Technical Brief 1: Pursuing Low-Cost Operational Changes to Mitigate Nitrous Oxide at Two Halton Region WRRFs

INTRODUCTION Nitrous oxide (N2O) is a potent greenhouse gas (GHG) with a global warming potential of nearly 300 times that of carbon dioxide. It is a major contributor to Scope 1 direct process emissions from wastewater treatment plants. In Ontario, Canada, the provincial electrical grid is already decarbonized, with a grid intensity <40 g-CO2-e/kWh, compared to the global average around 480 g-CO2-e/kWh (Ritchie and Rosado, 2020). Owing to low GHG emissions from electricity consumption, the Regional Municipality of Halton (Halton Region or Region) has estimated that N2O from treatment contributes a majority portion of Scope 1 GHG emissions at their wastewater treatment plants (WWTP). The Region's plant operations staff understand that N2O emissions may be an indicator of 'unhappiness' of the biomass and sub-optimal environmental conditions for treatment. Optimizing these conditions and removing stressors can reduce N2O emissions which is why operational optimization can be an effective mitigation approach. However, N2O generation mechanisms are complex due to multiple production and consumption pathways covering high dissolved oxygen (DO), low DO, and anoxic conditions (Figure 1; readers are referred to Ye et al. (2022) for a comprehensive explanation of N2O generation). Owing to variability of N2O production across space (i.e. different areas of a bioreactor) and time (daily, seasonally, yearly), Halton Region determined N2O needs to be measured to know what is happening. Therefore, the Region has undertaken a 2-year project with the goal of pursuing 'low hanging fruit' N2O mitigation strategies through operational optimization which is achievable based on current global experience. The objectives of the work include: (1) Measuring N2O emission factors (EFs) for Mid-Halton and Skyway WWTPs; (2) Identifying the operating parameters influencing N2O emissions & identifying low-cost operational optimization mitigation techniques; (3) Verifying the effectiveness of operational changes to reduce N2O emissions. METHODS In early 2024 Halton Region installed online dissolved nitrous oxide sensors (Unisense Environment, Aarhus, Denmark) at their two largest WWTPs, Skyway and Mid-Halton, with sensor data recorded starting early February 2024. The Skyway WWTP is rated 140 ML/d average day flow, with liquid treatment comprised of primary clarification, nitrifying activated sludge configured as long single pass tanks operated fully aerobic, dual-point ferric addition for phosphorus removal, deep bed tertiary sand filters, and sludge treatment comprised of thickening, anaerobic digestion, and dewatering. The Mid-Halton WWTP is rated 125 ML/d average day flow, with liquid treatment comprised of primary clarification, nitrifying activated sludge configured as 5-pass serpentine tanks operated with a year-round anoxic zone plus a seasonal anoxic swing zone, ferric addition for phosphorus removal, and sludge treatment comprised of thickening, anaerobic digestion, and dewatering. Figure 2 illustrates the bioreactors of both plants along with locations of dissolved oxygen and N2O sensors. Halton Region will conduct quarterly nitrogen profiling along the length of the bioreactors containing N2O sensors to understand where nitrogen conversion and N2O generation risk is highest. The Region created a standard operating procedure (SOP) and field data collection sheet for operations staff. The first bioreactor profiling was conducted at both plants on July 8, 2024 and the second one on October 28, 2024. Three sources of data are used: online SCADA data, lab data from the Region's laboratory information management system (LIMS), and a plant operations spreadsheet with daily summary data and operating information. A 'data lake' is generated by aggregating all data sources together through data merging and processing and calculating derived data (e.g. N2O mass emission rate, ammonia turn-over rate). Data analytics were performed including correlation 'heat maps' and other data synthesis. Artificial intelligence (AI) and machine learning (ML) methods are being applied through Cobalt Water Global's N2ORisk DSS which helps understand risk conditions for N2O, and is creating site- and location-specific ML models that will be used as soft-sensors later in the project when mitigations are tested. RESULTS Figures 3 and 4 illustrate the first 3 months of dissolved N2O concentrations