Seasonality of nitrous oxide emissions at six full-scale wastewater treatment plants

The main focus of wastewater treatment has been to meet the effluent requirements specified by the environmental legislation while operating the treatment process cost-effectively. Recently, ambitious goals for climate change mitigation have been set at the country, municipality, and water utility levels (Pijuan & Zhao 2022). As a result, water utilities are increasingly incorporating climate change mitigation into their strategies and plant operations (Pijuan & Zhao 2022).

In wastewater treatment, direct nitrous oxide (N₂O) emissions from the treatment process can contribute over 60% to the total greenhouse gas emissions of a municipal wastewater treatment plant (WWTP) (Daelman et al. 2013; Kosonen et al. 2016; Maktabifard et al. 2022a). N₂O is produced during nitrification and denitrification as a part of biological nitrogen removal (Kampschreur et al. 2009). It is an ozone-depleting greenhouse gas with a global warming potential 300 times that of carbon dioxide (CO₂) (IPCC 2019). N₂O emissions exhibit a complex multivariate dependency on wastewater treatment process variables, such as dissolved oxygen (DO) level (Kampschreur et al. 2008), nutrient accumulation (Tallec et al. 2006), pH (Su et al. 2019), system shocks (Burgess et al. 2002), and the microbial community in the activated sludge (AS) (Gruber et al. 2021b).

High variation in the N₂O emission level has been measured at full-scale WWTPs around the world. For example, an emission factor (EF) range of 0.01–1.8% relative to influent total nitrogen load was measured by Ahn et al. (2010) in the US. The EFs measured by Gruber et al. (2020) varied between 1.0 and 2.4% at Swiss WWTPs. The high variability of N₂O emissions makes it difficult to represent the emission level using fixed EFs (Cadwallader & VanBriesen 2017; Gruber et al. 2020). Moreover, applying fixed EFs in carbon footprint analysis or life cycle assessment (LCA) (e.g. by Liao et al. 2020; Maktabifard et al. 2022a) likely leads to an unreliable estimation of the contribution of N₂O emissions. Fixed EFs may also erroneously harmonize the N₂O emissions to a similar level for each WWTP irrespective of their treatment process configuration and operational conditions. Thus, plant-specific N₂O monitoring is needed to reliably quantify the emission level (Vasilaki et al. 2019).

The focus of this paper was on studying how the N₂O emission level varies seasonally at Finnish WWTPs by conducting relatively short-term N₂O monitoring at several full-scale WWTPs under summer and winter conditions. Even though there is a consensus on long-term monitoring being recommendable (Ribera-Guardia et al. 2019; Vasilaki et al. 2019; Gruber et al. 2021a), the primary objective of this study was to assess the feasibility of fixed EFs in N₂O emission estimation rather than develop new shorter-term seasonal EFs for Finnish WWTPs. Therefore, the measured N₂O emission levels were critically compared to the fixed EF proposed by the Intergovernmental Panel on Climate Change (IPCC) (2019). Additionally, available process data were analyzed to explain the variability in N₂O emissions.

N₂O measurements were carried out at five full-scale AS plants (plants A–E) in Finland. Long-term N₂O monitoring data were received from a sixth Finnish WWTP (plant F) to support the study and plan the N₂O monitoring at plants A–E. Plants A–D are outdoor plants, whereas plants E–F have been built underground. Table 1 introduces the size of the plants in population equivalents (PE) and the requirements for nitrification or nitrogen removal set by the environmental permit for each plant.

The treatment process at each treatment plant begins with screening, grit removal, and primary clarification. The pre-treatment is followed by an AS process consisting of bioreactors and secondary clarifiers. The bioreactors are divided into several zones. The reactors typically have a denitrification–nitrification configuration with anoxic zones in the beginning, followed by aerobic zones. However, some of the plants operate them for part of the year as only nitrifying as they need to switch the denitrifying anoxic zones to aerobic zones to ensure there is enough aerobic treatment volume for the completion of nitrification even at low temperatures. The plants are especially prone to omit denitrification during the cold period, if their environmental permit requires only a sufficient nitrification rate but not total nitrogen removal, such as plants B and D (Table 1), as they are typically designed with more limited treatment capacity than WWTPs where total nitrogen removal is required. The sludge age is adjusted according to the wastewater temperature, and pH is controlled with alkali addition. Phosphorus is precipitated chemically at each plant. Additional details of the treatment process and the used chemicals are included in Table 1.

