Enhanced biological phosphorus removal (EBPR) is a biological wastewater treatment process facilitated by polyphosphate-accumulating organisms (PAO). The absence of isolates that have the PAO phenotype has limited the scope of studies into the physiology of these industrially significant and metabolically unique organisms. This thesis outlines findings into the physiology and ecology of EBPR in mixed microbial cultures, which contribute to the fundamental understanding of the process. The first experimental approach used in these studies was to investigate the microbial abundance of identified PAOs and GAOs in full-scale and lab-scale EBPR processes, and correlate these data with chemical monitoring methods both at a “macroscale” and “microscale”. The “macroscale” studies consisted of process optimisation experiments that found propionate to be a more effective and stable carbon source than acetate. The “microscale” study investigated the activity of Competibacter, growing in dense aggregates. This study discovered that the structure of the granules affected the distribution of activity by limiting the supply of oxygen and that the activity of the Competibacter in turn affected the structure of the aggregate. The second experimental approach was to target key facets of the microbial physiology of PAOs and GAOs at a molecular level. Environmental gene expression studies were used to investigate the stimulus for the expression of a putative Accumulibacter polyphosphate kinase gene (ppk). This study found that the expression of this gene was repressed by high external phosphate concentrations, which suggests that the pho regulon is functioning in Accumulibacter. In another study, previously published models were integrated and elaborated to develop a model for the membrane transport processes in PAOs and GAOs, which give them the unique ability to sequester VFA without an electron acceptor. These studies confirmed that the proton-motive force (PMF) drives the uptake of VFA by both PAOs and GAOs and postulated fundamental differences in the molecular mechanisms that PAOs and GAOs use to create a PMF in the absence of respiratory electron transport. The studies also explain the molecular basis for findings in other studies that PAOs have a competitive advantage over GAOs at increased pH. The third experimental approach was to attempt to isolate organisms significant to EBPR. Some measure of success was achieved: colonies of Competibacter were obtained in pure culture but the growth could not be sustained further than the growth of micro-colonies just visible to the eye. EBPR microbiology, like many other subjects of inquiry in environmental microbiology, has benefited greatly from developments in molecular methods to identify and describe microbial communities. However, the investigation of microbial physiology in the environment remains a challenge; this thesis has taken up that challenge. Discoveries regarding the benefits of propionate as a carbon source and the basis for the competitive advantage that PAOs derive from an increased pH have potential application for practitioners of EBPR plants. Furthermore the findings make a contribution to the fundamental understanding of the physiology of EBPR organisms that may in the future lead to entirely novel approaches to EBPR optimisation.

The objective of this research was to optimize the bio-P process as applied to a weak sewage with respect to HRT in each of the process zones. This goal was to be achieved by changing the HRT of the various zones with all other operating characteristics being held constant. The experimental work during this study involved two initially identical process trains operated in the University of Cape Town (UCT) mode. The aerobic zones of both trains were divided into four equal sized complete-mix cells to allow observations of phosphate uptake and poly-β-hydroxyalkanoate (PHA) consumption under aerobic conditions. After steady-state was established, the anaerobic HRT was reduced to 50% of the original value in the experimental module by reducing the anaerobic reactor volume. At the same time, the mixed liquor of both trains was drained, mixed and reapportioned to the two processes, thereby assuring equivalent starting conditions. Results of this study showed that both processes performed identically prior to the anaerobic HRT change. After the anaerobic HRT change, there was a forty day period where P removal and effluent P were the same in both process trains. This was so, even though the anaerobic P release was considerably less in the experimental module. Subsequently, a change in influent sewage type corresponded to a change in P removal and effluent P in the two process trains. An examination of the process parameters showed that the anoxic zone of the experimental module, after the anaerobic HRT change and the sewage change, consistently removed less P or released more P than in the control module. As a result, the control module out-performed the experimental module. Batch tests and tests to better characterize the influent sewage were then conducted in an attempt to determine the reasons for the different P removal characteristics. Under the test conditions, it appeared that the original anaerobic HRT was excessive. This was preferable to an insufficient anaerobic HRT, such as in the experimental module, however. The anoxic zone may have been too large, too small or just right for optimum P removal depending on the influent sewage characteristics. Optimizing the bio-P process by reducing the aerobic zone HRT appeared to have the greatest potential. / Applied Science, Faculty of / Civil Engineering, Department of / Graduate

