Nitrification/denitrification in a submerged anaerobic membrane bioreactor (SAMBR) for nitrogen removal

Abstract

Due to serious drawbacks in the conventional nitrification-denitrification process for nitrogen removal from wastewaters, such as high energy and carbon inputs and significantly increased sludge production, the more efficient and cost-effective Anammox process is starting to be used more widely. Anammox is an anaerobic biological process where around equimolar amounts of NH4+ and NO2- are oxidised/reduced, respectively, to produce dinitrogen gas without using a carbon donor. In the first part of this study, the start-up and performance of the Anammox process were evaluated in a 3-litre submerged anaerobic membrane bioreactor (SAMBR). The start-up using seed culture from an anaerobic digester (Anglian Water, UK) was relatively quick (60 and 70 d) for SAMBR 1 (HRT = 2d) and SAMBR 2 (HRT = 4d), respectively, compared to other reports in the literature. Both reactors showed quite high NH4+ and NO2- removal efficiencies of over 80% and 65%, respectively, resulting in a molar ratio of NH4+ to NO2- consumption of 1:0.9, which was comparable to recently reported values in the literature which are lower than the originally cited ratio of 1:1.32. Despite different HRTs, there were no significant differences in performance between the reactors. The use of a flat sheet membrane panel (~0.4 m pore size-Kubota, UK) was shown to be capable of shortening the start-up period for Anammox compared to continuous flow through reactors such as conventional CSTRs. The second major part of this work examined the novel process of partial pre-oxidation of NH4+ using nanofiltration (NF) hollow fiber membrane modules, to provide a feasible alternative to conventional partial nitrification preceding the Anammox process. Prior to investigating the feasibility of this process, ammonium oxidising bacteria (AOB) were enriched from both activated sludge and full-scale SAMBR sludge, in batch reactors, in order to determine whether it was possible to use anaerobic sludge as a source of nitrifiers. The enriched AOB demonstrated stable and high nitrifying activity throughout the enrichment period of 200-300 days, with average NH4+ removal efficiencies of over 90%. In the pre-oxidation process, AOB in the shell-side of the membrane unit was shown to be capable of oxidising the NH4+ mainly to NO2- which then diffused back into the tube side, resulting in a mixture of NH4+ and NO2- in the exit stream from the membrane unit. It was found that only flow rates of above 3.0 L/h were feasible, with a maximum NH4+ flux in the range of 8 – 10 g/m2 h. After 48 hours of operation, and at a flow rate of 5.0 L/h, an approximately equimolar ratio of NH4+ to NO2- was observed in the exit stream, and this would meet the requirement for the Anammox process as suggested by previous reports. This study has demonstrated the potential benefits of applying the Anammox process in a SAMBR for the treatment of nitrogen-containing wastewater as it could reduce the process start-up period, and the operation can be carried out at a short HRT. The application of a membrane process for the pre-oxidation of NH4+ was found to be reasonably promising at a laboratory-scale, and practically viable at a scale similar to actual SHARON reactor (Whitlingham STC, UK) based on an estimation of the number of HF modules needed. However, a proper optimisation study of the process is strongly recommended so that its feasibility could be further examined at a larger scale linking both processes together.