A brief description of membrane bioreactor (MBR) historical evolution has been presented with emphasis on continual decline of treatment costs and energy requirements. Although MBR can operate at biomass (MLSS) concentrations 5 to 10 times higher than activated sludge these concentrations are limited in practice by increasing biomass suspension viscosity that in turn increases “reversible” membrane fouling and decreases oxygen transfer rates. “Irreversible” fouling is a major operational challenge since it depends on subtle interactions of membranes with various fractions of soluble microbial products resulting from microbial metabolism.

Abstract and Findings

With the development of urbanization, the excess amount of reactive nitrogen released to the environment caused many environmental issues. In San Francisco, over 60% of the nitrogen load comes from municipal wastewater treatment plants, which can provoke algal bloom and threaten the ecosystem. Sidestream is the wastewater derived from anaerobic sludge digestion. Due to its high nitrogen concentration, the sidestream needs to be treated separately. One optimized biological nitrogen removal process is the nitritation and ANAMMOX process. Compared to the traditional nitrification and denitrification process, the optimized process requires less aeration, zero-carbon inputs and produces much less sludge. The cost-effective nitritation and ANAMMOX process has been successfully applied for sidestream nitrogen removal. The thermal hydrolysis process (THP) is a pre-treatment process for anaerobic digestion. The pre-treatment can enhance anaerobic digestion efficiency. The wastewater produced with THP pre-treatment is called THP-sidestream (THP-S). However, previous work found that THP-S inhibit the nitritation and ANAMMOX processes. While the rate-limiting step and inhibition mechanisms remained unknown, a two-stage nitritation and ANAMMOX process were built to determine and study the inhibition mechanism of each process. Also, 16s rRNA and metagenome were applied to study the changes in microbial community structures after long-term THP-S treatment, providing biomarkers for tracking process performance.

The Nitritation process was established in the sequencing batch reactors (SBRs). The short-term effects of THP-S on ammonium-oxidizing bacteria (AOB) were evaluated. A higher THP-S percentage resulted in a lower specific ammonium removal rate by AOB. The Luong inhibition model best described the inhibition of AOB by THP-S. Long-term acclimation could not eliminate the negative effects of THP-S on the performance of the nitritation process. The nitrite accumulation percentage reduced by 55% using 100% of THP-S as feed. The AOB concentration decreased by 95% as the total biomass concentration increased. High-F: M-ratio batch tests also indicated that THP-S suppressed the growth of AOB.

ANAMMOX process was established in the membrane bioreactors. The short-term study revealed that the 1/20 diluted THP-S caused a 28% decrease of specific ANAMMOX activity. The MBR achieved a volumetric nitrogen loading rate of 3.64 kg/(m3·d) with undiluted regular sidestream (RS) feed, while the reactor crashed with 70% diluted THP-S as feed. The ratio of produced NO3--N to consumed NH_4+-N also decreased compared with RS feeding. Candidatus Brocadia was the dominant ANAMMOX bacteria species with the average abundances of 33.3% (synthetic wastewater), 6.42% (RS) and 2.51% (THP-S). The abundances of metagenome bins for dissimilatory nitrate reduction to ammonium (DNRA) increased in the system with THP-S compared with RS. The reason for the inhibition of ANAMMOX by THPS could be the high content of organic carbon in THP-S, which caused the over-population of heterotrophic bacteria, i.e., DNRA bacteria, leading to ANAMMOX bacteria washout.

To recover the nitritation process treating THP-S, the nitritation process was operated under extended sludge retention time (SRT), longer hydraulic retention time and elevated dissolved oxygen set-points. With the changes in operational strategies, the stable operation of nitritation treating THP-S was achieved. The nitritation pre-treatment can effectively reduce the concentrations of proteins and CODs, especially refractory CODs in both RS and THP-S. The subsequent ANAMMOX process also showed stable performance treating both nitritation pre-treated RS and nitritation pre-treat THP-S. However, rapid ANAMMOX process failure was observed having additional COD added into THP-S feed, implying that the high COD levels in THP-S can inhibit the ANAMMOX process. 16s rRNA sequencing revealed that the abundance of Nitrosomonas increased and the abundances of denitrification associated bacteria decreased with the extended SRT and aeration. This finding illustrates that nitritation treating THP-S can benefit from the new operational strategies. Similar bacterial community structures were observed in ANAMMOX RS samples and ANAMMOX THP-S samples. Additionally, ANAMMOX RS samples contained higher abundances of anaerobic digestion associated bacteria, indicating more refractory organics in RS. With extra COD added into influent, the ANAMMOX THP-S samples did not show many differences. The two-stage nitritation/ANAMMOX process can be a solution for treating THP-S.

The single-stage nitritation/ANAMMOX process, which can save operational cost while reducing footprints of the wastewater treatment plant, becomes the new trend in industrial application. Integrated fixed-biofilm activated sludge (IFAS) pilot was applied for the first time treating THP-S. Nitritation/ANAMMOX process was established in the IFAS pilot. Feeding with RS, the IFAS pilot achieved a much higher nitrogen removal rate (NRR) of 3.03 gN/L/d than the SBR pilot with NRR of 0.67 gN/L/d. The stable operation was achieved by the IFAS pilot treating THP-S, of which the NRR reached 1.38 gN/L/d. On the contrary, significant increases of effluent nitrogen concentrations were observed in the SBR pilot treating THP-S. DNA was extracted from IFAS biofilm samples, IFAS suspension samples, and SBR samples. According to 16s rRNA sequencing, the structures of the bacterial community were quite stable in IFAS biofilm samples despite changing the feed from RS to THP-S. With most ANAMMOX bacteria observed in IFAS biofilm samples and most AOB observed in IFAS suspension samples, the distribution reduced the diffusion limits of substrates, resulting in higher NRR of IFAS. Also, mutually symbiotic heterotrophic bacteria such as Limnobacter, Bryobacter, Truepera, and Sandaracinus were detected in the IFAS samples, benefiting the stable operation of nitritation/anammox process. When fed with THP-S, ANAMMOX bacteria was gradually washed out from the SBR pilot with the occurrences of fermentation bacteria such as Fastidiosipila, vandinBC27, and Methanosarcina. The results indicated that IFAS configuration, which promoted the build-up of the robust bacterial ecosystem, would be helpful for establishing stable nitritation/anammox treating THP-S.

This dissertation studied the inhibition of THP-S on separate nitritation and the ANAMMOX process for the first time and demonstrated the inhibition kinetics and inhibition mechanisms of THP-S on both processes. These findings are directly relevant to industrial applications. Recommendations for future research include applying different immobilization techniques on nitritation/ANAMMOX process to avoid washout while receiving better nitrogen removal efficiency, and developing THP-S pre-treatment methods to remove organic contents.