This thesis assesses the potential for enhanced biological phosphorus (Bio-P) removal in Canadian wastewater treatment plants. Retrofit designs incorporating Bio-P removal were prepared for nine wastewater treatment plants across Canada, and were compared against chemical phosphorus removal technologies. Incremental capital and operating costs were calculated and internal rates of return (IRR's) for the capital investment required to install the Bio-P removal facilities were calculated. Based on these results, an assessment of the potential use for the technology in Canada is made. Of the nine plants studied, results indicate that Bio-P removal is economically superior to chemical phosphorus removal for the Calgary Bonnybrook, Edmonton Gold Bar, Saskatoon Mclvor Weir and Regina wastewater treatment plants. In general, Bio-P removal appears to offer significant economic advantages to plants located in Alberta and Saskatchewan because of the high cost of phosphorus removal chemicals in these provinces. The present low cost of phosphorus removal chemicals in Ontario and Quebec likely limits the viability of Bio-P removal to large (greater than 300,000 m³/d), suitably configured plants. In British Columbia, where Bio-P removal is presently used in the Okanagan Valley, the absence of widespread provincial phosphorus removal standards makes future Bio-P installations unlikely. The potential for Bio-P removal in Manitoba, the Maritimes and the Yukon and Northwest Territories is again limited by the absence of phosphorus removal standards in these parts of Canada. Results also indicate that the use of an anoxic/anaerobic/ aerobic process in the bioreactor, in conjunction with primary sludge fermentation through gravity thickening, is very applicable to Canadian plants and offers potential capital and operating cost savings relative to other Bio-P processes. The common practice of anaerobic sludge digestion, combined with sludge dewatering and land application, was found to be unfavourable from a Bio-P perspective unless the resulting supernatant/filtrate streams can be re-used or disposed of outside of the mainstream treatment process. Through the preparation of the retrofit designs, it was determined that certain aspects of Bio-P technology require additional research in order to optimize treatment plant design. These include kinetic modelling; short SRT Bio-P removal; the anorexic/anaerobic/aerobic process; the use of gravity thickening for primary sludge fermentation; and phosphorus release during anaerobic digestion. / Applied Science, Faculty of / Civil Engineering, Department of / Graduate

Passive reactive filters have the potential to provide effective phosphorus (P) removal from stormwater or agricultural drainage, or to act as an add-on P-removal technology for decentralized or small community wastewater treatment systems. Passive filters require minimal energy consumption and human maintenance. Direct-reduced iron (DRI), a steel-making intermediate, was investigated as a passive filter media for wastewaters phosphorus reduction. Phosphorus is a biologically active element that is in excess in many natural waterways due to intensive human activity. Eutrophication can occur when P concentrations exceed 0.02 mg/L in freshwater lakes and rivers. The harmful consequence of this phenomenon includes oxygen deprivation, fish death and cyanobacteria-produced toxins. There is a pressing need to limit phosphorus over-discharge into natural waterways. DRI is a novel media in the application of wastewater treatment and was characterized to have a porous structure with high metallic iron content. The phosphorus retaining mechanisms in batch and column studies suggest a combination of adsorption and surface crystal formation as the dominant removal mechanisms. Batch studies demonstrated increasing removal capacity with P concentration with a plateau observed at 21 mg P/g DRI relating to initial 3000 mg P/L. Media rejuvenation was investigated through chemical treatment with two iron solutions (Fe2(SO4)3, FeCl3) and two acidic solutions (H2SO4 and HCl) at varying molarity. P removal capacity could be fully recovered with 0.05 M Fe3+ or 0.4 N H+ (HCl/H2SO4), while a 37.6% P recovery was also achieved in an acidic solution at 1.2 N H+ (HCl/H2SO4). A column study utilizing three media sizes of DRI (3.5, 11, 19 mm) and one media size of activated alumina (AA) (7.5 mm) was conducted for 315 days using synthetic P solution varying from 2 to 10 mg/L and hydraulic retention times (HRTs) varying from 0.7 – 15 h. The results demonstrated that removal efficiency increased with HRT and decreased with increasing media size and concentration with minimum HRTs to maintain an 80% removal efficiency varying from 4.4 to 15 hrs for DRI and 3.9 hrs for AA for influent P concentrations of 10 mg/L and below. After 1 year of column operation, the DRI media had demonstrated a minimum removal capacity of 1.82 mg P/g DRI, which can be used as a conservative design parameter. A short duration column study (34 days) utilizing municipal lagoon effluent exhibited similar removal efficiencies to the synthetic column study under the same operational conditions. The 10 years lifespan DRI filter with 80% removal rate in the treatment of stormwater, municipal lagoon effluent, septic tank effluent and dairy wastewater application would have been estimated to have filter volumes of 0.24, 4.69, 15.3 and 36.2 m3, respectively.

