Nitrite build-up effect on nitrous oxide emissions in a laboratory-scale anaerobic/aerobic/anoxic/aerobic sequencing batch reactor

Biological wastewater treatment processes with biological nitrogen removal are potential sources of nitrous oxide (N2O) emissions. It is important to expand knowledge on the controlling factors associated with N2O production, in order to propose emission mitigation strategies. This study therefore sought to identify the parameters that favor nitrite (NO2 ) accumulation and its influence on N2O production and emission in an anaerobic/aerobic/anoxic/aerobic sequencing batch reactor with biological nitrogen removal. Even with controlled dissolved oxygen concentrations and oxidation reduction potential, the first aerobic phase promoted only partial nitrification, resulting in NO2 build-up (ranging from 29 to 57%) and consequent N2O generation. The NO2 was not fully consumed in the subsequent anoxic phase, leading to even greater N2O production through partial denitrification. A direct relationship was observed between NO2 accumulation in these phases and N2O production. In the first aerobic phase, the N2O/NO2 ratio varied between 0.5 to 8.5%, while in the anoxic one values ranged between 8.3 and 22.7%. Higher N2O production was therefore noted during the anoxic phase compared to the first aerobic phase. As a result, the highest N2O fluxes occurred in the second aerobic phase, ranging from 706 to 2416 mg N m h, as soon as aeration was triggered. Complete nitrification and denitrification promotion in this system was proven to be the key factor to avoid NO2 build-up and, consequently, N2O emissions.


INTRODUCTION
High nitrogen (N) concentrations in effluents may cause eutrophication and deterioration of recipient water bodies. N overloads favor microalgae and water plant growth, which may release toxins into the water (von Sperling, 2005). Although non-toxic, geosmin and 2methylisoborneol (2-MIB), two products released by cyanobacteria, can influence drinking water organoleptic characteristics, representing an obstacle to water treatment (Freitas et al., 2008). In order to prevent eutrophication, wastewater treatment plants (WWTPs) must be improved to ensure that N loads to receiving water bodies are within the limits stipulated by local legislation (Yang et al., 2017).
An economically viable and widely studied alternative for N removal is the application of biological processes involving nitrification and denitrification (von Sperling, 2005). However, the possibility of N2O release exists in both reactions, thus resulting in an anthropogenic source of this gas into the atmosphere (Wrage et al., 2001). In the troposphere, N2O is a chemically stable and long-lived greenhouse gas, with a global warming potential about 265 times that of carbon dioxide (CO2) (IPCC, 2014). Furthermore, in the stratosphere, N2O is the most emitted gas from anthropogenic sources displaying ozone (O3) depletion potential (Ravishankara et al., 2009). The highest N2O emission rates in WWTPs occur on those that apply biological processes, especially those that operate nitrification and denitrification processes in activated sludge (IPCC, 2019).
During nitrification under aerobic conditions, the ammonium ion (NH4 + ) is converted to hydroxylamine (NH2OH), which is, in turn, oxidized to nitrite (NO2 -) with the participation of ammonia-oxidizing bacteria (AOB) under alkaline conditions, while NO2is oxidized to nitrate (NO3 -) by nitrite-oxidizing bacteria (NOB). During denitrification under anoxic conditions, facultative heterotrophic bacteria convert NO3into molecular N (N2) (Wrage et al., 2001). N2O production is commonly attributed to three pathways: (1) partial nitrification, as a by-product of NH2OH oxidation; (2) nitrifier denitrification, which can occur under oxygen-limiting conditions, as an intermediate product; and (3) heterotrophic denitrification, where N2O is an intermediate product but can be released when the process is incomplete (Duan et al., 2017;Terada et al., 2017). Variations in N2O production and emissions occur according to the type of applied treatment process and configuration and operational parameters (Law et al., 2012).
N2O generation is usually associated with dissolved oxygen (DO) concentrations, NH4 + and NO2accumulation, pH and organic carbon availability (Duan et al., 2017;Vasilaki et al., 3 Nitrite build-up effect on nitrous … Rev. Ambient. Água vol. 16 n. 2, e2634 -Taubaté 2021 2019). Pijuan et al. (2014) reported N2O emissions almost ten-fold higher when altering a pilot plant system from continuous operation to a sequencing batch reactor (SBR). The authors attributed the N2O increases to the transient conditions between the anoxic and aerobic stages. Rodriguez-Caballero et al. (2015) also reported higher emissions in a full-scale SBR due to anoxic/aerobic transition. The authors also pointed out NO2build-up and the length of the aeration phases as contributors. Thus, SBR may become a potential source of emissions when applying operational conditions that favor N2O generation.
