Enhanced Biological Phosphorus Removal

Abstract

Enhanced biological phosphorus removal (EBPR) is a biological process for efficient phosphate removal from wastewaters through intracellular storage of polyphosphate by polyphosphate-accumulating organisms (PAO) and subsequent removal of PAO from the system through wastage of sludge. In comparison to physical and chemical phosphorus removal processes, the biological process has several advantages such as high removal efficiency, low cost, and no chemical sludge production, but disturbances and prolonged periods of insufficient phosphate removal are still observed in conventional treatment systems and the applicability of the process for the treatment of saline waters remains unclear. In this PhD project two different aspects of the enhanced biological phosphorus removal were studied. In the first part of the research, potential existence of functional diversity among PAO clades and its influence on process performance was investigated, whereas in the second part of the study, salinity effects were assessed on the metabolism of polyphosphate-accumulating organisms (PAO) and glycogen-accumulating organisms (GAO). Functional diversity among PAO clades (Chapter 2 to 5) Although genetic diversity among PAO clades has been observed in past studies, PAO were often considered to behave functionally the same. Several recent studies suggested that PAO clades may be functionally different and that some PAO clades can perform better phosphate removal than other clades. Considering the significant role of the EBPR process in nutrient removal and recovery processes and the potential effect of PAO clades prevalence on the process performance, there was a need to investigate the potential existence of functional differences among PAO clades. The objective of this part of the study was to assess the existence of functional differences among PAO clades regarding the anaerobic metabolism in relation to their storage polymers and regarding the denitrification pathways. In Chapter 2, it was demonstrated in short-term experiments that significant functional differences exist between PAO I and II, with respect to the anaerobic volatile fatty acid (VFA) uptake metabolism. Although both PAO clades were able to shift their metabolism from a mixed poly-P and glycogen dependent metabolism to a metabolism that fully relies on glycogen, the HAc-uptake rate of both PAO clades decreased significantly where the decrease of HAc-uptake rates was most pronounced for PAO I. Consequently, at poly-P depleted conditions, the HAc-uptake rate of PAO II was four times faster than that of PAO I, whereas that of PAO I was slightly faster at poly-P non-limiting conditions. In addition, under conditions where poly-P was not limiting for the anaerobic HAc-uptake, PAO II performed a mixed metabolism that was partially dependent on glycogen and partially on poly-P for the generation of energy required of the HAc uptake, while PAO I relied to a much bigger extent on poly-P for the generation of energy. These findings are of major importance because they contribute to explain and clarify the controversy concerning the different stoichiometric and kinetic values observed in EBPR systems and are relevant for the development of operational guidelines for combined chemical and biological phosphate removal processes. In Chapter 3, the effect of the storage polymers on the metabolism of PAO II was assessed in long-term experiments and the results were compared to a previous study which was, based on the reported stoichiometry and kinetics, presumably conducted with a PAO I dominated biomass culture. The study supported the observations in short-term experiments regarding the functional diversity between PAO I and II. In addition, it provided interesting insights in the role of storage polymers on the regulation of the anaerobic HAc-uptake metabolism. As the influent P/C ratio increased, the poly-P content of the biomass increased while its glycogen content decreased. At higher P-contents, the kinetic P-release rates for HAc-uptake and maintenance increased. In parallel, the HAc-uptake rates increased up to an optimal poly-P/glycogen ratio of 0.3 P-mol/C-mol. Above that optimal ratio, the HAc-uptake rate decreased. The stoichiometry of the anaerobic conversions showed that a metabolic shift occurred from a glycogen dependent metabolism towards a poly-P dependent metabolism when the poly-P content of the biomass increased. The changes in the HAc-uptake rates suggest that at low poly-P contents the ATP formation rate is the rate limiting step, while at high P-contents (and, thus, low glycogen contents) the NADH production rate becomes the rate limiting step for HAc-uptake. Electron microscopy showed that poly-P is stored in the form of large granules in each PAO cell and, therefore, the rate of poly-P consumption may be surface area limited. Therefore, a decrease in the poly-P content of the biomass could limit the ATP production and thereby trigger the ATP production from glycogen conversion at a smaller rate. The findings contribute to a better understanding of the Accumulibacter clades metabolism under dynamic conditions and clarify population dynamics observed in previous studies. To confirm the observations in Chapter 2 and 3, it was assessed in Chapter 4 if certain PAO clades had the ability to proliferate under conditions where the phosphate concentrations were just enough for assimilation into biomass. In a SBR system, inoculated with activated sludge, a mixed PAO-GAO culture was enriched after 16 SRT that comprised of 49% PAO and 46% GAO of the total bacterial population. More specifically, all PAO were closely related to 'Candidatus Accumulibacter phosphatis' Clade II. Under anaerobic conditions, the mixed PAO-GAO culture performed a typical GAO metabolism in which all energy for HAc-uptake was produced by the conversion of glycogen. This study confirmed the findings in chapter 2 and 3 that PAO in general can perform a glycogen dependent metabolism but that PAO II had a competitive advantage over PAO I under phosphate limiting conditions. Under aerobic conditions PAO II were capable of instantly taking up excessive amounts of phosphate when additional phosphate was added to the reactor. The study also demonstrated that from a practical perspective, PAO may remain in PAO II dominated activated sludge systems under phosphate limiting conditions for periods of up to 16 SRT for instance due to overdosing of iron while still being able to take up phosphate aerobically when phosphate becomes available in the influent. Chapter 5 focussed on the denitrification pathways of PAO. Several literature studies suggest that PAO I is able to use both nitrite and nitrate as external electron acceptor while PAO II is only able to use nitrite. The results from those previous studies are contradictory and inconclusive as no studies were conducted with EBPR