Determining the efficiency of the anammox process for the treatment of high- ammonia influent wastewater

Abstract

Domestic wastewater contains a high nutrient load, primarily in the form of Carbon (C), Nitrogen (N), and Phosphorous (P) compounds. If left untreated, these nutrients can cause eutrophication in receiving environments. Biological wastewater treatment utilizes a suspension of microorganisms that metabolize this excess nutrient load. Nitrogen removal in these systems are due to the synergistic processes of nitrification and denitrification, each of which requires its own set of operating parameters and controlling microbial groups. An alternative N-removal pathway termed the anammox process allows for total N-removal in a single step under anoxic conditions. This process, mediated by the anammox bacterial group, requires no organic carbon, produces negligible greenhouse gases and requires almost 50 % less energy than the conventional process, making it a promising new technology for efficient and cost-effective N-removal. In this study, a sequencing batch reactor (SBR) was established for the autotrophic removal of N-rich wastewater through an anammox-centric bacterial consortia. The key microbial members of this consortia were characterized and quantified over time using molecular methods and next generation sequencing to determine if the operational conditions had any effect on the seed inoculum population composition. Additionally, local South African wastewater treatment plants were screened for the presence of anammox bacteria through 16S rRNA amplification and enrichment in different reactor types. A 3 L bench scale SBR was inoculated with active biomass (~ 5 % (v/v)) sourced from a parent anammox enrichment reactor, and maintained at a temperature of 35 °C ± 1 °C. The reactor was fed with a synthetic wastewater medium containing no organic C, minimal dissolved oxygen (< 0.5 mg/L), and N in the form of ammonium and nitrite in the ratio of 1:1.3. The reactor was operated for a period of 366 days and the effluent ammonium, nitrite and nitrate were measured during this period. The hydraulic retention time was controlled at 4.55 days from Day 1 to Day 250, and thereafter shortened to 1.52 days from Day 251 to Day 360 due to an increased nitrogen removal rate (NRR). During Phase I of operation (Day 1 to Day 150), the reactor performance gradually increased up to an NRR of ~160 mg N/day. During Phase II (Day 151 to Day 250), the overall reactor performance decreased with the NRR decreasing to ~90 mg N/day, while Phase III (Day 251 to Day 366) displayed a gradual recovery of NRR back to the reactor optimum of ~160 mg N/day. The accumulation of nitrate in the effluent during the latter parts of Phase II and Phase III, coupled with oxygen ingress (~2.1 mg/L) in the same period, indicated that it was not the anammox pathway that was dominating N-removal within the reactor, but more likely the second half of the nitrification pathway mediated by the nitrite oxidizing bacteria (NOB). This was further confirmed through molecular analysis, which indicated that the bacterial population had shifted significantly over the course of reactor operation. Quantitative PCR methods displayed a decrease in all the key N-removing population groups from Day 1 to Day 140, and a marginal increase in anammox and aerobic ammonia oxidizing bacteria from Day 140 – Day 260. From Day 300 onwards, NOB had started dominating the system, simultaneously suppressing the growth of other N-removing bacterial groups. Despite this, the NRR peaked during this period, indicating an alternative mechanism for ammonia removal within the reactor system. A total population analysis using NGS was also performed, which corroborated the QPCR results and displayed a population shift away from anammox bacteria towards predominantly NOB and members of the phylum Chloroflexi. The proliferation of aerobic NOB and Chloroflexi, and the suppression of anammox bacteria, indicated that DO ingress was indeed the primary cause of the population shift within the reactor. Despite this population shift, N-removal within the reactor remained high. New pathways have recently emerged which implicate these two groups as potential N oxidizers, with specific NOB groups showing the ability for oxidation of ammonia through the comammox process, and members of the Phylum Chloroflexi being capable of nitrite reduction. This could imply that an alternate pathway was responsible for the majority of N-removal within the system, in addition to the anammox and conventional nitrification pathways. Additionally, in an attempt to detect a local anammox reservoir, eleven wastewater systems from around South Africa were screened for the presence of anammox bacteria. Through direct and nested PCR-based screening, anammox bacteria was not detectable in any of the activated sludge samples tested. Based on the operating conditions of the source wastewater systems, a subset of three sludge samples were selected for further enrichment. After 60-110 days of enrichment in multiple reactor configurations, only one reactor sample tested positive for the presence of anammox bacteria. Although this result indicates that anammox bacteria might not be ubiquitous within every biological wastewater system, it is more likely that anammox bacteria might only be present at undetectable levels, and that an extended enrichment prior to screening is necessary for a true representation of anammox bacterial prevalence in an environmental sample.