No description available.

pH control is a tool to improve some aspects of Enhanced Biological Phosphorus Removal (EBPR) process. Filipe et al (2001a, 2001b, and 2001c) found strong evidence that the stability of EBPR systems can be improved by increasing the pH of the anaerobic zone, thereby creating conditions where phosphorus-accumulating organisms (PAOs) are able to take up acetate faster than glycogen-accumulating organisms (GAOs). They explained this observation by comparing the growth rate of phosphorus-accumulating organisms (PAOs) and glycogen-accumulating organisms (GAOs) and found that pH has little effect on PAOs growth rate but adversely affects GAOs growth rate when it increases (at pH values greater than 7.25, PAOs would take acetate faster than GAOs would). They used synthetic wastewater rich in acetic acid. In this study, we used real wastewater and the dominant volatile fatty acid available to microorganisms was propionic acid in continuous EBPR system. It was found that lower anaerobic zone pH (6.5 vs. 7.2) reduced the anaerobic P release both on an MLVSS specific basis and also on a non-specific (absolute value for the process) basis. In addition, the observed yield was significantly decreased. Aerobic P uptake was lower in the low-pH system (on a non-specific basis) due to the lower observed yield, and thus lower MLVSS concentration. Net P uptake was hard to interpret because of the effect of P release in the secondary clarifier of Train 2 (high pH). However, on a specific basis it was clear that net P uptake was either equal or better in the low-pH system regardless of how the secondary clarifier data was interpreted. Carbon transformations were not impacted in as consistent a fashion as anaerobic P release was. On a specific basis, PHA content remained unchanged although the PHV/PHB ratio was impacted with much lower PHV content in the low-pH system. Glycogen content and the amount of labile glycogen (delta glycogen) were higher in the low-pH system, in spite of the fact that MLVSS P content did not decrease. However, due to the impact of the low observed yield at low pH, absolute values resulted in higher PHA content for the process reactors as a whole, higher glycogen content, and unchanged labile glycogen. Low pH resulted in increased biomass P content, however the lower observed yield offset this on a process basis so that effluent P levels were nearly equal. So low pH improved P removal on a specific basis, but not on a process basis. Since it is unknown if the low observed yield is repeatable, and due to the impact of the secondary clarifier in the high pH system, it cannot be concluded that the effect of low pH on net P removal would be similar in other EBPR systems.