measured at sensor location #1 at the Skyway and Mid-Halton WWTPs. As of writing this abstract, the second batch of data (months 4 through 8) are being analyzed. Four operational factors anticipated to relate to N2O generation are discussed. Dissolved Oxygen Both plants have high DO concentrations, routinely >4 mg/L, whereas the Cobalt Water N2O DSS recommends DO of 1-2.2 mg/L (Figure 5). There are periods of time when N2O generation is well aligned with DO changes (Figure 6). Optimizing aeration is the most common N2O mitigation technique owing to energy savings often being a co-benefit. Several European plants such as Eindhoven (Netherlands) have realized N2O reduction through DO adjustment (Porro et al., 2017) and this operational change is applicable to both Mid-Halton and Skyway WWTPs. Ammonia Turn-Over Rate Ammonia turn-over rate influences N2O generation at Mid-Halton and Skyway WWTPs. Figure 4 illustrates the Mid-Halton bioreactor MLSS compared to N2O concentration. The drop in both concentrations around March 14, 2024 was related to bringing the neighboring bioreactor online after a maintenance period. Operations staff built up the mass inventory prior to that time. With another bioreactor online the ammonia concentration and loading rate dropped resulting in less N2O. Bioreactor profiling was useful for understanding the nitrogen turnover rate along the length of the bioreactors, illustrated in Figure 7 for Mid-Halton and Figure 8 for Skyway. Although there is uncertainty in the nitrogen values at Mid-Halton (total nitrogen balance drops and then rises again), the highest ammonia turn-over rate was at positions 3, 4, and 5 at Mid-Halton (Figure 7) while it was positions 1 and 2 at Skyway (Figure 8). N2O production at a given sampling location was correlated with ammonia conversion in the upstream location (Figure 9). Skyway's profile was unexpected, and on investigation it was determined the ferric dosing prior to aerated grit tanks was resulting in extremely high primary clarifier performance with 80% TSS and 68% BOD removals. This extreme performance resulted in low organic loading to the Skyway bioreactors allowing immediate nitrification at the front of the tanks. These observations suggest a potential operational change to increase the MLSS so the biomass-specific ammonia turn-over rate will decrease. This mitigation was observed to reduce N2O emissions by 74% at the Avedøre WRRF (Andersen et al., 2022). Similarly, increasing the return activated sludge rate to dilute nitrogen concentrations and shorten hydraulic retention to push activity further down the train may also be useful. Diurnal Centrate Loadings At Skyway, the Region dewaters biosolids using a centrifuge about 6 hours per day for 5 or 6 days per week. The centrate is stored and then returned to the headworks over a four-hour period around 4-8pm. This slug load of ammonia reaches the aeration tanks in the late evening around midnight. The impact of this time-varying ammonia loading is observable in the N2O concentrations of Figure 10. Each day when there is dewatering there is a corresponding spike in N2O generation around midnight, while on the weekend when there is no dewatering there is no midnight N2O spike. At Skyway, extending the duration of centrate return would decrease the slug load ammonia. Rodriquez-Caballero et al. (2015) lowered N2O in a full-scale SBR with this technique. This approach is not applicable at Mid-Halton since dewatering centrate is already fully equalized and returned over a 24-hour period. pH There was also an observed correlation between N2O and pH at both plants (Table 1). This relationship could also be observed in the data (Figure 11). DISCUSSION/CONCLUSIONS Based on the first 3 months of data, Halton Region has observed N2O generation at the Skyway and Mid-Halton WWTPs is correlated as expected with DO levels, ammonia loadings, mass inventory levels, and pH. It is already clear there will be several operational changes that can be tested during year 2 to assess low-cost N2O mitigation. Interestingly, the N2O emission factors are quite different between the two WWTPs and differ from the IPCC default (Table 2). These results confirm the site-specific nature of N2O emissions and reaffirms the need for WRRFs to measure N2O to know their actual emission rate. At time of writing data from June to September is being analyzed. The conference presentation will include all data available and the Region will have decided which operational changes to test.