the operational problems were probably caused by either a micro/macronutrient deficiency, or a methanol acclimation problem. Lack of biomass growth, poor settling characteristics of the sludge and overall slow reaction rates led us to believe that there was a problem with the reactor. 4.3 Excess Aeration Experiments - Getting Nitrification To Work In order to investigate the possibility of a shock load of methanol to the reactor in the initial feed phase, the methanol concentration in the feed was reduced from 0.25 mL/L to 0.1 mL/L. The effects of this shift were very significant; a significant growth in the biomass was noted: the MLSS concentration increased from 1 g/L to more than 2.0 g/L in about two weeks, while little or no growth had been observed before. In addition to rapid biological growth, it was observed that the pH variations were still very significant. In fact, despite efforts to reduce the amplitude of pH variations with more alkalinity addition, the pH still decreased from 8.6-8.8 at the end of the anaerobic phase, to about 7.0-7.4 at the end of the aerobic phase. Attempts to control the pH, using bicarbonate instead of sodium carbonate, failed and resulted in the pH dropping below 6.8 (which is often reported as a critical pH for nitrifiers (Tchobanoglous et al., 2003)). Complete nitrification was observed for the first time, and denitrification seemed to occur. The drop in pH suggests that a base is being consumed or an acid is being produced. It also suggests that the system is poorly buffered. If the carbonate was the primary buffer (because the proteins and the phosphate are also low concentrations) then a drop in pH with a production of small amounts of acid could be expected because the pKa of the carbonate is about 6.2 and 10. The range of pH 7 to 9 is at least a pH unit away from either pKa so the buffering is limited by pKa as well as the low chemical concentration. Phosphate is a bit better at 6.6 but there is nothing to hold the pH around 9 except the ammonium at a pKa of 9.3. Since the ammonium concentration is low it would not be much of a buffer. The results are presented in the Figures 14 and 15. In Figures 14 and 15, Phase 1 was the end of the previous cycle (settling), Phase 2 was a combination of feed and anoxic/anaerobic cycle, Phase 3 was the aerobic phase, Phase 4 was the settling phase and Phase 5 was the beginning of the next phase. From Figure 14, with the help of the ORP and DO signals, it can be observed that the feed was added at 12:00, causing a drop in ORP and DO and an increase in pH (the pH of the feed was about 9.0). The ORP drop ended at point "A" where the slope changed drastically; it coincided with the beginning of the aerobic phase. At point "B", the ORP curve had another 33 inflexion point, and according to our hypothesis, marked the end of the carbon oxidation phase. This also appears to be the point where denitrification stops. This hypothesis was corroborated by the COD data in Figure 15, which shows a stabilization of the COD curve. The remaining 40-50 mg/L of COD is believed to come mainly from the macro and micro-nutrients in the primary feed. From Figure 14, from point B, the slope of the ORP increased significantly until it reached point C, which is believed to be the end of the nitrification phase. This hypothesis is supported by the DO data, pH data (pH increases when the nitrification is complete due to the stripping of H2CO3 (C02)), levelling off of the NO x curve and disappearance of residual ammonia (see Figure 15). It is surprising to note that the NO x and NH 4 concentrations decreased simultaneously at the beginning of Phase 3, which was aerobic. It must also be noted that the dissolved oxygen concentrations in the first part of the aerobic phase were very low. The end of the NO x concentration decrease at 13:40 seems to coincide with the end of the carbon oxidation/uptake phase. This data could suggest a simultaneous nitrification and denitrification phenomenon taking place, as long as there is a denitrification substrate readily available. It has been reported in the literature (Lemaire et al., 2006) that simultaneous nitrification and denitrification is possible, transforming the ammonia into nitrous oxide, without a significant step through nitrates. It was surprising to note that the NO x concentration kept decreasing at the same rate as ammonia in the aerobic phase, until the point "B". By doing a nitrogen balance, it was noted that a total of 5 mg N/L was removed in the 4 h cycle, and that about 3.8 mg N/L was removed between 12:00 and 13:30, suggesting that SND might, in fact, require low dissolved oxygen concentrations. Two distinct kinetics of nitrification can also be observed; one before point B and one after. The competition for oxygen between carbon oxidizers and nitrifiers could explain the slower nitrification rate, until all of the substrate is oxidized. The COD profiles were as expected; however, there was still a residual COD of 45-50 mg/L that seemed to be present throughout the experiments. Finally, the phosphorus levels remained constant throughout the 4h cycle. This suggests the complete absence of additional phosphorus release or accumulation. This was somewhat unexpected, since phosphorus release typically occurs under anaerobic conditions. A complete removal of all NO x would thus have to occur, before the necessary anaerobic conditions can take place, and initiate phosphorus release. This specific experiment showed that nitrification and denitrification worked reasonably well. However, in light of these results, it was obvious that the reactor received excessive 34 amounts of air; the ORP levels reached 300 mV and did not drop below 25 mV, the carbon and nitrogen were both completely oxidized before the end of the aerobic phase, the pH decreased and then increased, indicating the end of alkalinity consumption, and the dissolved oxygen concentrations reached more than 4 mg/L. These factors, collectively suggested that the microorganisms might have been under oxic stress. These results initiated the next experiments, focusing on lowering the aeration levels to achieve anaerobic conditions. 35 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 Time (abs) Figure 14: pH, DO and ORP profiles over one cycle of the methanol-fed anoxic-aerobic SBR. NOx (mg N/L) 0-PC4 NH4 [mg N/L) COD (mg/L) 120 11:30 12:00 12:30 13:00 13:30 14:00 14:30 Time (abs) 15:00 15:30 16:00 16:30 Figure 15: NOx, orthophosphates, ammonia and COD concentration profiles over one cycle of the methanol-fed anoxic-aerobic SBR. 4.4 Setting Up Anaerobic Conditions - Getting Denitrification To Work The previous data demonstrated that the alternation of anoxic/aerobic conditions did not trigger biological phosphorus accumulation and release. This was confirmed by several sampling programs (data not shown). The problem was diagnosed as an inability to completely denitrify, that prevented the ORP levels from dropping. It was previously hypothesized that high methanol concentrations may have had a toxic effect on the bacteria, thus inhibiting denitrification. It was also previously hypothesized that high DO concentrations might have exerted an oxic stress on the microorganisms, in addition to potentially inhibiting a simultaneous nitrification and denitrification effect. In order to address this problem, the feed pattern was changed; instead of injecting the carbon substrate all at once at the beginning of the cycle, a fraction of it would be injected in the primary feed and the rest would be injected slowly, as a secondary feed, throughout the anaerobic and aerobic cycles. It was surmised that this low-concentration, secondary feed regime would depress the dissolved oxygen levels, while maintaining low carbon concentrations in the reactor; thus, nitrification could still take place without significant competition from the heterotrophs. The initial attempt failed, since methanol was virtually absent from the primary feed and it was realized that methanol had to be present in significant concentrations at the beginning of the anoxic phase. Further modifications were made to the operation and the "optimal conditions" were found to be lA of the substrate load injected at the beginning of the anoxic phase (0.1 mL/L * 3.3 L = 0.33 mL), and % of the carbon load injected in the reactor throughout the anoxic/anaerobic/aerobic phase (1 mL). Alternating anoxic, anaerobic and aerobic conditions were reached within a few weeks after the feeding mode change. Nitrification and denitrification were observed to be complete under those new operating conditions and a significant change in phosphorus levels throughout the cycles could be observed. The results of the sampling program are presented in Figure 16 and Figure 17. Two inflexion point "knees", A and B, were observed from the ORP profiles shown in Figure 16, indicating the end of the carbon oxidation (A) and the end of nitrification (B). This was confirmed by examining the COD and ammonia profiles from Figure 17. The inflection point at the end of denitrification around 12:10 (transition between anoxic and anaerobic) was so subtle that it could barely be noticed from the ORP graph. 