Enhanced nitrogen removal in centrate sidestream treatment through modifications of aeration and free ammonia concentration

Abstract

High ammonia wastewater streams at a wastewater treatment plant (WWTP), such as the off-water (i.e. centrate) from anaerobic digestion, can negatively impact the performance of the plant's mainstream biological treatment basins by overloading them with nitrogen. The high ammonia load and temperatures (typically 15-35°C) of the centrate stream make it possible to pretreat the high ammonia stream in a sidestream treatment basin prior to combining it with the mainstream treatment train of the plant, which can significantly improve the effluent quality for the plant. This research focused on determining an operational strategy for flow-through sidestream centrate treatment to (1) promote nitrogen removal in existing treatment tanks through operational changes alone; (2) increase nitrogen removal in the sidestream basins while maintaining ammonia oxidation performance; and (3) characterize the effects of free ammonia (FA) and dissolved oxygen (DO) concentrations of nitrification and denitrification in the system. Strategies were investigated in bench-scale batch tests and in a full-scale sidestream system at a WWTP. Two full-scale sidestream treatment (SST) basins were used to systematically test different operational strategies in a side-by-side comparison with conventional sidestream treatment operation to induce short-cut nitrogen elimination. In one of the SST basins, modifications were made to the aeration pattern and the RAS-to-centrate ratio. The basin in conventional operation was fully aerated the length of the basin and aeration was controlled using a pH set point; the ratio of return activated sludge (RAS) to centrate was approximately 12. In the modified basin, aeration was provided using alternating aerated and anoxic zones and the RAS-to-centrate ratio was varied to test the effect of different free ammonia (FA) concentrations on performance. Although the aeration patterns were different between the basins, the amount of aeration to both basins was equal, approximately 2000 scfm, and having two aeration zones off in the modified basin in effect created a similar environment to that of the conventional basin where aeration was turned on and off based on pH. The DO depth profiles across the main compartment of each basin indicate that both basins had similar DO concentrations in the main compartment, typically below 1 mg O2/L. Despite aerating the conventional and modified basins using different aeration regimes, similar DO conditions resulted. Laboratory batch tests suggest that a range 4 to 5 mg FA-N/L is the optimal FA range to inhibit nitrite-oxidizing bacteria while ammonia-oxidizing bacteria (AOB) remain active. Full-scale monitoring, however, indicates that AOB activity was reduced at FA concentration greater than 2 mg FA-N/L with the best removals occurring while FA concentrations were between 0.5 and 2 mg FA-N/L. At FA concentrations greater than approximately 4 mg FA-N/L, complete inhibition of nitrification was observed. Overall, NH4+ and TIN removals in the conventional basin were 83% with a standard deviation of 10% (2613 ppd (std. dev. 706 ppd)) and 67% ± 9% (2327 ppd ± 720 ppd), respectively, while NH4+ and TIN removals in the modified basin were 55% ± 17% (1755 ppd ± 652 ppd) and 42% ± 18% (1366 ppd ± 693 ppd), respectively, when FA concentrations in the basin were between 0.5 and 2 mg FA/L. Modified operation led to a decrease in the percent and ppd of NH4+ and TIN removed as compared to conventional operation at a controlled RAS-to-centrate ratio of 12. Model 1 ANOVA of these removal rates determined that the mean removals for each basin were significantly different. The conventional basin monitored in this study also performed better for NH4+ and TIN removal as compared to the historical performance of the conventionally operated basin, which achieved removals of 47% ± 15% (2160 ppd ± 1055 ppd) and 28% ± 16% (1530 ppd ± 1030 ppd) for NH4+ and TIN removal, respectively. The improvement in removal of the study conventional basin over the historical conventional basin is likely due to controlling the RAS-to-centrate ratio to get more stable loading to the basin and due to the RAS-to-centrate ratio being approximately 3.5 times lower in the study basin as compared to the historical basin.