Evaluation of the effect of influentammonium : nitrite ratio on anammoxreactor efficiency

Abstract

The anaerobic ammonium oxidation (anammox) process has been recognized as an energy-efficient and cost-effective alternative to the conventional nitrification-denitrification route. The anammox process offers many advantages over the conventional processes such as less oxygen demand, non-requirement of external carbon source, and low operational cost. However, the major limitation of this process is the extremely slow growth rate of anammox bacteria and the need for stringent metabolic and reactor conditions leading to a long start-up period, which hinders its application in wastewater treatment. This study focused on evaluating the effect of key substrates (ammonium and nitrite) on anammox performance (nitrogen (N) removal) and community structure in anaerobic sequencing batch reactors (ASBR). For this, three 1L reactors containing different ammonium: nitrite ratios namely; Reactor 1 (1 NH4 + - N: 1.32 NO2 - -N), Reactor 2 (2 NH4 + -N: 1 NO2 - -N) and Reactor 3 (1 NH4 + -N: 2 NO2 - -N) were operated for 320 days using enriched anammox bacterial seed inoculum. The N removal performance of the reactors was assessed over time based on chemical and microbial analysis. From the results, the highest nitrogen removal efficiency (NRE) was observed in Reactor 3 containing high NO2 - -N (68.1 ± 7.7 %), followed by Reactor 1 containing the reported anammox stoichiometric substrate ratio (66.3 ± 13.3 %) and Reactor 2 containing high NH4 + -N (64.1 ± 7.2 %) on the 320th day of reactor operation. By using different substrate ratios, a significant variation (α= 0.05; P= 0.0004) in NRE in the three reactors was observed. Overall, the observed NO2 - - N (consumed)/NH4 + -N (removed), NO3 - -N (produced)/NH4 + -N (removed) ratios in Reactor 3 (1.38 ± 0.35 and 0.51 ± 0.34) was closer to the reported anammox stoichiometry ratio compared to Reactor 1 (0.88 ± 0.35 and 0.91 ± 0.48) and Reactor 2 (0.69 ± 0.32 and 0.72 ± 0.26) indicating a better anammox enrichment in Reactor 3. The inhibitory impact of free ammonia (FA) and free nitrous acid (FNA) concentration was monitored throughout the operational period. The FA concentration did not have a negative effect on anammox bacteria and AOB since the observed FA inhibitory concentration was below the reported inhibitory concentration of 1700 µg/L for anammox bacteria in all three reactors. As for FNA, Reactor 3 recorded the highest FNA concentrations (27.3 – 27.4 µg HNO2 - -N/L) throughout the study period. This FNA concentration did not negatively affect anammox bacteria on the 170th day, since anammox population was increased. However, long-term exposure resulted in anammox inhibition on the 320th day indicated by reduction of anammox bacteria. Whereas, nitrite oxidising bacteria (NOB) were not negatively affected by the observed FNA concentration, since their activity and growth was observed throughout the operation. As for Reactor 1 and 2, the FNA concentration (5.5 – 5.9 µg HNO2 - -N/L) was below inhibitory concentration on the 170th day. However, on the 320th day, the FNA concentration (6.2 – 7.3 µg HNO2 - -N/L) was above the reported inhibitory value resulting in anammox inhibition. A detailed exploration of the changes in the microbial community structures within the three reactors were studied by quantitative polymerase chain reaction (qPCR), sequencing and phylogenetic analyses. Using qPCR, Reactor 3 (1:2) with high NH4 + -N concentration showed high abundance of anammox bacteria followed by Reactor 2 (2:1) with high NO2 - -N concentration and Reactor 1 (1:1.32) having balanced NH4 + -N: NO2 - -N respectively on the 170th day. Thereafter, a shift from anammox bacteria abundance towards proliferation of AOB and NOB was observed on the 320th day. The AOB population was favoured by the fluctuating DO concentrations (0.39 ± 0.19 – 0.49 ± 0.20 mg/L). High AOB population observed in Reactor 1 (1:1.32) followed by Reactor 3 (1:2) and Reactor 2 (2:1) on 170th and 320th day. The NOB population was high in Reactor 3 (1:2) followed by Reactor 1 (1:1.32) and Reactor 2 (2:1) respectively throughout the operational period. High throughput sequencing analysis further showed a shift in the microbial community structure on 170th day with an increase in phylum Planctomycetes population from 0.76 % to 3.30 % in Reactor 1, 21. 32 % in Reactor 2 and 22.26 % in Reactor 3. The population of Proteobacteria increased from 6.38 % to 6.70 % in Reactor 1, 21.63 % in Reactor 2 and 21.73 % in Reactor 3. On the 320th day, Planctomycetes population decreased drastically to 2.84 %, 0.36 % and 4.91 % in Reactors 1, 2 and 3, respectively. Whereas Proteobacteria population further increased to 28.95 %, 24.15 % and 23.86 % in Reactors 1, 2 and 3, respectively. The Nitrospira population were below 0.10 % on the 170th day, however, an increase was observed on the 320th day from 0.01 % to 2.84 %, 7.38 % and 1.09 % in Reactors 1, 2 and 3, respectively which are in accordance with the qPCR results. In conclusion, different substrate ratios showed a significant influence on the overall N removal performance as well as on the selection of nitrifiers during the initial 170 days of operation. However, the long term operation of the reactors negatively affected the performance as well as community structure irrespective of the ratio used. Furthermore, the intermittent spike in DO and FNA concentrations (above inhibitory levels) could have affected the growth of anammox bacteria adversely. A further study based on continuous reactor operation is recommended for further verification of the results and prediction of unstable reactor episodes and possible process inhibitions in real-time.