38 The pH exhibited a profile similar to what was reported in the previous sampling programs; a significant drop in pH from 8.4 - 8.6 to about 7.0 - 7.1. Surprisingly enough, the pH seemed to decrease slightly, even during the anoxic phase, when alkalinity recovery would be expected. This drop when the ammonia levels drop may be additional C O 2 waste from the metabolism. This observation still remains unexplained and presents an interesting "mystery". This could potentially be explained by a dual substrate utilization by ammonia oxidizers (NH3" and methanol). Similar to what was observed in previous experiments, it is interesting to note that the NO x and NFL* concentrations decreased simultaneously in the first 45 minutes of Phase 3 (aerobic) of Figure 16. Termination of this simultaneous nitrification and denitrification phenomenon was previously thought to be caused by a carbon shortage, since it coincided with the point A from Figure 16. It must be noted that, in this experiment, some substrate was injected throughout the anoxic/anaerobic and aerobic phases, so the presence of carbon should theoretically not be an issue, unless if the rate of carbon addition was too low. No denitrification was observed after the point "A", and this was probably related to the dissolved oxygen levels exceeding 2 mg/L. A very interesting feature of Figure 17 - is the appearance of a slight variation in the orthophosphate concentration; an increase in the order of 30 % in the anaerobic phase and a small decrease in the aerobic phase. It can be observed in Figure 17 that phosphorus concentration changes seemed related to the ORP levels. Since phosphorus uptake is thought to be proportional to the amount of released phosphorus, a longer anaerobic period could increase the amount of phosphorus released and subsequently stored in aerobic conditions. In fact, the study conducted by Louzeiro et al. (2002) successfully correlated the duration of the phosphorus release period with the amount of phosphate released per gram of MLVSS. The change in the feeding pattern from "all at once" to lA of the carbon load at the beginning and % of the carbon injected slowly throughout the cycle seemed successful in creating favourable conditions for initiating the enhanced/luxury biological phosphorus removal process. Simultaneous nitrification and denitrification conditions only seem to last for a relatively short period (~ 45 minutes), until the bulk of carbon oxidation is complete; then, denitrification completely stopped. In this scenario, increasing the length of the anaerobic phase would make sense since it would potentially release more phosphorus, and more phosphorus would subsequently be taken up in the aerobic phase. 39 14 Phase # 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 Time (abs) Figure 16: pH, DO and ORP profiles over one cycle of the methanol-fed anoxic-aerobic SBR. o 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 Time (abs) Figure 17: NOx, orthophosphates, ammonia, COD and TOC concentration profiles over one cycle of the methanol-fed anoxic-aerobic SBR. 4.5 Experiments With Low Dissolved Oxygen Levels Data from the previous experiments suggested the presence of some sequential phosphorus release and uptake in alternating anaerobic/aerobic conditions, with methanol feed. The release and uptake were, however, partial, with only a variation of maximum +/- 40% of the phosphorus concentration in the reactor. It was then hypothesized that over-oxidation in the aerobic phase might negatively affect the phosphorus uptake. The system by then was completely nitrifying the ammonia in the aerobic phase, but this required extremely high ORP levels (> 200 mV). The aeration levels were reduced in order to keep lower ORP levels throughout the cycle. Practically, the volumes of air were gradually decreased until the peak ORP value reached about 60-100 mV. The results of the sampling programs are presented in Figure 18 and Figure 19. As in previous runs, DO readings exhibited a high level of variability. The accuracy of the DO probes below 1 mg/L was typically low, so the average of readings in a time period were used, to determine the approximate DO reading. As expected, in both runs, the DO concentrations were significantly lower than in previous sampling programs, where the DO exceeded 2 mg/L in the aerobic phase. Average DO readings from Figure 18 increased from about 0.5 to about 1-1.5 mg/L in the late aerobic phase. It was also noticed from Figure 18 that the maximum ORP levels were around 100 mV, which is significantly lower than the ORP observed in previous experiments. This could mean that nitrification would not be complete. This suspicion was confirmed by the data presented in Figure 19; ammonia levels at the end of the aerobic phase are still around 5 mg N/L. The pH exhibited the same pattern as found in previous experiments. Very good phosphorus removal could be observed in Figure 19. In order to differentiate between the phosphorus required for microbial growth and "luxury" phosphorus uptake, the following calculations were performed: To obtain a rough estimate of the stoichiometric nutrient requirements for growth, the relationship in equation (5) was used (Tchobanoglous et al., 2003): COHNS + 0 2 + bacteria + energy -> C5H7NO2P1/12 (5) In order to maintain a 30 d SRT and a constant TSS concentration in the reactor, it was necessary to purge 0.33 L/d. Thus, the biomass produced per day was: 0.33L/d * 3.5 g/L =1.166 42 g/d. Having 1.166 g/d / 6 cycles/d meant 194 mg of biomass was synthesized in each cycle. Using molecular weight ratios, it was found that stoichiometric growth needs are approximately 10 mg/L of carbon (thus 100 mg in the whole reactor, which translates into approximately 300 mg of methanol), 1.17 mg/L of hydrogen, 2.5 mg/L of nitrogen, 5.34 mg/L of oxygen and 0.4 mg/L of phosphorus. It thus appeared that most of the injected carbon was being used for energy (700 mg out of 1000 mg), and that a relatively small amount (0.4 mg/L) of phosphorus was required for growth. Very good phosphorus removal could be observed in this experiment, with low dissolved oxygen concentrations. In fact, the phosphorus (PO4-P) concentration more than doubled during the anaerobic release phase, increasing from 1.9 mg/L to about 4.7 mg/L. The subsequent P-uptake was also impressive, reducing the phosphorus concentration down to 0.114 mg/L at the end of the aerobic phase (Figure 19). Interestingly, the phosphorus concentration remained high throughout the first part of the aerobic phase, but started decreasing as soon as the ORP reached an inflexion point, (Point A in Figure 18) and the DO started increasing. This is thought to be caused by the additional depletion of methanol in the reactor. During the aerobic phase, carbon oxidation is suspected to have depressed the DO and redox conditions low enough for the reactor to behave like an anoxic zone, even though oxygen was being added to the system. In fact, nitrification only increased when the DO started to rise at the ORP "knee" (Point A in Figure 18). The second "knee", labelled as point B in Figure 18, is due to the end of the secondary methanol feed at time 15:15. Denitrification, however, was complete and seemed to occur very quickly; the nitrate concentrations dropped from 4 mg/L to about 0.01 mg/L, in less than four minutes. In summary, decreasing the air supply to the reactor had a very positive effect on phosphorus removal, even if nitrification was incomplete. Phosphorus removal data from Figure 19 strongly suggests that the phosphorus variation is indeed a Bio-P "phenomenon". Consequently, it was suggested that further attention should be put on microscopic examinations of stained samples, to detect polyphosphate granules and/or try to measure PHB's in these samples. 