Measurement campaigns of 4–19 days were conducted at plants A–E to measure the direct N₂O emissions. The long-term N₂O measurement data from plant F supported the planning of the measurements, as the plant was assumed to be a good representation of the typical conditions and treatment process of Finnish full-scale WWTPs. The emission level at plant F exhibits distinctive seasonality, and thus the emission monitoring at plants A–E was conducted primarily with one measurement campaign in the warm period and one campaign in the cold period. A threshold water temperature of 12 °C was applied to separate warm and cold period conditions. Generally, the average water temperature has been detected to stay below 12 ± 1 °C for more than 6 months of the year at Finnish WWTPs (Kruglova et al. 2014). Additionally, the current Government Decree on Urban Wastewater Treatment 888/2006 does not require total nitrogen removal from WWTPs when the water temperature is below 12 °C (Ministry of the Environment Finland 2006). The water temperature at the plants varied from 12 to 20 °C during the warm period conditions and from 7 to 12 °C during the cold period conditions, respectively. Table 2 details the length and measurement locations of each measurement campaign. The nitrification rate was high (90–100%) during all measurements.

The N₂O measurements at plants A–E focused only on the aerobic zones, as they have been found to emit the majority of the N₂O emissions at plant F in comparison to the anoxic zones. For example, Maktabifard et al. (2022b) studied the dominant N₂O emission pathways at plant F with a process model and found that the anoxic zones produce only around 7% of the N₂O emissions, as the majority (93%) of the emissions originates from the aerobic zones. In addition, only a low contribution has been found from secondary clarifiers to the total N₂O emissions (Mikola et al. 2014). The process configuration and conditions as well as the wastewater quality at plants A–E were similar to plant F. The goal of the monitoring campaigns was to measure the first and last aerobic zones based on the assumed emission pattern within the zones (Blomberg et al. 2018; Maktabifard et al. 2022b). As the focus of the monitoring at plants A–E was on the first and last aerobic zones and the emissions for the middle aerobic zones were linearly interpolated, there are a few essential conditions for the validity of this approximation: (1) the majority of N₂O emissions is produced during nitrification, (2) nitrification proceeds gradually in the aerobic zones of the plug flow reactors, and (3) the aeration control should aim at maximizing the nitrification rate.

Plant F measures the N₂O emissions from the exhaust air pipe, as it is an underground plant, and all the off-gas from the treatment process is conducted to the open air via the pipe. The off-gas is analyzed with an FTIR-based Gasmet Continuous Emissions Monitoring System (CEMS) II. Measurement data from 2019 to 2022 were used in this study. Periods with atypical emission levels were defined by the plant operators, and these data were excluded from the analysis to consider only the typical seasonality of the emission level at plant F.

The off-gas N₂O concentrations for the unmeasured aerobic zones were estimated based on the measured zones with linear regression analysis. If only one zone was measured, the same N₂O concentration was assumed for all aerobic zones. The plants analyze the water quality, such as influent total nitrogen, with flow proportional, 24-h composite samples which are collected biweekly, weekly, or monthly.

Relevant process data were collected to support the analysis and find explanatory factors for the variability in the measured emission levels. To study the variability in the loading of the AS processes, the specific total nitrogen loading in the aerobic treatment volume (kg-N/kg-MLSS/d) was calculated. The specific loading relates the incoming total nitrogen load to the AS process to the mass of mixed liquor suspended solids (MLSS) in the aerobic zones. Additionally, dissolved nitrite (⁠⁠) was analyzed from grab samples near the off-gas hood to study its correlation to N₂O emissions. The standard for the analysis was SFS 3029:1976.