The objective of this dissertation was to provide a controlled comparison of identical continuous flow BNR processes both with and without prefermentation in order to provide a stronger, more quantitative, technical basis for design engineers to evaluate the potential benefits of prefermentation to EBPR in treating domestic wastewater. In addition, the even less understood effect of prefermentation on denitrification kinetics and anoxic phosphorus (P) uptake was studied and quantified. Other aspects of BNR performance, which might change due to use of prefermentation, will also be addressed, including anaerobic stabilization. Potential benefits to BNR processes derived from prefermentation are compared and contrasted with the more well-known benefits of primary clarification. Finally, some biokinetic parameters necessary to successfully model both the activated sludge systems and the prefermenter were determined and compared for the prefermented versus the non-prefermented system. Important findings developed during the course of this dissertation regarding the impact of prefermentation upon the performance of activated sludge treatment systems are summarized below: • For a septic COD-limited (TCOD:TP < 40:1) wastewater, prefermentation was found to enhance EPBR by 27.7% at a statistical significance level of alpha=0.05 (95% confidence level). • For septic P-limited (TCOD:TP > 40:1) wastewaters, prefermentation was not found to improve EBPR at a statistical significance level of alpha=0.05 (95% confidence level). • The increased anaerobic P release and aerobic P uptakes due to prefermentation correlated with greater PHA formation and glycogen consumption during anaerobiosis of prefermented influent. • Improvements in biological P removal of septic, non-P limited wastewater occurred even when all additional VFA production exceeded VFA requirements using typical design criteria (e.g. 6 g VFA per 1 g P removal). • Prefermentation increased RBCOD content by an average of 28.8% and VFA content by an average of 18.8%, even for a septic domestic wastewater. • Prefermentation increased specific anoxic denitrification rates for both COD-limited (14.6%) and P-limited (5.4%) influent wastewaters. This increase was statistically significant at alpha=0.05 for COD-limited wastewater, but not for P-limited wastewater.

Harmful algal blooms (HABs) can be detrimental to ecosystems, human health, and economies. The low levels of phosphorus remaining in the effluent of municipal wastewater treatment plants can contribute to HAB formation. To achieve more complete phosphorus removal, an effluent treatment method has been proposed that uses phosphorus-deprived, filamentous microalgae to quickly assimilate soluble phosphorus to low concentrations. This study investigated two parameters that influence the feasibility of such a system: (1) the biomass growth productivity of algal cultures during the phosphorus deprivation period and (2) the correlation between the duration of this period and the phosphorus uptake rate by the biomass when contacted with the water to be treated. A single strain of filamentous algae, Tribonema minus, was used. Two experiments lasting 8-9 days compared the biomass productivity of cultures of T. minus grown in phosphorus-replete and -deplete media. While no significant difference in productivity was observed between treatments, further studies should be done to confirm this finding. In addition, 39 uptake contact experiments were conducted. The soluble phosphorus uptake rate was measured for algae deprived of phosphorus for 0 to 12 days of growth. The highest observed uptake rate was 3.83 mg P/g VSS-h, during the first three hours of contact, by biomass that had been phosphorus-deprived for 12 days. The correlation between phosphorus deprivation duration and 3-h uptake rate was 0.34 mg P/g VSS-h per day of deprivation (R2 = 0.81). Additional development efforts seem justified based on these results.

Phosphorus (P) in stormwater runoff has detrimental effects on water quality and ecosystem health when it reaches surface waters and promotes algal blooms. Constructed wetlands (CWs) have been utilized to combat this problem by containing stormwater and removing excess nutrients. Including filter materials in the design of CWs has shown promise for increasing their capacity for nutrient removal. This mesocosm scale study was conducted outdoors over a 12-month period to evaluate the effectiveness of three filter materials in their ability to adsorb phosphorus, retain water, and support plant life. The filter materials examined were electric arc furnace (EAF) slag, engineered biochar, and sand. All treatments demonstrated positive plant response and the ability to retain water. The EAF slag and biochar removed significant amounts of P from effluent and appear to be suitable materials for integrating into CW design. Sand was found to be ineffective as a P filter.