43 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 Time (abs) Figure 18: pH, DO and ORP profiles over one cycle of the methanol-fed anoxic-aerobic SBR. 4^ 4.6 Total Phosphorus/Total Nitrogen Balance Knowing that at the time of these specific experiments, the biomass concentration was 3.5 g/L, and that 100 % of that biomass was made up of cell material (it was previously found that about 95 % of TSS were VSS), one can calculate the phosphorus and nitrogen incorporated into the biomass and attempt to determine the amount of phosphorus/nitrogen accumulated in the organisms. The typical composition of bacteria cells is typically 50 % carbon, 22 % oxygen, 12 % nitrogen, 9 % hydrogen and 2 % phosphorus (Tchobanoglous et al, 2003). Applying these percentages to the biomass, this would produce a total phosphorus concentration of about 80 mg P/L and about 480 mg N/L, just to make up for the cell material. The data presented in Table 3 seems to suggest that most of the phosphorus was in a non-soluble form and that the total phosphorus concentration, (once the soluble part has been removed), was very close to the theoretical phosphorus concentration for cell mass alone. Soluble phosphorus was found to be mainly composed of orthophosphates. The total nitrogen data suggests that most of the nitrogen was also in a non-soluble form. The total nitrogen concentration seemed higher than the theoretical nitrogen concentration required for cell mass alone. The difference was about 18 %, which could be statistically significant; further testing is recommended to confirm this finding. " Table 3: TKN and TP concentration in main reactor biomass; samples 1-3 were unfiltered from the anoxic phase, 4-6 were filtrate from the anaerobic phase, while 7-9 were Sample TP (mg P/L) TKN (mg N/L) 1 105.5 585 2 88 590 3 83 595 4 3.98 10.7 5 3.98 11.3 6 3.16 - 11.4 7 81.5 575 8 83.5 555 9 83.5 565 46 4.7 Batch Tests The previous experiments have been fairly conclusive in terms of demonstrating that the phosphorus uptake and release phenomenon was indeed biological. In order to investigate a larger numbers of conditions on the system, a series of batch test experiments were also conducted. The batch tests were of 4 h duration, to mimic the behavior of the full-size reactor. Due to technical issues, pH, DO and ORP could not be monitored in real-time. The purpose of these batch tests was to first check if the batch tests were representative of real conditions and, second, to investigate the effect of various nutrient loading rates on the phosphorus removal efficiency when fed with methanol. 4.7.1 Batch Test #1 The first batch test was aimed at determining the possible impact of yeast extract on phosphorus release/uptake. To determine this uptake, different concentrations of yeast extract were added to the feed of the batch test apparatus. A typical ratio of about 133 mL of feed to about 400 mL of sludge was used in the batch test. No secondary feed was used during this experiment. The batch test lasted for 5 hours, with the first two hours under anaerobic conditions and the last three hours under aerobic conditions. Dissolved oxygen concentrations could not be measured due to technical difficulties. Aeration was determined by trial and error with a couple test runs, but the volumes of air were in the range of 50 mL/second The tested concentrations were 0, 10, 30, 70, 130 and 210 mg/L yeast extract. The results are presented in Figure 20 and 21. 47 350 Time (minutes Figure 20: NOx concentrations vs time for a batch test examining the impact of yeast extract addition 150 200 T i m e (minutes) 250 0 mg/L Yeast 10 mg/L yeast 30 mg/L Yeast 70 mg/L Yeast 130 mg/L Yeast 210 mg/L Yeast — i — i — i — i — i — 300 350 Figure 21: Orthophosphate concentrations vs time for a batch test examining the impact of yeast extract addition From Figure 20, it can be inferred that the concentration of yeast extract did not seem to have any significant impact on NO x profiles, over a 4 h period. No unusual trend could be observed from the experimental data. The NOx that was present at the beginning of the 48 experiment came from the sludge that was taken from the bioreactor. The NO x concentration decreased between time 0 and time 120. The system took a surprisingly long time to reach a low NO x concentration. On the other hand, the NOx concentration started increasing as soon as air was turned on. This suggests that the carbonaceous oxygen demand was not very large at the end of the anaerobic zone, suggesting a low concentration of reduced compounds such as methanol. The concentration of yeast extract did not seem to have any noticeable impact on phosphorus release/uptake, either, as it can be observed from Figure 21. The pattern of the curves was independent of the applied concentration; phosphorus released between time 0 and time 120, then either stabilization or a reduction of phosphate levels. No clear trends could be seen from this batch test on the impact of yeast extract concentration on NO x or orthophosphate levels. 4.7.2 Batch Test #2 This batch test was designed similar to the previous batch test, but in much lower oxic conditions than Batch Test #1. Air flows were halved compared with the air flows associated with Batch Test #1 (it was suspected that the Batch test #1 sludge had been excessively aerated prior to the experiment). All other conditions were kept identical. The results are presented in Figures 22 and 23. 0 5 0 100 150 2 0 0 2 5 0 3 0 0 Time (minutes Figure 22: NOx concentration vs time for different yeast extract concentrations in a batch test 49 1 A i r AHI " A r" - • — 0 mg/L Yeast 10 mg/L Yeast - 4 — 3 0 mg/L Yeast Anaerobic - e — 7 0 mg/L Yeast Aerobic 1 — i — i — i — i — i — i — i — i — i — i — t j I O U 11 iy/1_ i e n s i . —A— 2 1 0 mg/L Yeast — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — 0 5 0 1 0 0 150 2 0 0 2 5 0 3 0 0 Time (minutes) Figure 23: Orthophosphate concentration vs time for different yeast extract concentrations for a batch test It can be observed from Figure 22 that, since the sludge was relatively "reduced" prior to the experiment, the initial NO x concentrations were quite low and dropped rapidly to zero in less than an hour. The NO x increased rapidly following the start of aeration. Data from Figure 22 also suggests that there might be some sort of correlation between the yeast extract concentration and the NO x levels at the end of the aeration phase. In fact, yeast extract is expected to contain complex nitrogen compounds that are likely to be hydrolysed to ammonia, then nitrified. This contrasts with Batch Test #1, which found no correlation. It was also observed that the initial phosphorus levels were significantly higher than those observed on the Batch Test #1 experiment (6.0 vs 1.5 mg P/L). This is thought to be due to the near-absence of phosphorus uptake in the aeration period prior to the batch test. Marginal phosphorus release was observed for most yeast extract concentrations and a subsequent uptake in the aerobic phase; as such, the phosphate profiles look different than those observed in Batch Test#l. The variability between experimental conditions, the difficulty of collecting and storing sufficient quantities of sludge prior to an experiment and lack of reproducibility of results led us to believe that large amounts of sludge taken directly from the full-size reactor and tested right away on the batch test apparatus were needed. The main problem with this approach is that this significant sampling of sludge could compromise the quasi-equilibrium established in the full-size pilot scale reactor. As such, the batch test experiments were terminated. 50 4.8 Staining and Microscopic Examinations 4.8.1 Staining of Methanol Bio-P Sludge Polyphosphate (Neisser) as well as PHB stains were performed on one long anaerobic batch test (March 2, 2005). This experiment was aimed at trying to release all the phosphate contained in the cells and observe macroscopically (ortho-P levels) and microscopically (microscopic examinations of stained samples) the sludge. Figure 24 shows a microscopic slide of the polyphosphate stained sludge before the experiment. The deep blue particles are polyphosphate granules. Figure 24: Polyphosphate stain of a sludge sample prior to a prolonged anaerobic cycle Note: 1 mm is equal to 1pm in the final image 51 i Figure 26: Polyphosphate stain of a sludge sample after a prolonged anaerobic cycle Note: 1 mm is equal to 3um in the final image 52 Figure 27: Polyphosphate stain of a sludge sample after a prolonged anaerobic cycle (lOOOx magnification) Note: 1 mm is equal to lum in the final image From Figure 24, one can observe large egg-like structures. The structures seem to be aggregated in little clusters of several "eggs". The distribution of the polyphosphate granules seem to be mainly on the inside of the floe. From Figure 25, it can be observed that the egg-shaped structures are no longer visible. No dark blue spots can be observed on the photograph. This would be consistent with the behaviour of PAOs under extended anaerobic conditions; much of the phosphate would be released to the liquid. Also observed were some "empty shells", that resemble bacillus, in Figure 25. These may be the same "egg-shaped" structures that were observed in Figure 24, but empty of polyphosphates. The Figures 26 and 27 were taken on the same visual frame, but with different objectives. Figure 26 was taken with a normal lOOOx objective, while Figure 27 was taken at the same magnification, but with a phase contrast objective, allowing us to see the "inside" of the structures. Despite the fact that the structures look empty to the microscope, it is believed that 53 they may have contained some polyphosphates. Furthermore, the brown empty shells of Figure 25 may be the "empty" state of the egg-like structures. In order to obtain a more precise size estimate of the structures that were observed under the microscope, a stage micrometer was employed. The stage micrometer consists of a microscopic slide on which inscribed lines are exactly 0.01 mm (10 pm) apart. A picture of the micrometer at the appropriate scale is presented in Figure 28. Figure 28: Micrometer viewed at a lOOOx magnification Note : 1 division = 10 pm With this new tool, it was estimated that the egg-like structures were around 1-2 pm in diameter, while some of the clusters were up to 8 pm in diameter. The pale brown structures that were observed in Figure 25 were between 0.5 and 1 pm in diameter. The diameters of the individual structures would be appropriate for bacterial cells, but would be too small for other microorganism classes, such as protozoa (20-300 pm) or yeast (30-50pm). 