Knowledge of parameters that affect N2O production during the nitrification and denitrification stages is necessary to improve the sustainability of the process (Blum et al. 2018). Therefore, this study evaluated and identified the parameters responsible for NO2buildup and its effects on N2O production and emission in an SBR operated under anaerobic/aerobic/anoxic/aerobic conditions.

SBR operation
The study was carried out in a laboratory-scale anaerobic/aerobic/anoxic/aerobic SBR ( Figure 1A). The different SBR system phases were adjusted and regulated by a programmable logic controller (PLC), favoring higher DO and oxidation reduction potential (ORP) control. The reactor comprises 8.1 L of working volume and treats 4 L during each 8-hour cycle ( Figure 1B). An air compressor pump was used to provide system aeration, with an air flow rate of 120 L h -1 . Peristaltic pumps were used for the feeding and discharge of raw and treated wastewater, respectively. A mixed liquor volume was removed from the reactor during each cycle by a peristaltic pump, to guarantee a solid retention time (SRT) of 30 days. The reactor was inoculated with sludge from a WWTP designed to treat sanitary wastewater from a 2,500 population equivalent (PE) and fed with synthetic wastewater, which was prepared by adapting the formulation used by Holler and Trösh (2001). The synthetic wastewater was composed of casein peptone (500 mg N L -1 ), beef extract (323 mg N L -1 ), dibasic potassium phosphate (35 mg N L -1 ), sodium chloride (24 mg N L -1 ), urea (23 mg N L -1 ), calcium chloride dihydrate (23 mg N L -1 ) and magnesium sulfate heptahydrate (11 mg N L -1 ). SBR stabilization took 2 months. After this period, samples were collected for efficiency and N2O production and emission assessments.

Sampling and analysis
Monitoring took place for six consecutive weeks, where one sampling of one cycle was performed (one cycle per week). Throughout the sampling stage, raw and treated wastewaters were collected for chemical oxygen demand (COD), dissolved organic carbon (DOC) and total nitrogen (TN) analyses. In addition, volatile suspended solids (VSS) were analyzed in the discharged mixed liquor. After each metabolic phase, mixed liquor samples were collected for NO2and NO3analysis. Dissolved and emitted N2O sampling were also carried out during this period.
COD and VSS were determined according to APHA et al. (2012). DOC and TN analyses were performed on a TOC-L and TN Analyzer model TOC-L/TNM-L (Shimadzu). NO2and NO3analyses were performed by a Personal IC ion chromatograph (Metrohm) with conductivity detector, using a Polyvinyl alcohol with quaternary ammonium groups column (Metrosep A Supp 5 -150x4.0 mm); a solution of Sodium Carbonate (3.2 mmol L -1 ) and Sodium Bicarbonate (1.0 mmol L -1 in 5% acetone) was used as anion carrier and a solution of Tartaric Acid (4 mmol L -1 ) and Dipicolinic Acid (0.75mmol L -1 ) was used as cation carrier. A HI 9828 multiparameter probe (Hanna) was used to monitor reactor DO, ORP, temperature and pH.
A technique similar to the one applied by Brotto et al. (2015) was used for the emitted N2O sampling. Emitted N2O was collected using a modified lab-scale upturned funnel partially submerged in the reactor and a syringe ( Figure 1A). The N2O flux (F) was calculated using Equation 1: (1) Where Qupturned funnel stands for the emerging air flow rate passing through the upturned funnel, Δ[N2O] is the difference between the N2O concentration determined in the reactor and the concentration present in the atmosphere and Aupturned funnel is the superficial area of the upturned funnel.
Dissolved N2O was collected from the mixed liquor using a syringe and the concentration in the liquid was determined by the headspace gas method (de Mello et al., 2013). Dissolved N2O concentrations (C) were calculated with Equation 2: Where K0 is the N2O solubility coefficient (Weiss and Price, 1980), Chs is the N2O concentration stripped from the liquid (final), P is the atmospheric pressure, R is the universal gas constant, T is the liquid temperature (K) and Car is the N2O concentration present in the atmosphere (initial).