Low dissolved oxygen (DO) biological nutrient removal (BNR) is becoming a viable option to improve the energy efficiency of BNR. To properly model and design BNR processes for low DO operation, it is critical to fully understand how nitrifier, denitrifier, and polyphosphate accumulating organism (PAO) oxygen kinetics adapt in a shift from traditional DO operation (2 mg O2/L or more) to low DO operation. Research characterizing how oxygen kinetics shift over time in activated sludge biomass adapting to low DO is limited. Therefore, a method to characterize oxygen kinetics for nitrifiers, denitrifiers, and PAOs simultaneously is lacking. Here a method was developed to simultaneously measure the oxygen kinetics of nitrifiers, denitrifiers, and PAOs. This method, termed the SND and P-Uptake Oxygen Kinetics test, was able to estimate the ammonia oxidizing bacteria (AOB) oxygen half-saturation coefficient, ammonia maximum removal rate, denitrifier oxygen inhibition coefficient, total inorganic nitrogen (TIN) maximum removal rate, PAO oxygen half-saturation coefficient, phosphorus maximum uptake rate, and a simultaneous nitrification and denitrification (SND) optimum operation point. Three tests were conducted on the Virginia Initiative Plant (VIP) BNR Activated Sludge Pilot while it was operating at a process DO of 2 mg O2/L, and one test while it was operating at 1.5 mg O2/L. The measurements among the three initial tests showed high similarity in their parameter estimates. Estimated oxygen half-saturation and oxygen inhibition coefficients were compared to current suggested ranges and were within the expected magnitudes. At 2 mg O2/L, denitrifier oxygen inhibition coefficients and PAO oxygen half-saturation coefficients were estimated to be remarkably low here, under 0.4 and 0.1 mg O2/L, respectively. AOB oxygen half-saturation coefficients were variable here in the range of 0.62 to 2.57 mg O2/L, seeming to vary with available ammonia concentrations. Upon comparison with a previously developed respirometric test for nitrifier oxygen kinetics, termed the Declining DO test, the AOB oxygen half-saturation coefficient from the SND and P-Uptake Oxygen Kinetics test and the Declining DO test, when both were conducted on the VIP BNR Pilot, showed a similar trend. This provided validation for the AOB oxygen kinetics here and the usefulness of the test developed here. Additionally, measuring and plotting AOB and denitrifier oxygen kinetics together produced an intersection point where ammonia removal rates were equal to TIN removal rates. This intersection point was an optimum point for SND during the conditions of the test. This method can be used to characterize and track oxygen kinetic changes in a BNR system adapting from high to low DO. / Master of Science / Aerating biological processes in wastewater treatment plants is necessary to facilitate nitrogen and phosphorus removal but is extremely costly. Traditional dissolved oxygen concentrations in these processes are 2 mg O2/L or higher. Operating processes with low dissolved oxygen (DO) concentrations, less than 1 mg O2/L, can cut costs significantly. However, designing processes at low DO concentrations requires knowledge of how microorganisms utilize substrate with lower oxygen availability and how substrate utilization develops when gradually decreasing the DO concentration in a process. Here, a method was developed to measure the parameters describing the relationship between substrate utilization and DO concentration for the microorganisms responsible for nitrogen removal (nitrifiers and denitrifiers) and phosphorous removal (polyphosphate accumulating organisms). Additionally, the method provides an optimum DO setpoint for simultaneous nitrification and denitrification (SND) during testing conditions. This method, termed the SND and P-Uptake Oxygen Kinetics test, was able to estimate the following parameters simultaneously: ammonia oxidizing bacteria (AOB) oxygen half-saturation coefficient, ammonia maximum removal rate, denitrifier oxygen inhibition coefficient, total inorganic nitrogen (TIN) maximum removal rate, PAO oxygen half-saturation coefficient, and phosphorus maximum removal rate. Three tests were conducted on the Virginia Initiative Plant (VIP) BNR Activated Sludge Pilot while it was operating at a process DO of 2 mg O2/L, and one test while it was operating at 1.5 mg O2/L. The measurements among the three initial tests showed high similarity in their parameter estimates. Estimated oxygen half-saturation and oxygen inhibitions coefficients were compared to current suggested ranges and were within the expected magnitudes. Upon comparison with a previously developed test for nitrifier oxygen kinetics, termed the Declining DO test, the AOB oxygen half-saturation coefficient from the SND and P-Uptake Oxygen Kinetics test and the Declining DO test when both were conducted on the VIP BNR Pilot showed a similar trend, providing validation for the usefulness of the test developed here.