cultures highly enriched with specific PAO clades under appropriate conditions. In chapter 5 the oxidative pathways (oxygen, nitrite and nitrate) of a PAO I culture were investigated in combination with different VFA feed (HAc and HPr), firstly after a cultivation period in anaerobic/anoxic mode and secondly after a cultivation period in anaerobic/anoxic/oxic mode. After cultivation in anaerobic/anoxic/oxic mode, the enriched culture was not able to take up P in the presence of nitrate, despite the observation of low denitrification rates. In the presence of oxygen and nitrite, rapid P-uptake was observed. The big difference in denitrification rates with nitrite and nitrate together with observation that side populations were still present in the highly enriched biomass, resulted in the hypothesis that the side population in the biomass might have been responsible for NO3 to NO2 conversions, where the carbon source was mainly obtained from released soluble microbial products. This hypothesis was further supported by a comparison of literature values from studies conducted with PAO enrichment with various degrees of PAO enrichment. This comparison showed that the biomass specific P-uptake rate in the presence of nitrate increases when the fraction of side populations increases. In addition, the P-removal/N-removal ratio in many past EBPR studies under anaerobic/anoxic/oxic conditions with nitrite was higher than the P-removal/N-removal with nitrate, suggesting that PAO in general are not capable of using nitrate as external electron acceptor and are dependent on the partial denitrification activity of other organisms. Overall, this research revealed that significant functional diversity exist in the metabolism of PAO regarding the anaerobic metabolism while the study suggest that for the denitrification pathways among the Accumulibacter clades PAO I and II, functional differences may not exist. The differences in the anaerobic metabolism contribute to a better understanding of metabolic differences observed in past studies, provides more insight in population dynamics and are from a practical perspective in particular relevant for the development of nutrient recovery and/or combined chemical and biological P-removal systems. The findings of the denitrification pathways of PAO I provide a better understanding of the role of PAO in combined nutrient removal systems, helps to explain practical issues such as anoxic P-release in full scale wastewater treatment plants and support the development of measures to mitigate such issues in the performance. In addition to the functional diversity, the research provided more insight in the role of the storage polymers on the regulation of the anaerobic substrate uptake metabolism, which also leads to a better understanding of EBPR processes in full scale treatment plants under dynamic conditions. Although this study provided clear insights in the functional diversity of PAO clades and their metabolism, it is just the starting point of research focused on the functional diversity of PAO clades. To enable future research on the functional diversity, reliable selection methods or methods for isolation of PAO should be developed to obtain highly enriched or pure cultures with specific PAO clades. Impact of salinity on the metabolism of PAO and GAO during short-term exposure (Chapter 6 and 7) Saline wastewater can be generated by industry, intrusion of saline water in the sewerage or when saline water is used directly as alternative water source for non-potable purposes such as flushing toilets. To prevent the environment from severe environmental issues like hypoxia and eutrophication, the nutrients (C, N and P) need to be removed from saline wastewaters before its discharge to the receiving water bodies. However, salinity may negatively affect the microorganisms responsible for the nutrient removal in biological nutrient removal systems. This study assessed the effect of salinity on the metabolism of the microbial populations that prevail in EBPR systems (PAO and GAO). In Chapter 6, the short-term salinity effects on the anaerobic metabolism of PAO and GAO were assessed. It was demonstrated that salinity affected both PAO and GAO, with PAO being the most sensitive organisms. With increasing salinity the HAc uptake rates were inhibited while the maintenance requirements increased (up to 4% salinity) for both PAO and GAO. Interestingly, elevated salinity levels seemed to induce a shift from poly-P to glycogen consumption for HAc uptake and maintenance by PAO, whereas the stoichiometry of GAO related to the anaerobic HAc-uptake was unaffected. In addition, a structured model was developed, which could successfully describe the salinity effects on the different metabolic processes of PAO and GAO. In Chapter 7, the short-term salinity effects on the aerobic metabolism of PAO were assessed. The metabolism was very sensitive to even low salinity concentrations. An increase from 0.02 to 0.18% salinity led to a decrease in the specific oxygen consumption, PO4 and NH4 uptake rates of 25%, 46% and 63%, respectively. At 0.35% and higher salinity concentrations, the PO4-uptake, NH4-uptake and glycogen recovery were fully inhibited. Biomass growth was the most inhibited parameter, followed by poly-P formation and glycogen synthesis. The aerobic maintenance energy requirements increased up to a threshold concentration of 2% salinity, above which it rapidly decreased. To supply additional energy to cover the increasing maintenance requirements, P was released at salinity concentrations higher than 0.35%. The aerobic maintenance P-release followed a similar trend like the maintenance oxygen consumption. The inhibition model developed in this study could successfully describe the observed salinity effects on the different metabolic processes in this study. Overall, the research demonstrated that the EBPR process, in particular the aerobic stage, may be very sensitive to salinity. The findings suggest that any saline discharge equivalent to more than 5% seawater (with 3.4% salinity) addition or 15% brackish water (with 1.2% salinity) by either seawater toilet flushing, industrial discharges or saline intrusion can cause serious upsets of the EBPR process. This indicates that the EBPR process may not be applicable for saline wastewater treatment and that salinity may even be a relevant inhibition factor for activated sludge systems, treating conventional domestic wastewaters. The presented data is however based on short-term (hours) experiments using sodium chloride. To give the organisms the opportunity to acclimatize to higher salinity concentrations or giving the system the possibility to select for more salt tolerant PAO strains, future studies should focus on the long-term salinity effects on EBPR cultures, considering different salt compositions as well as different compositions of organic carbon in the synthetic wastewater.