More Than Just Energy Savings: Understanding the Benefits of Low DO Operation

Introduction
Although advanced aeration controllers based on either ammonia (ABAC) or the ratio of ammonia to nitrate/nitrite (AvN) are becoming more common in WRRFs, the biological mechanisms of low DO operation are still not well documented. Low DO is generally referred to as operating below 0.5 mg/L but more specifically at a DO concentration that results in conditions that promote simultaneous nitrification and denitrification or SND, which can vary based on loading rates, process configuration, temperature, and floc morphology. Operating at low DO and achieving SND provides the opportunity for several process improvements in addition to achieving 10-30% aeration savings compared to conventional DO control. It can simplify operations and save additional energy by eliminating the need for multiple mixed liquor recycles and reduce supplemental carbon requirements. SND has also been shown to improve biological phosphorus removal. It is only when these mechanisms are understood can they be fully exploited to intensify treatment. This presentation will cover the current state of the industry as well as recent findings from an ongoing Water Research Foundation study (WRF 5083) investigating the benefits and mechanisms of low DO operation. The overarching research questions of this study that will be addressed are: 1. What setpoints represent low DO? 2. How are nitrification, denitrification, and phosphorus removal kinetics impacted by low DO operation? 3. What role does carbon cycling play in successful low DO BNR systems? 4. How do we maintain good settleability in low DO BNR systems?
Participating Facilities
The WRF study includes several participating facilities and pilot plants that are at different stages of planning, construction, or operation. The following facilities are either in operation or nearing the end of a demonstration phase. Rochester, MN: The Rochester WRF consists of a two-stage high purity oxygen plant and an activated sludge system operated in parallel. The activated sludge plant was operated in an A2O configuration and low DO operation was achieved by inputting low DO setpoints in the DO controllers and adjusting the DO setpoints as needed to prevent excessive ammonia bleed through. Hydrocyclone based wasting was also investigated to improve settling. City of Pueblo, CO: The James R. Dilorio WRF consists of a 19-MGD secondary facility using the Johannesburg BNR process. Aeration is controlled using a continuous AvN controller that results in an average low DO concentration of 0.2-0.5 mg/L. The facility has been achieving about 11 mg/L TIN and 0.6 mg/L TP with little chemical addition. Hydrocyclone based wasting is used at this facility to promote sludge densification. Operating at low DO has resulted in $150k/year in aeration savings and $300k/year in chemical savings. City of St. Petersburg, FL: The Southwest WRF is a 20-MGD A/O plant treating primary effluent. The plant operates at low DO concentrations and low SRT conditions to control nitrification and effluent N and P concentrations. Low DO operation has resulted to poor sludge settling affecting secondary clarifier capacity during the wet weather season. To improve the settling characteristics, hydrocyclone based wasting was recently implemented.
Results and Discussion
Using mixed liquor samples from the St. Petersburg facility, the specific denitrification rate was measured at different DO concentrations and the results are presented in Figure 1. As shown in this figure, denitrification was not measured until the DO concentration was less than 1 mg/L, but the rate was only approximately 20% of the maximum denitrification rate. It was not until the DO was lowered to 0.5 mg/L that the denitrification rate increased to about 70% of the maximum rate. Although denitrification at low DO was observed in the St. Petersburg batch tests and full-scale plant, at the Rochester WRF demonstration facility, SND was typically only observed in the first aerobic zone and not in the proceeding zones even though the DO was maintained at 0.3-0.4 mg/L as shown in profile in Figure 2. This was likely due to the type of substrate available after the first aerobic zone. An example of the impact of substrate type on denitrification is shown in Figure 3 where lower denitrification rates were measured after the readily biodegradable COD was consumed. These observations suggest that even lower DO concentrations are required in downstream zones that lack sufficient rbCOD and may rely on hydrolysis of slowly degradable COD. Testing at several facilities indicate that the maximum nitrification rates are not significantly reduced when adapted for low DO operation, but there is clear increase in the oxygen affinity. This results in higher nitrification rates at lower DO concentrations compared to conventional DO systems. The maximum nitrification rate is reduced by approximately 20-30% and apparent half saturation coefficient decreases to approximately 0.1 to 0.2 mg/L as shown in Figure 4 for the St. Petersburg facility. Overall, this only results in about 40% reduction of the nitrification rate at 0.2 mg/L DO. However, this rate can be increased by increasing the DO without losing a lot of overall nitrification capacity. Another observation in facilities operating at low DO concentrations is improved and more stable biological P removal. For example, the Rochester WRF had consistently lower OP concentrations as demonstrated in Figure 2. Lower nitrate in the RAS and more efficient use of influent carbon were obvious contributing factors, but higher P-uptake rates were also observed in low DO batch activity tests as shown in Figure 5. It is still unclear at this time what the mechanisms are, but it is hypothesized that low DO concentrations result in lower aerobic utilization rates of internally stored carbon that promotes higher P-uptake rates. These tests will be repeated at other facilities and the mechanisms investigated further in the pilot facilities. Poor settleability has often been associated with low DO operation, but these observations were from oxidation ditch plants that typically have poor settleability due to lack of proper selectors and plug flow conditions. As shown in Figure 6, if a process has well designed selectors that adequately reduce the influent rbCOD before the first low DO zone, the settleability remains good. Settling will be documented for the participating facilities.