54 4.8.2 Staining of UBC Pilot Plant's Sludge In order to compare our biomass quality with an existing BNR plant sludge, (the UBC Pilot Sewage Treatment Plant) a slide stained for polyphosphates is presented in Figure 29. It was noticed that the sludge is completely purple, and that the density of the colour is much higher. It was also noticed that the individual cells were at least 2-3 times smaller than the egg-like structures that were observed in the methanol Bio-P sludge. The clusters observed in Figure 29 were also much more compact, and almost look like "grapes". Figure 29: Polyphosphate stain of UBC Pilot Sewage Treatment Plant sludge under a 1000 x microscopic magnification Note: 1 mm is equal to lum in the final image Most of the PHB stains that were performed on the sludge samples were negative to PHB or were not clear. It is very interesting to note that a lot of the PHBs do not seem to be aggregated in clusters as the black dots that were observed for the polyphosphate staining, but 55 seem to be fairly diffuse. On the other hand, in some floes, the PHB slides do clearly exhibit a cluster-like pattern and could potentially match the polyphosphate pattern. As a consequence, the analysis of PHB staining is not included in this discussion. 4.9 Investigation of The Effect of pH on Bio-P Following previous attempts to run batch tests, it was determined that experimental conditions in batch tests were too difficult to control, to obtain reliable and reproducible data. Moreover, with the batch test apparatus, it was impossible to perform several "cycles" back to back, thus proving difficult to have "reference conditions". As such, further experimentation reverted back to the main reactor The purpose of this experiment was to determine if the phosphorus release/uptake that was being observed could be caused by a pH effect. If the apparent phosphorus removal mechanism was a chemical reaction/precipitation, it is very likely that it would be pH dependent. The reactor was followed over three cycles - the first cycle as a reference, and two experimental cycles. The results are presented in Figures 30, 31, 32. In the experimental cycles, the pH was maintained fairly constant between 8.6 and 9.0, with dilute NaOH. As demonstrated in Figure 30, the first cycle exhibited the typical pH drop, from around 9.0 to about 7.2. The two subsequent cycles showed that the pH stabilization strategy worked well; the pH seemed to remain between 8.7 and 9.0. The ORP profiles exhibited a very interesting pattern; they remained flat throughout most of the aerobic phase, then suddenly increased to levels around 100 mV. The ORP levels with pH control were shifted downwards by roughly 75 mV, compared to the control cycle. At least one study (Okouchi S. et al., 2002) has shown a relationship between pH and ORP. This effect, which was similar between the cycle #2 and cycle #3, was possibly due to the pH effect. The DO levels remained very low (< 0.5 mg/L) throughout the experimentation period, and only Cycle #1 and #2 showed signs of nitrification (measurable concentrations of NO x in the reactor). The orthophosphate levels for the first cycle exhibited only a very small release during the anaerobic phase, followed by a small uptake during the aerobic phase. The orthophosphate profiles for the second and third cycles showed steadily increasing phosphate levels, without any drop during the aerobic phase. This increased the orthophosphate concentration to 6.0 mg/L, from roughly 3.0 mg/L, at the beginning of the first cycle. This would suggest that "a stabilization of pH" could be detrimental to phosphorus uptake during the aerobic phase. 56 Furthermore, the ammonia removal efficiency decreased significantly with time; the final ammonia concentration for the first cycle was below 2 mg/L, while it was roughly 4.0 and 6.0 mg/L for the second and third cycles, respectively. This was attributed to poor nitrification rates at a higher pH, with the optimal pH in the 7.5 to 8.0 range (Tchobanoglous et al., 2003). It is interesting to observe that ammonia levels decreased without creating significant amounts of nitrates, except for the first cycle. Even allowing for bacterial assimilation and some ammonia stripping, this result suggested the probability of simultaneous nitrification and denitrification, since the ammonia reduction is in the order of 4.5 mg N/L. It could also suggest that the rate of nitrification was outpaced by the rate of denitrification, where the pH effect might have had less effect on the denitrifiers (heterotrophs, primarily). Formaldehyde seemed to be present in this reactor, with initial levels around 4-5 mg/L in the beginning of the anaerobic phase and a peak at the end of the anaerobic phase, between 6.5 and 9 mg/L. Methanol is believed to be metabolized into formaldehyde by the alcohol dehydrogenase. This formaldehyde is subsequently thought to be metabolized into formate and other metabolites. (Korotkova et al, 2001) However, the observed formaldehyde levels increased with increased phosphorus levels, and seemed to correlate with the amount of oxygen present; it accumulated under anaerobic conditions and was degraded under aerobic conditions. Methanol seemed to be consumed at roughly the same rate it was added in the secondary feed, and no significant methanol accumulation occurred throughout the three sequential cycles. Finally, as was observed in previous experiments, VFAs were not present in measurable concentrations throughout the three cycles. Glycogen, formaldehyde and orthophosphate concentration profiles are presented in Figure 32. Formaldehyde and glycogen profiles do not seem to be linked to each other or even linked to the orthophosphate concentration. However, caution has to be exercised interpreting this observation, since this experiment was under pH-controlled conditions, so the data might not be representative of typical cycles. It is very interesting to note that a lot of the PHBs do not seem to be aggregated in clusters, as the black dots observed for the polyphosphate staining, but seem to be fairly diffuse. On the other hand, in some floes, the PHB slides do clearly exhibit a cluster-like pattern and could potentially match the polyphosphate pattern. Follow-up work proposes to use the "three cycles back to back" type of experiment, as proposed by Simm (2003) to collect data and work on the following topics: impact of secondary feed (with/without) on the observed bio-P; impact of a stable ORP on the bio-P (if it can be 57 stabilized), impact of inhibited nitrification (either an inhibitor or little/no ammonia); impact of a different substrate on bio-P: e.g. first try different doses of acetate, then try ethanol/glucose; and study maximum phosphorus uptake/release over long, uninterrupted periods of time. It would also be interesting to observe formaldehyde levels to see if they ever reach potentially bacteriocidal levels, over long anaerobic phases. 58 Time (abs) Figure 30: pH, DO and ORP profiles over tree cycles of the methanol-fed anoxic-aerobic SBR. 0 "vy — 8:00 9:00 10:00 11:00 12:00 N O x (mg N/L) - * g O - P 0 4 (mg P / L ^ NH4 (mg N/L - V F A [acet ic] mg/L a — F o r m a l d e h y d s (mg/L) A — Inject. M e O H (g) \ — M e O H measL r e d (g) 3.50 ^ 3.00 2.50 3 2.00 1.50 1.00 0.50 V 0.00 I o 0) T3 0) i — 3 ro a> E -6 a> < 13:00 14:00 15:00 Time (abs) 16:00 17:00 18:00 19:00 20:00 Figure 31: NOx, orthophosphates, ammonia, VFA and methanol concentration profiles as well as the cumulative methanol addition profile, formaldehyde concentration over three cycles of the methanol-fed anoxic-aerobic SBR. 