RESULTS AND DISCUSSION
The SBR reached average COD, DOC and TN removal efficiencies of 89, 91 and 79%, respectively, similar to those reported by other authors (Jia, et al., 2012;Rodriguez-Caballero et al., 2015). Jia et al. (2012) reported removal efficiencies of 91 and 85% for COD and TN, respectively, in an anaerobic/aerobic SBR designed for simultaneous nitrification and denitrification (SND), with a 6-hour cycle. Rodriguez-Caballero et al. (2015) also reported removal efficiencies near 90%, for both COD and TN in a full-scale SBR operated with alternate aerobic and anoxic phases in 4.5-hour cycles.
Despite the high TN removal obtained during this study, a high concentration of NO2at the end of both the aerobic and anoxic phases was noted, with higher values in the aerobic phases ( Figure 2). Too low NO3concentrations (<0.16 mg N L -1 ) during the cycle were also observed. These results can be an indicator that both nitrification and denitrification were only partial, as reported by other authors (Guo et al., 2009;Stenström et al., 2014;Du et al., 2016). Stenström et al. (2014) also observed NO2build-up in a SBR applying both nitrification and denitrification, indicating higher NO2production during nitrification and consumption reduction during denitrification throughout the study. Guo et al. (2009) observed NO2concentrations over 25 mg N L -1 and NO3concentrations under 6 mg N L -1 in an SBR aerobic designed for partial nitrification. In the present study, the NO2accumulation rate reached 96%. Du et al. (2016), for an anoxic SBR designed for partial denitrification, reported that NO2accumulated due to decreased NO2enzyme reduction activity. It is known that high nitritation rates may result in decrease of pH, provoking a reaction shift of nitrite and nitrous acid upon pH below 5, which may impact the denitrification (Todt and Dörsch, 2016). However, the pH values measured in the present study were above 5, which may not have been sufficient to increase nitrous acid production. This reinforces the theory that partial nitrification was responsible for the accumulation of nitrite.  Figure 3A presents NO2production (first aerobic phase), consumption (anoxic phase) and build-up (operational cycle) throughout the study. During the first aerobic phase, NO2 -consumption rates were high and varied from 43.1 to 63.8 mg N h -1 . In the next phase (anoxic), the consumption rates decreased throughout the sampling period, from 40.4 to 20.3 mg N h -1 . The combination of high partial nitrification rates (NO2generation) during the first aerobic phase with decreased NO2consumption in the subsequent anoxic phase led to NO2build-up at the end of each operational cycle. The NO2build-up rate increased from 29 to 57% at the end of each operational cycle and expresses the NO2percentage that was not consumed during the anoxic phase in relation to that produced in the previous phase (aerobic). Wu et al. (2011) observed NO2accumulation rates near 90% in an anaerobic/aerobic SBR designed for partial nitrification with lower biomass concentrations, while Stenström et al. (2014) reported that NO2accumulation can lead to increased N2O production and emission, as observed herein.  Figure 3B indicates NO2production (first aerobic phase) and consumption (anoxic phase) rates and the N2O/NO2ratio for each phase (first aerobic and anoxic) during the study. The increased NO2build-up rate coincides with increased N2O production rate, raising the N2O/NO2ratio ( Figure 3B). During the aerobic phase, N2O production accounted for 0.5 to 8.5% of NO2production, where the main production mechanism was nitrification. During the anoxic phase, values were substantially higher and corresponded to the evolution of the denitrification process, ranging from 6.3 to 22.7%. Therefore, denitrification was responsible 7 Nitrite build-up effect on nitrous … Rev. Ambient. Água vol. 16 n. 2, e2634 -Taubaté 2021 for the higher N2O production in this type of system, being extremely important for the control of the operational conditions of the anoxic phase to mitigate N2O emissions in the following phases, mainly the aerobic stage. Rodriguez-Caballero et al. (2015) reported NO2build-up as a key factor in increasing N2O production during nitrifying denitrification. Mampaey et al. (2016) described the anoxic phase as an important N2O generation factor, in addition to high NO2concentrations. According to these authors, the anoxic phase was responsible for 70% of N2O production in a SHARON reactor. Therefore, effective operational parameter control, in order to minimize NO2accumulation and, consequently, N2O supersaturation in the liquid, is necessary for N2O emission mitigation measures (Kampschreur et al., 2009;Vasilaki et al., 2019).
