A Full-Scale Demonstration of SND, Post Denitrification With Internally Stored Carbon and Anammox Potential For Energy and Carbon-Efficient BNR

Introduction
Availability and transformation of organic carbon at different stages in the activated sludge process impact the efficiency of biological nitrogen (N) and phosphorus (P) removal. Conventional biological nutrient removal (BNR) flowsheets often do not take full advantage of influent carbon and favorable influent carbon-to-nitrogen (C/N) and carbon-to-phosphorus (C/P) ratios for N and P removal. The reasons for the inefficiency of conventional BNR are rigid design guidelines without adequate considerations for local conditions, suboptimal operational strategies based on limited anecdotes, lack of understanding of carbon transformations at different redox conditions, gaps in the knowledge of bio-P, including selection factors for desired microbial community members, and inappropriate use of large safety factors resulting from a lack of process understanding. To this end, the main research objectives of this study were to: - Maximize influent carbon utilization for N and P removal rather than aerobic oxidation; - Integrate biological P removal and simultaneous nitrification and denitrification (SND) for energy and carbon efficient complete nutrient removal; and - Test potential for partial denitrification anammox (PdNA) in mainstream application for N polishing (with and without supplemental carbon).
Methodology The 26 mgd Seneca Water Resource Recovery Facility (WRRF) employs a 5-stage Bardenpho process, secondary clarifiers, and filters to meet stringent nutrient limits (< TN of 4 mgN/L, TP of 0.27 mgP/L). One of the five process trains was converted in Spring 2021 to a test train which involved the following changes: - Internal mixed liquor recycle (IMLR) reduced from 400% to 200% of the influent flow; - Ammonia-based aeration control (ABAC) control to maintain controlled dissolved oxygen (DO) levels in all aerated zones, including reaeration zone (Ammonia setpoint of 1.5 mg/L at the end of the aerobic zone; minimum DO setpoint of 0.2, maximum DO setpoint of 1.5 mg/L); - Methanol addition discontinued in the post-anoxic zone; and - Decreasing the size of post anoxic zone volume from 17% to 9% of the reactor basin. It is important to note that the plant does not have a primary clarifier, and the test train does not have dedicated clarifiers (combined return activated sludge [RAS]). Therefore, any microbial adaptation resulting from the operational conditions of the test train would impact the rest of the plant and vice-versa. The experimental plan included: - Reactor profiling for ammonia, OP, nitrate, and nitrite by taking two grab samples (morning and afternoon) every week; - Influent and effluent monitoring of carbon, nitrogen, phosphorus species; - Batch testing to determine the specific denitrification rates (without supplemental carbon), which involved collecting samples from the end of the aerated zone (just before the post-anoxic zone); - Monitoring airflow in the test train and the rest of the trains; - Collecting DNA samples for microbial population analysis; and - Collecting Polyhydroxyalkanoates (PHA) samples to understand the fate of carbon transformations for post anoxic denitrification (PHA analysis to be performed per Oehmen et al., 2005).
Results and Discussion Effluent quality The test train produced excellent effluent quality, as seen in Figure 1, without the use of supplemental chemicals. Based on weekly profiles, the average test train effluent ammonia was less than 0.2 mgN/L, TIN was 1.9 mgN/L, and OP was less than 0.2 mgP/L. Aluminum sulfate is dosed slightly downstream before the secondary clarifiers for additional phosphorus removal. SND and post-anoxic denitrification were responsible for enhanced nitrogen removal in the test train. The high degree of P uptake at low DO conditions resulted in very low effluent P levels. SND and aeration savings The low DO operation (~0.3 mg/L) achieved by ABAC resulted in significant aeration savings; about 30% less airflow was measured in the test train compared to the other trains operated at higher constant DO (~1.5 mg/L), which is presented in Figure 2. In addition, the low DO operation resulted in conditions favorable for SND, which removed an average of 5 mgN/L in the test train's aerated zones, as seen in Figure 3. Post-denitrification The test train removed > 4 mgN/L via denitrification in the post-anoxic zones without supplemental carbon and achieved similar effluent nitrate concentrations compared to the other trains, which do require supplemental carbon. The specific denitrification rate (SDNR) test results indicated that the low DO conditions resulted in high rates (Figure 4). These rates are much higher than the denitrification attributed to endogenous respiration, as seen in Figure 4. More efficient utilization of internally stored carbon may have resulted in more carbon availability for post-denitrification at low DO conditions.
Anammox potential The test train at the Seneca WRRF has demonstrated the ability to meet nutrient removal requirements without chemical addition. However, it appears that there is further opportunity to remove more N via the anammox pathway in the post-anoxic zone. The internally stored carbon-driven post-denitrification nitrite accumulation due to partial denitrification, likely aided by the shorter hydraulic retention time (HRT) of the post-anoxic zone (Figure 5). Further anoxic ammonia removal without the use of methanol at the Seneca WRRF would increase the treatment capacity and may be desired in the future. In the full paper, the findings from the following ongoing efforts will also be presented: 1. Anammox potential batch testing using the samples from the end of the aerated zone (to show that internally stored carbon generated nitrite is viable for anammox activity); 2. PHA profiling from the test reactor; and 3. Microbial analysis via 16S rRNA amplicon sequencing.
Conclusion - Optimization of ammonia-based aeration control (ABAC) resulted in low dissolved oxygen (DO) operation (<0.3 mgO2/L on average, saving > 30% in aeration energy), leading to significant simultaneous nitrification/denitrification (SND) (~60% nitrogen [N] removal in the low DO zones). - Low DO operation was compatible with biological P uptake, resulting in low effluent orthophosphate (OP) (< 0.2 mgP/L). - Reduction in internal mixed liquor recycle from 400% to 200% saved on pumping energy. - The efficient use of internally stored carbon in the low DO SND zone led to the carbon availability for denitrification in the post anoxic zone (> 4 mgN/L removed) without supplemental carbon. - Post denitrification rate as high as 2.7 mgN/gVSS/h was observed, which was significantly higher than what is expected from endogenous respiration. - The internally stored carbon driven post denitrification resulted in some nitrite accumulation (~0.8 mgN/L) via the partial denitrification pathway, which could be favorable for further N polishing via anammox - The demonstration train achieved effluent total inorganic nitrogen (TIN) < 2 mgN/L and OP < 0.2 mgP/L without supplemental carbon and metal salts addition.