The EFs calculated for plants A–E as well as the off-gas N₂O concentrations in the last aerobic zone (Table 3) varied significantly, both between the plants and between two measurement campaigns conducted at the same plant. The EFs relative to influent total nitrogen varied from 0.04 to 1.0% in the warm period and from 0.2 to 1.8% in the cold period. Similarly, EFs were computed relative to the removed total nitrogen load to study how nitrogen conversion could affect the emitted N₂O. However, these EFs were close to the EFs related to the influent total nitrogen load, as the plants generally achieved relatively high total nitrogen removal (around 60–80%). The only notable exception was the EF from the cold period monitoring at plant B, as the EF relative to the influent total nitrogen load was 1.3%, whereas the EF related to the removed total nitrogen load was 28%. This was due to the low total nitrogen removal rate (around 7%), as there is no requirement for total nitrogen removal in the environmental permit of plant B. Thus, the study only considers the EFs relative to the influent total nitrogen load as they are more commonly used and comparable to other studies.

A comparison to reported EFs from measurements at other full-scale WWTPs and the fixed EF by IPCC (2019) (Table 4) revealed a similar emission level as was measured from plants A to F (Table 3). The IPCC factor is close to the highest measured EFs, but it is clearly an overestimation of the EFs below 0.5% that were measured especially during the warm period (above 12 °C). Over year-long measurements by Daelman et al. (2015) and Gruber et al. (2020) resulted in a high emission level of around 1–3%, while the EFs from shorter measurement campaigns by Ahn et al. (2010), Mikola et al. (2014), and Valkova et al. (2021) had a wide range from close to 0% to above 2%.

The results indicate that specific nitrogen loading (kg-N/kg-MLSS/d) in the aerobic treatment volume did not explain the high variation in the N₂O-N EFs, as no correlation was found between the variables. Similarly, the nitrogen removal rate did not link to the N₂O emissions at plants A–F. Plants achieving over 80% nitrogen removal produced both low (EF 0.04–0.09%) and high (EF 1.7%) emissions. On the other hand, low EFs of 0.08 and 0.2% were measured with nitrogen removal of 40 and 60%, respectively. Moreover, the BOD₇/N ratio in the influent to the AS process (Table 1) varied between 1.2 and 3.6, but it did not correlate with the EFs. We assumed that a low BOD₇/N ratio could result in increased N₂O emissions if there was not enough biodegradable organic matter for the completion of heterotrophic denitrification. The hypothesis was supported by the study by Zhang et al. (2012), who found that a sequencing batch reactor (SBR) treating influent wastewater with a BOD₅/N ratio of 0.7 produced higher emissions than an SBR treating influent wastewater with a BOD₅/N ratio of 2.9. Surprisingly, a low EF of 0.08% was measured at plant D even when the BOD₇/N ratio was very low (1.2).

The N₂O-N EFs from six Finnish full-scale WWTPs showed significant seasonality in emissions, with the EFs varying from near 0% to almost 2%. The seasonality could be partly due to the variation in wastewater temperature. High N₂O emissions were typically measured at low-temperature water (below 12 °C), while the emission level could be very low under warm period conditions (12–20 °C). High emissions were measured from AS processes with a nitrification configuration, whereas both low and high emissions were measured from denitrifying–nitrifying AS processes. The dissolved at the end of the aerobic zones of the AS bioreactors was found to be a good indicator of the N₂O emission variability during the cold period conditions. On the other hand, the nitrogen removal rate, BOD₇/N ratio, or specific nitrogen loading could not explain the variation in the measured emissions.

The EFs are in line with reported EFs from other full-scale WWTPs, and the high variability in the EFs highlights the unrepresentativeness of the fixed EF suggested by the IPCC. Furthermore, the EFs calculated from short-term measurement data depict well the N₂O emission seasonality but can be uncertain if disturbances to the normal operation of the WWTP occur during the measurements. Therefore, longer-term measurements are recommended to study the root causes of the emissions and determine the length of low and high emission levels.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.