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 Time (abs) Figure 32: Formaldehyde, 0-PO4 and glycogen profiles over three cycles of the methanol-fed anoxic-aerobic SBR. 4.10 Effect of Nitrification and Denitrification on P Removal In order to study the possible effect of the nitrification and denitrification on phosphorus removal, it was hypothesized that the only viable way of stopping nitrification, and therefore denitrification, was to reduce the amount of alkalinity present in the reactor. Thus, the composition of the primary feed was modified slightly, with the elimination of the sodium carbonate. The pH of the primary feed without sodium carbonate was found to be fairly low - in the order of 6.2, and was adjusted to 8.0. The first cycle of the three cycles presented in Figure 33 and Figure 34 was a reference cycle with an unmodified feed, while the two others used the low-alkalinity feed. It was surprising to see that little change in the ORP, DO and pH was observed from Figures 33 and 34. Throughout the whole cycle, the ORP dropped to very low levels and barely rose above the 0 mV mark, while the DO remained between 0 and 2 mg/L. The pH readings remained relatively flat in cycles 2 and 3 compared to the reference cycle. The increase in ammonia levels in the anaerobic phase was unexpected and remains unexplained. However, it was noticed that the final levels of ammonia were around 12 mg N/L, high enough to declare nitrification as being incomplete. During the reference cycle, the ammonia removal was roughly 4 mg N/L, and in the experimental cycles, this removal was found to be in excess of 7 mg N/L, indicating a higher nitrogen removal during the experimental cycles despite nitrification inhibition. Very low levels of NO x were observed, even during the aerobic phase. The NO x levels only started rising from the point at which the methanol concentrations dropped below 2 mg/L. The orthophosphate data suggests that phosphorus release and uptake occurred in all three cycles, but it was noticed that the last two cycles had final orthophosphate concentrations below 0.2 mg P/L. Compared to the feed, which added about 2 mg P/L to the reactor, the effluent had a very low phosphorus concentration, with a removal efficiency of about 90 %. This suggests that nitrification/denitrification could be independent from the observed phosphorus uptake and release phenomenon. It was expected that by inhibiting nitrification and by keeping all other parameters constant, the oxygen demand exerted by nitrification would decrease, and thus the amount of oxygen available to the biomass would increase. This should have had the effect of insuring the rapid and complete oxidation of methanol in the aerobic phase. However, the opposite phenomenon occurred; in the third phase, there was a very significant accumulation of methanol, 62 indicating a poor methanol uptake. This phenomenon would probably need to be investigated further by inhibiting nitrification for a period long enough to rule out any "shock" effect, but not long enough to see a shift in the bacterial population. However, this is another supporting observation for the "dual substrate theory", meaning that nitrifiers could use both ammonia and methanol. 63 8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 17:00 17:30 18:00 18:30 19:00 19:30 20:0 Time (abs) Figure 33: pH, DO and ORP profiles over three cycles of the methanol-fed anoxic-aerobic SBR. 4 -8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 17:00 17:30 18:00 18:30 19:00 19:30 20:00 Time (abs) Figure 34: Formaldehyde, methanol, ammonia, nitrate and o-P04 profiles over three cycles of the methanol-fed anoxic-aerobic SBR. 4.11 Ethanol as a Replacement Substrate for Methanol In order to determine if the organisms that were present in these reactors were specific in terms of substrate that they can use for energy and cell growth, ethanol was substituted as a substrate, using the same carbon ratio. Ethanol has characteristics, as well as a molecular structure, very similar to methanol. However, the known metabolic paths are quite distinct. Ethanol goes through acetic acid to acetyl-CoA while methanol needs to be fixed to make it a bigger carbon molecule so the processing should be different, unless we are faced with a novel, unique pathway There have been reports that ethanol, at low doses, might have even faster degradation kinetics than methanol (Seong-Wook et al., 2006). Ethanol being a two-carbon compound, also raised the question of whether this carbon would be processed by the same organisms, using a similar pathway as methanol. The results of the experiment are presented in Figures 35 and 36. The first cycle was a reference cycle and was fed methanol. The last two cycles (#2 and #3) were fed ethanol as a sole carbon source. Acclimation of the biomass to ethanol was not undertaken. It can be observed from Figure 35 that, the "lag" period between the time at which the air is turned on and the steep rise in ORP levels is greatly shortened in the ethanol-fed cycles. The "lag" period was found to be nearly 120 minutes in the reference cycle, and was shortened to less than 60 minutes in the ethanol-fed cycles. This would suggest a very fast uptake of the substrate by the micro-organisms, which is consistent with previous reports of faster kinetics for ethanol than methanol (Seong-Wook et al., 2006). The nitrification process seemed to be impacted by the injection of ethanol, and this is reflected in Figure 36 where, for cycles #2 and #3, there was only a very small drop in ammonia levels, despite a significant increase in dissolved oxygen concentrations, indicating the end of the carbon oxidation phase. Ethanol seemed to remain present in measurable concentrations in the bioreactor, throughout the two experimental cycles. Phosphorus release seemed to be significantly impacted by the addition of ethanol. Whether the phenomenon was due to a disturbance to the biomass or due to the fact that the organisms readily used ethanol as a substrate, the phosphorus "release" doubled when comparing the reference cycle and the two experimental cycles. Phosphorus uptake in the experimental cycles was incomplete and phosphorus levels in the effluent were higher in the experimental cycles than in the reference cycle. However, in methanol, the P uptake was only 2-3 mg P/L in 1.5 hour while in ethanol, the P uptake was 5 or 6 66 mg P/L in 1.5 hour. This would suggest that the uptake was more effective with ethanol by a factor of 2x. The ethanol phosphorus uptake and release patterns looked very similar to acetate uptake and release patterns. 67 8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 17:00 17:30 18:00 18:30 19:00 19:30 20:0 Time (abs) Figure 35: pH, DO and ORP profiles over three cycles of an ethanol-fed anoxic-aerobic SBR. oc • N O x (mg H/L) Cycle # 90.00 80.00 70.00 60.00 50.00 f x o 40.00 UJ x " o 30.00 1 20.00 1 n n 0 8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 17:00 17:30 18:00 18:30 19:00 19:30 20:00 0 Time (abs) Figure 36: Formaldehyde, 0-P04 profiles over three cycles of the ethanol-fed anoxic-aerobic SBR. OS 4.12 Acetate as a Replacement Substrate for Methanol Throughout this project, the lab-scale bioreactors (Figure 37) were never fed with volatile fatty acids. Despite numerous experiments where VFA levels were measured for in the reactor, there was never "proof of the presence of acetate. Whether or not the reactor would be able to tolerate acetate remained a question. Typically, acetate is one of the substrates of choices and is widely recognized as the prime substrate in Bio-P reactors. However, acetate is a two-carbon compound, and unlike ethanol, has characteristics that are significantly different from methanol. Thus methanol was replaced in the primary and secondary feed of the reactor, on a 1 to 1 carbon ratio. The results of the experiments are presented in Figure 38 and 39. The initial control cycle with methanol was fairly standard - a pronounced drop in pH, drop of the ORP levels to about -250 mV, and a fast rise as soon as the carbon oxidation phase was over. The dissolved oxygen profile was fairly unusual - a low DO concentration until the end of the aeration cycle, then a rapid increase of the DO to about 2 mg/L. The experimental cycles with acetate were completely different than anything that had been witnessed up to this date. The ORP profile showed a steep decline in the anaerobic phase, and a steep increase in redox levels as soon as the air was turned on at the beginning of the aerated phase. The third phase was even more pronounced; ORP levels only dropped to about -125 mV, compared to the usual -250 mV, and started increasing as soon as the air was turned on. The dissolved oxygen levels began to rise almost at the same point as the ORP signal. The DO levels remained around 1 mg/L for most of the aerated cycle, and reached as high as 4 mg/L. The pH signal also started leveling off in cycles 2 and 3, compared to cycle #1. As it can be observed in Figure 39, the NO x levels remained fairly low throughout the first cycle, and increased only at the very end of the aerobic phase. For cycles #2 and 3, NO x levels started rising as soon as the air was turned on. According to the NO x data, nearly showing a straight line, it was assumed that the nitrification rate was constant throughout the two cycles. The data from the second and third cycles suggest that nitrification was functional, but that denitrification could have been affected. Acetate was identified as a poor denitrification substrate (Seong-Wook et al., 2006), and since it's a two-carbon compound, could have primarily affected heterotroph organisms such as denitrifiers. It is impossible to confirm the total amount of nitrification in the three cycles since, due to unforeseen circumstances, ammonia concentration data was not available for this experiment. 