The higher N2O production rate occurring simultaneously with NO2build-up in the system may be associated with a reduction in the reactor biomass (VSS). Figure 4 displays the NO2production (first aerobic phase) and consumption (anoxic phase) rates in parallel with the VSS concentrations in the reactor throughout the study period. A biomass concentration reduction of approximately 30% at the end of the sampling period was observed. The loss of biomass may be related to a mechanical problem in the mixer observed during the fifth week of sampling. The mixer malfunction led to sludge flotation during sedimentation, causing biomass losses through the treated wastewater discharge and, consequently, a sharp reduction in SRT. This event may have caused decreased efficiency of both the nitrification and denitrification processes, resulting in N2O and NO2accumulation. Wu et al. (2011) observed that decreased suspended solid (SS) concentrations in an SBR system favored NO2accumulation. Noda et al. (2003) reported higher concentrations of dissolved and emitted N2O in anoxic and oxide reactors with lower SRT. Other authors have also associated lower SRT with NO2build-up and increased N2O production and emissions (Hanaki et al., 1992;Kampschreur et al., 2009;Castellano-Hinojosa et al., 2018). As previously reported, solid losses in SBR alongside decreased SRT are likely to favor increasing NO2build-up and N2O generation. However, an additional effect was observed regarding the magnitude of the N2O transfer rate from the liquid to the atmosphere during the second aerobic phase. Figure 5 presents the N2O production rates (from the retained and notemitted portion) of the first aerobic and anoxic phases in parallel to the maximum N2O flux at the beginning of the second aerobic phase throughout the study period. The N2O flux peak occurred as soon as aeration began during the second aerobic phase, with substantially high values ranging from 706 to 2416 mg N m -2 h -1 . These findings are close to those reported by Ribeiro et al. (2017) in a conventional activated sludge WWTP with landfill leachate addition, where a maximum N2O flux of 1890 mg N m -2 h -1 was observed, which was correlated to decreased DO concentrations and partial nitrification. Figure 5. Non-emitted N2O production rate (mg N h -1 ) during the first aerobic and anoxic phases and maximum flux emitted during the second aerobic phase (mg N m -2 h -1 ).
In the present study, the peaks from the second aerobic phase represented the amount of N2O produced and retained (not emitted) from the previous phases (first aerobic and anoxic). The same behavior for N2O/NO2accumulation in the liquid phase was observed for the maximum N2O flux throughout the study period, with an increase in emitted N2O in parallel with an increased N2O production rate retained from the previous phases ( Figure 5). Other studies have reported the same liquid N2O accumulation problem during the anoxic phase and its implications for the next aerobic phase (Gustavsson and La Cour Jansen, 2011;Yang et al., 2017;Pijuan et al., 2014;Mampaey et al., 2016). Thus, alternation between the anoxic and aerobic phases can be a negative point for N2O emission mitigation in wastewater treatment processes applying N removal.
Therefore, operational control in order to favor lower N2O production during the aerobic phase and its rapid consumption in the subsequent anoxic phase are necessary to mitigate emissions in the following phases, especially in the aerated units. Otherwise, in systems with a subsequent aerobic phase, N2O produced and not consumed may be emitted. In systems without this subsequent step, N2O may be emitted during treated wastewater disposal into receiving water bodies. Other operational adjustments, such as NO2and SRT control, are extremely important to create favorable conditions for nitrification and denitrification processes without liquid N2O accumulation and subsequent emission.

CONCLUSION
This study correlates N2O production and emissions with the operational condition of an SBR undergoing anaerobic/aerobic/anoxic/aerobic phases. The main conclusions are: • Even with controlled DO and ORP, partial nitrification was observed during the first 9 Nitrite build-up effect on nitrous … Rev. Ambient. Água vol. 16 n. 2, e2634 -Taubaté 2021 aerobic phase, resulting in high NO2production. The NO2was not totally consumed during the anoxic phase, also indicating the partial denitrification.
• NO2build-up favored N2O production during both the aerobic and anoxic phases of the process. The increases observed in both parameters can be associated with decreased biomass concentrations.
• The anoxic phase was responsible for the highest N2O production rates. The N2O accumulated during this phase was released during the second aerobic phase, causing emission peaks as soon as aeration began.
• An environment able to sustain complete nitrification and denitrification is required, minimizing NO2accumulation and allowing for rapid N2O consumption, thus minimizing emissions.