70 The pattern of phosphate release and uptake observed in the reference cycle of Figure 39 was fairly common, but during the cycles #2 and #3, there was a very significant increase in phosphorus release during the anaerobic cycle, followed by a slower phosphorus uptake in the aerobic phase. The phosphorus uptake was approximately 4 mg/L, which was poorer than the ethanol system. This suggests that the phosphorus release and uptake were likely limited because the "anaerobic zone" was too short. The acetate concentration profiles shown on Figure 39 show a fairly fast acetate uptake rate since there is no significant accumulation of acetate in the reactor. The acetate uptake rate seemed higher in the aerobic phase than during the anaerobic phase. This experiment with acetate was unusual on several fronts; a significant disturbance of the ORP, pH and DO profiles, coupled with an enhancement of nitrification and phosphorus release in the anaerobic phase, was noticed. The disturbance was so significant that the reactor started foaming at the beginning of the aerobic period of cycle #2. Moreover, immediately after the cycle #2, the sludge stopped settling. The biomass remained in suspension and the formation of a "hazy blanket" was observed in the reactor. This blanket remained in the reactor for several months, following the acetate experiment. It is still not clear to this researcher if acetate really helped bio-P removal in this system or if the biomass reacted to a sudden change in environmental conditions. However, it was evident that acetate provided a sufficiently significant shock to the reactor, that the "normal" ORP, pH and DO patterns never returned. On the left side of Figure 37, a healthy biomass can be observed in the settling phase -what can be observed is a well-defined sludge blanket, a clear supernatant and a sludge having a brown-red tint. This previously healthy reactor deteriorated into what is shown on the right side of Figure 37. It was noticed that the sludge blanket is not well defined, and that the sludge concentration is lower than the reactor on the left. It was also noticed that the supernatant was very turbid and that there were more sludge deposits on the walls of the reactor. The reactor at the right of Figure 37 remained turbid for months. The reactor kept losing solids, due to the poor settlability, despite several attempts to "rescue" the reactor by subjecting it to a range of pH, carbon and nitrogen regimes, it never seemed to "recover". Given the texture of the liquid coming out of the reactor, it was suspected that the hazy material might be composed of exocellular polymer, but this was not confirmed in this study. The only way to implement a return of a "normal" biomass was to empty the reactor, disinfect it and re-seed it with refrigerated sludge. 71 72 8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 17:00 17:30 18:00 18:30 19:00 19:30 20:0 Time (abs) 0 Figure 38: pH, DO and ORP profiles over thee cycles of the acetate-fed anoxic-aerobic SBR. —J 8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 16:00 16:30 17:00 17:30 18:00 18:30 19:00 19:30 20:00 Time (abs) Figure 39: Acetate, NOx, 0-P04 and methanol profiles over three cycles of the acetate-fed anoxic-aerobic SBR. 4.13 Uniformity of the Biomass of the Two Reactors: Despite the fact that, at several points in time, two reactors were operating simultaneously, they always exhibited a significant difference either in pH, DO or ORP profiles or in an apparent difference in biomass concentration or type. It was found that the only way that both reactors could be used for the same experiment would be to empty them, mix their biomass contents and redistribute the sludge between the two. Although this was an "interesting observation", it was also time consuming and frustrating. 4.14 Characterization of Organisms Present in the Bio-Reactors Over the period of time spanning these experiments, microscopic examinations of various samples allowed us to observe different various types of organisms. All observations were made at least 6 months after the start of the methanol-only feed. Some photo examples of the observed organisms are shown in Figure 40 for the reader's interest only. Figure 40: Various organisms that were observed under optical microscopy throughout this study. All images taken at a magnification of lOOOx. Ciliate Note: 1 mm is equal to 3pm in the final image 75 Polyphosphate accumulating organisms Note: 1 mm is equal to 3um in the final image Phosphate-accumulating filamentous bacteria Note: 1 mm is equal to lum in the final image > 4 76 4.15 Molecular Mechanisms of Methanol Conversion This initial study would be incomplete without looking at the potential molecular mechanisms explaining the enhanced phosphorus uptake and release that were observed in several of the runs. In fact, according to conventional wisdom, biological phosphorus removal should not occur with methanol as a sole substrate. The data was fully examined but there was no evidence found of the presence of acetate in any reactor. It was thus hypothesized that acetate could be produced in-situ, in minute concentrations, that would be taken up as fast as it would be produced. By carefully studying the established molecular mechanisms happening inside these microorganisms, it was found that acetate could "theoretically" be produced from Acetyl-CoA. Acetyl-CoA is an essential component of carbon metabolism. This metabolite plays a vital role of an "intermediate" between various anabolic and catabolic pathways, such as the synthesis of fatty acids, cholesterol (not present in bacteria) and even amino acids (Smith, 2007). The reactions describing the formation of acetate can be shown by Equations (6, 7 and 8) (Van Hellemond, 1998): Acetyl-CoA + ADP + Pi <r -» acetate + ATP + CoA (6) 77 Acetyl-CoA + Pi <- -> acetyl phosphate + CoA (7) Acetyl phosphate + ADP <r -> acetate + ATP (8) From Figure 41, it can be seen that Acetyl-CoA plays a significant role in the metabolism that leads to the formation of acetate. However, the data presented is only valid for glucose substrate, something that had never been tested in these experiments. Thus, a way of tying methanol into the metabolic picture had to be found. What was subsequently found was more than a molecular pathway between methanol and the formation of acetate. In fact, it had been previously reported in the literature that there was also an indirect pathway between methanol and PHBs. The full metabolic pathway is presented in Figure 42. glucose • PEP CO, Figure 41: Biochemical pathways : 1- Phosphotransferase system, 2- Pyruvate kinase, 3-Pyruvate dehydrogenase, 4- phosphoenol pyruvate carboxylase, 5- pyruvate oxidase, 6-Lactate dehydrogenase, 7- Acetyl CoA synthase, 8- Phosphotransacetylase, 9- acetate kinase, 10- isocitrate lyase, 11- Malate synthase Source: Adapted from Tomar et al., 2003 78 Methanol Formaldehyde I Methylene H4F Assimilation Pec, MeaA GlyA NADP -NADPH Butyryl-CoA 1 adaA Crotonyl-CoA 2 crsA. tas& (L)£-HydrdxybutyryK*>A NAD NADH +-Ace^yt-CoA 6 phaA Succinate 4-Succlnyl-CoA — Acetoacetyl-CoA 10 - NADPH • NADP A c e t y l p h o s p h a t e Aceto acetate t RbsB (D)£-Hydroxybutyryl-CoA-NADH 4 s| m NAD 9 PHB-cycle A c e t a t e p-Hydroxybutyrate HjO Figure 42: Metabolic pathway between methanol, formaldehyde, acetate and PHB. Source : Adapted from Korotkova et ah, 2001. Korotkova clearly shows a relationship between methanol and PHB. There is no clear relationship to polyphosphate in her diagram. In the normal acetate BEPR the polyphosphate is incorporated in the acetylphosphate that is converted to acetyl-CoA. In the serine cycle the methanol is never directly phosphorylated or converted to methanol-CoA so the involvement of phosphate must be less direct and the acetyl CoA arises from malyl-CoA rather than malyl-phosphate. However, the serine cycle shown in Fig 42 is missing several steps. In one step the compound glyoxylate is converted to glycine that is converted to serine by methylene H4F made from methanol. This links methanol to the pathway. The serine is then converted to hydroxypyruvate which is reduced to glycerol by NADPH. This might link PHB degradation to the pathway because PHB degradation makes NADPH. However it is important to recognize that this pathway uses NADPH and competes with the PHB synthetic pathway for NADH. This means that under conditions where serine is converted to hydroxypyruvate you would likely see 79 PHB degradation rather than synthesis. The glycerol is then phosphorylated to 2-phosphoglycerate with ADP. That phosphoglycerate is then converted to phosphoenolpyruvate which is dephosphorylated and combined with carbon dioxide to make oxalacetate which is converted to malate. The malate is then converted to malyl-CoA by using ATP. The malyl-CoA then hydrolyses into acetyl-CoA and glyoxylate. The glyoxylate goes back into the start of the cycle and the acetyl-CoA is used for synthesis of cell material and PHB. In this pathway the polyphosphate could be used to make ATP for the conversion of glycerol to 2-phosphoglycerate or to make ATP to add the malate to malyl-CoA. In this way the polyphosphate might be working as an energy storage that allows the accumulation of methanol as PHB in anaerobic conditions but the connections are less direct than in the acetate system and the NADH balances are different. It is believed that this metabolic pathway could "adequately" explain the biological phosphorus removal mechanisms observed under low DO conditions in the presence of methanol. This possible metabolic pathway would lead to the formation of acetate, as well as PHB. The formation of these two compounds could return methanol bio-P in the realm of known science, since these two have been extensively studied in the past. The formation of PHB is linked to the "P" storage since PHB is a carbon storage molecule. Methanol could thus be metabolized into acetate and further incorporated in the form of PHBs or the PHBs could be synthesized directly from methanol. Further research is planned along this line of thinking, as noted in the following. In order to confirm these findings, methanol bio-P sludge will be subjected to a number of tests to validate this hypothesis. These tests could take the form of tracking radio-labelled methanol inside the microorganisms, inhibiting the degradation of some of the intermediates, to be able to measure their concentrations. This investigation will go even further by trying to detect the presence of the "genes coding", for some specific proteins such as acetyl phosphatase, acetate kinase, etc. However, these enzymes are fairly common, so their presence would not prove that polyphosphate was used to make ATP during the anaerobic phase The microbial genes associated with the proposed phosphorus removal pathway should be identified and tracked in laboratory-scale reactors. This extended research program will form the basis of a doctoral study between the Department of Microbiology and the Department of Civil Engineering, at UBC and the Department of Civil and Environmental Engineering of McGill University. 80 CHAPTER V - SUMMARY AND CONCLUSIONS In conclusion, this study began with very little information regarding an elusive phosphorus removal mechanism that was previously reported, when an SBR system was fed with methanol. A laboratory-scale bioreactor was fed with synthetic sewage for several months, in order to try and replicate these findings. Not only was the reported phosphorus removal phenomenon replicated in several runs, but the operating conditions that appear to make it happen were identified. The phosphorus removal that occurred was quantified, and microscopic examinations were performed to pinpoint the potential organisms responsible for the phosphorus removal; also measured were the concentrations of some chemical intermediates in the pathway that was most likely responsible for initiating the phosphorus removal. It was found that an alternation of anaerobic and aerobic conditions was essential to trigger the phosphorus removal mechanism. It was also found that, even for acclimated organisms, a methanol shock load could occur, at fairly low concentrations. The shock load in these experiments seemed to have stopped biomass growth, nitrification, denitrification and phosphorus removal. Separating the methanol dose into a primary methanol feed and a secondary, continuous methanol feed was found to be the key to the overall health of the reactor. It was also found that the oxidation-reduction potential (ORP) was the best real-time indicator of the process health, indicating clearly when the reactor was in a denitrifying, nitrifying or even over-oxidizing phase. Over-oxidizing conditions were detrimental to the removal of phosphorus. When the reactor had ORP levels higher than approximately 100 mV, the conditions were considered to be "over-oxidizing". Under microscopic examination, it was discovered that the organisms deemed responsible for phosphorus removal were relatively large in size (~ 1-2 um) and tended to be found in clusters. It was not clear whether or not they formed PHBs. Also, no detectable concentrations of VFAs were measured in the reactor; several experiments gave us false positives for acetate, but were quickly disproven with further analysis. Also confirmed was the presence of formaldehyde in the reactor, which was demonstrated to be an intermediate in the degradation of methanol, eventually converting it into biomass or carbon dioxide. The batch tests that were performed were inconclusive, due to multiple technical problems; these included sample volume drawing down the fluid level in the flasks, ensuring an even distribution of air into the various flasks due to different hydraulic heads, the problem 81 associated with providing a uniform mixing to all the flasks at the same time, the problem associated with distributing evenly the flow between 6 flasks connected to the same peristaltic pump and the inability to monitor any parameter in-line such as pH, DO or ORP. The strategy of analyzing three full cycles back-to-back worked very well, since it was possible to use the first cycle as a reference, then change the conditions and observe the results over the next two cycles. This method, however, proved to be time-intensive, since three cycles back to back involve 12 hours of sampling per experiment. It was found that feeding the microorganisms with ethanol still resulted in enhanced removal of phosphorus. In fact, this feeding strategy with ethanol was found to boost phosphorus release, even compared to methanol, a result which was highly unexpected. Results collected from these experiments could not be used to propose a mechanism to explain this phenomenon. The ethanol effect might in fact relate to the balancing of the NADH on the cycles. It is possible that the enhanced phosphorus and uptake might be due to some sort of shock load or stress, but this phenomenon will require more thorough investigation. It was also found that feeding the bacterial mass with acetate upset the reactor in a very significant fashion, to the point that the reactor had to be emptied and disinfected before being re-seeded with a refrigerated sample. In fact, after being exposed to acetate, the biomass started foaming and completely stopped settling; also observed was the formation of a hazy compound in the reactor that appeared to be exocellular polymers. There was no conclusive evidence of the link between the methanol bio-P removal and the nitrogen cycle. It was found, however, that low oxic conditions - often resulting in incomplete nitrification - favored the methanol-induced biological phosphorus removal process. There has been some speculation that a simultaneous nitrification and denitrification phenomenon might be connected to phosphorus removal, but this avenue remains to be investigated, in future experiments. Finally, several classes of organisms were identified with the help of optical microscopy. These classes range from nematodes, filamentous bacteria to ciliates. A genetic "fingerprint" of the various organisms present in the reactor might help to identify the specific species of these organisms. 82 5.1 For Future Investigation In terms of avenues for future research, it is believed that molecular probes would be very interesting to employ, to confirm that the large egg-shaped organisms that were observed throughout the microscopic examination, are indeed bacteria. It is believed that, due to their size, some of these organisms could be either protozoa or large bacteria. It is also believed that studying detailed molecular mechanisms of the methanol biological phosphorus removal could be interesting. This could mean investigating the formation of PHB and acetate from the methanol building block. For example, some specific enzymes could be "blocked", in order to observe the accumulation of an intermediate that would confirm or disprove the presence of the specific metabolic pathway. Tracking of "intermediates", using radio-labelled methanol, could also help to determine the fate of the carbon in the cell. RNA probes could also be used to detect the presence or absence of specific genes coding, for targeted proteins synthetizing or breaking down acetate. Other possibilities should also be looked at; the possibility of phosphorus storage in the glycocalyx of some bacteria (adsorption), shifts in bacterial population makeup by extending the anaerobic cycle, tracking N 2 0 to determine if SND is, indeed, taking place. There could also be a study on the origin and makeup of oxygen demand. The role of nitrifiers could also be assessed using nitrapyrin, finding a way of designing the batch reactors to work to simplify the testing in a controlled way. The role of the serine cycle in the reactor could also be investigated. An investigation could also be done on the role of other CI sources such as methane, methylamine, formaldehyde, formic acid to try to understand the organisms that might be involved. The significant pH changes should also be investigated to determine if it is caused by the accumulation of formate or other intermediates in the process. An investigation on the best secondary feed regime could also be performed to determine if secondary feed can be abolished in an established biomass. A form of "ORP conditioning" could also be interesting to see if it could trigger phosphorus storage and release, without changing reactor concentrations of nutrients. 83 REFERENCES Arvin, E. 1983. Observations Supporting Phosphate Removal by Biologically Mediated Chemical Precipitation: A Review. Wat. Sci. Tech., 15(3/4): 43-63. Bennion, H., Juggins, S. and Anderson, N.J. 1996. Predicting Epilimnetic Phosphorus Concentrations Using an Improved Diatom-Based Transfer Function and Its Application to Lake Eutrophication Management. Environ. Sci. Technol. 30(6): 2004-2007. Bilanovic D., Battistoni P., Cecchi F., Pavan P. and Mata-Alvarez J. 1999. Denitrification Under High Nitrate Concentration and Alternating Anoxic Conditions. Wat. Res. 33(15): 3311-3320. Bitton G, 1994. Wastewater Microbiology , Wiley-Liss, New York. Chistoserdova, L. Vorholt, J.A. and Lidstrom, M.E. 2005. A Genomic View of Methane Oxidation by Aerobic Bacteria and Anaerobic Archaea. Genome Biol. 6:208. Cho E.S., Ahn K.H. and Molof A.H. 2004. Improvement of Denitrification by Denitrifying Phosphorus Removing Bacteria Using Sequentially Combined Carbon. Wat. Sci. Tech.. 50(8), pp 33-40. Clesceri L. S., Greenburg A. E. and Eaton A. D. (1998) Standard Methods for the Examination of Water and Wastewater. 20th ed. American Public Health Association, Washington D.C. Crowther, G. 2004. Fundamentals of General Microbiology III. Course material for Micro 412. University of Washington. http://facultv.washington.edu/crowther/Teaching/Guest/methylotroph handout.doc. Accessed on April 15, 2007 Eikelboom, D.H. 2000. Process Control of Activated Sludge Plants by Microscopic Investigation. IWA Publishing, London. 150 p. Glindemann, D., and Bergmann, A. 1995. Spontaneous Emission of Phosphane from Animal Slurry Treatment Processing, Zbl. Hyg. Umweltmed. 198, 49. Glindemann, D., and Bergmann, A. 1996. Phosphine in the Lower Terrestrial Troposphere, Naturwissenschaften. 83, 131, 1996. Glindemann, D., Stottmeister, U., and Bergmann, A. 1996. Free Phosphine from the Anaerobic Biosphere. Environ. Sci. Pollut. Res. 3(1), 17, 1996a. Grady, C.P.L, Daigger, G.T., and Lim, H.C. 1999. Biological Wastewater Treatment. Marcel Dekker. 1092 p. Greenburg, A. E., Klein, G. & Kaufman, W. J. 1955 Effect of Phosphorus on the Activated Sludge Process. Sewage Ind. Wastes, 27. 84 Hies, P. and Mavinic, D.S. 2001. The Effect of Decreased Ambient Temperature on the Biological Nitrification and Denitrification of a High Ammonia Landfill Leachate. Wat. Res. 35(8): 2065-2072. Jenkins, D., Richard, M.G. and Daigger, G.T., 1993. Manual on the Causes and Control of Activated Sludge Bulking and Foaming. 2 n d Ed. Lewis Publishers. Chelsea, MI. Khan, F.A. and Ansari, A.A., 2005. Eutrophication: An Ecological Vision. Bot. Rev. 71(4), pp 449-482 Korotkova, N. and Lidstrom, M.E. 2001. Connection Between Poly-B-Hydroxybutyrate Biosynthesis and Growth on CI and C2 Compounds in the Methylotroph Methylobacterium extorquens AMI. J. Bacteriol. 183(3), 1038-1046. Lemaire, R., Meyer R., Taske, A., Crocettia, G.R., Kellera, J., Yuan Z. 2006. Identifying Causes for N?Q Accumulation in a Lab-scale Sequencing Batch Reactor Performing Simultaneous Nitrification. Denitrification and Phosphorus Removal. J. Biotechnol. 122(1), pp 62-72. Louzeiro, N.R., Mavinic, D.S., Oldham, W.K., Meisen A. and Gardner I. S. 2002. Methanol-induced Biological Nutrient Removal Kinetics in a Full-scale Sequencing Batch Reactor. Wat. Res. 36(11): 2721-2732. Marco, P. 2004. Methylotrophy Versus Heterotrophy: a Misconception. Microbiol. 150, 1606-1607. Morris, D.L. 1948. Quantitative Determination of Carbohydrates with Dreywood's Anthrone Reagent. Sci. 107(2775). 254-255. Morse, G.K., Lester, J.N. and Perry, R. 1993. The Economic and Environment Impact of Phosphorus Removal from Wastewater in the European Community. Selper Publications, London. Morton, S. C. and Edwards, M. 2005. Reduced Phosphorus Compounds in the Environment. Crit. Rev. Sci. Tech. 35: 333-364. Murray, R.G.E. 1981. Manual of Methods for General Bacteriology. ASM. Washington, DC. Murrell, J.C., McDonald, I.R., Bourne, D.G., 1998. Molecular Methods for the Study of Methanotroph Ecology. FEMS Microbiol. Ecol. 27'. 103-114. Naidoo, M.V. and Buckley, C.A. 1999. Municipal Wastewater Characterization: Application of Denitrification Batch Tests. Final Report to the Water Research Commission for Project No. K5/820. NIOSH. 1994. Manual of Analytical Methods. Method 3500 Fourth Edition. 85 Okouchi, S., Suzuki, M., Sugano, K., Kagamimori, S. and Ikeda, S. 2002. Water Desirable for the Human Body in Terms of Oxidation-Reduction Potential (ORP) to pH Relationship. J. Food Sci. 67(5). 1594-1598. Russell, R.A., Holden, P.J., Garvey, C.J., Wilde, K.L., Hammerton, K.M., Foster, L.J. 2006. Investigation of the phase morphology of bacterial PHA inclusion bodies by contrast variation SANS. Physica B.{\) pp859-861. Seong-Wook, O., Geonka, K. and Young, K. 2006. Determination of Electron Donors by Comparing Reaction Rates for In Situ Bioremediation of Nitrate-Contaminated Groundwater. J. Env. Sci. Health [A]. 41(10). 2359-2372. Shapiro, J., Levin, G.V., and Zea, H. G. 1967. Anoxically Induced Release of Phosphate in Wastewater Treatment. J. Water Pollut. Control Fed. 39: 1810-1818. Shiskowski, D. 2004. Investigation of Nitrous Oxide as a Nitrification Monitoring and Aeration System Control Parameter in Sequencing Batch Reactor Wastewater Treatment Systems. Thesis. University of British Columbia, 286 p. Smith, K. 2007. Activation of Acetate to Acetvl-COA. Viewed online on Jan. 15, 2007 at http://www.clemson.edu/genbiochem/People-Faculty-SmithK.html Tchobanoglous, G. and Burton F.L. 1991. Wastewater Engineering. Treatment, Disposal and Reuse. McGraw-Hill, New York. Tchobanoglous, G., Burton, F. L., Stensel, H.D. and Eddy, M. 2003. Wastewater Engineering : Treatment and Reuse. Boston, McGraw-Hill. Tetreault, M.J., Benedict, A.H., Kaempfer, C. and Barth, E.F. 1986. Biological Phosphorus Removal: A Technological Evaluation. /. Water Pollut. Control. Fed., 58(8), 823-837. Tomar, A., Eiterman, M.A. and Altman, E. 2003. The Effect of Acetate Pathway Mutations on the Production of Pyruvate in Escherechia coli. Appl. Microbiol. Biotechnol. 62:76-82. U.S. EPA Process Design Manual for Phosphorus Removal. 1976. EPA-625/1-76-001 a, Cincinnati, OH. Van Hellemond, J. J., Opperdoes, F.R. and Tielens, G.M. 1998. Trypanosomatidae Produce Acetate via a Mitochondrial Acetate:Succinate CoA Transferase. Proc. Nat. Acad. Sci. USA. 95: 3036-3041. Yeoman, S., Stephenson, T., Lester, J.N. and Perry, R. 1998. The Removal of Phosphorus During Wastewater Treatment. Environ. Poll, 49, 183-233 (1988a) Zeng, R.J., Lemaire, R., Yuan, Z. and Keller, J. 2003. Simultaneous Nitrification, Denitrification, and Phosphorus Removal in a Lab-Scale Sequencing Batch Reactor. Biotechnol. Bioeng. 84(2): 170-178. 86 Zhao, H.W. 1997. Oxidation-reduction Potential and Organic Carbon Sources as Two Control Parameters for Simultaneous Nitrification and Denitrification in Biological Nutrient Removal Process. Department of Civil Engineering. Vancouver, University of British Columbia: 263 p. Zhao, H.W., Mavinic, D.S., Oldham, W.K. and Koch, F.A. 1999. Controlling Factors for Simultaneous Nitrification and Denitrification in a Two-stage Intermittent Aeration Process Treating Domestic Sewage. Wat. Res. 33(4): 961-970. 87 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100876/manifest

Comment

Related Items