The continuous measurements of N2O emissions from the aeration tanks of three activated sludge wastewater treatment plants (WWTPs) operated with biological nitrogen removal (BNR) and non-BNR were performed during the different operating conditions of several parameters, such as aeration, dissolved oxygen (DO) profiling and organic shock loading (with landfill leachate). The nitrification process is the main driving force behind N2O emission peaks. There are indications that the variation of the air flow rate influenced N2O emissions; high N2O emissions denote over-aeration conditions or incomplete nitrification, with accumulation of NO2 concentrations. Thus, continuous measurements of N2O emissions can provide information on aeration adequacy and the efficiency of complete nitrification, with major focus on DO control, in order to reduce N2O emissions. An additional concern is the observed propensity of WWTPs in developing countries to receive landfill leachates in their wastewater systems. This practice could have adverse effects on climate change, since wastewater treatment during periods of organic shock loading emitted significantly higher amounts of N2O than without organic shock loading. In short, non-BNR WWTPs are subject to high N2O emissions, in contrast to BNR WWTP with controlled nitrification and denitrification processes.

One of the major challenges for wastewater treatment plants (WWTPs) with biological nitrogen removal (BNR) is the adequate control of operating parameters responsible for complete nitrification and denitrification processes and, consequently, undesirable nitrous oxide (N2O) emissions. BNR WWTPs are important to minimize the release of oxidized nitrogen (N) forms into water bodies (Strokal & Kroeze 2014) and the emissions of N2O to the atmosphere (Foley et al. 2010). On the other hand, N2O can be emitted in large amounts from BNR WWTPs with incomplete denitrification processes (Foley et al. 2010), contributing as additional anthropogenic sources of this gas to the atmosphere. N2O is a powerful greenhouse gas (IPCC 2013) and a stratospheric source of nitric oxide (NO), one of the gases responsible for the depletion of stratospheric ozone (Crutzen 1979; Ravishankara et al. 2009), therefore, any contribution other than those from natural processes is undesirable. According to Law et al. (2012), global N2O emissions from wastewater treatments are projected to increase by 13% during 2005–2020.

N2O production from WWTPs commonly comprises three pathways: (1) nitrification of ammonium or a hydroxylamine (NH2OH) oxidation step (releasing N2O as a by-product); (2) heterotrophic denitrification by nitrite or nitrate reduction (with N2O as an intermediate); (3) nitrifier denitrification by oxidation followed by reduction in a single step, under limited dissolved oxygen (DO) conditions: NH4+→NO2→NO→N2O (as an intermediate)→N2 (Wrage et al. 2001; Wunderlin et al. 2012). N2O emissions from WWTPs depend on certain operating parameters, such as organic matter (biochemical and chemical oxygen demand (BOD and COD)) and total N (TN) loads, DO concentrations, aeration rates, hydraulic retention time (HRT) and sludge retention time (SRT), among others (Kampschreur et al. 2009; Law et al. 2012). Few studies regarding the direct relationship between N2O emissions and operating parameters in full-scale WWTPs are available (Ahn et al. 2010; Law et al. 2012; Aboobakar et al. 2013; Rodriguez-Caballero et al. 2014; Daelman et al. 2015), especially in tropical countries (Brotto et al. 2010, 2015; de Mello et al. 2013; Ribeiro et al. 2013, 2015). These reports have shown a significant variability of the N load emitted as N2O, ranging from 0 to 25.4%, attributed to differences in WWTP designs or in the applied sampling strategies (Daelman et al. 2013).

The Intergovernmental Panel on Climate Change (IPCC) suggests an emission factor (EF) of 3.2 g N2O person−1 year−1 for WWTPs with controlled nitrification and denitrification processes (IPCC 2006). However, this N2O EF comes from a single study carried out in a small town in New Hampshire (USA) with weekly grab samples obtained in the morning from an activated sludge WWTP (Czepiel et al. 1995). Other studies conducted in other parts of the world have found different N2O EFs, especially when determined during different times of the day (Ahn et al. 2010; Rodriguez-Caballero et al. 2014; Daelman et al. 2015). Rodriguez-Caballero et al. (2014) and Daelman et al. (2015) reported a 24-h online sampling strategy showing diel variability of N2O emissions. Moreover, a short-term sampling strategy can demonstrate relationships between N2O emissions and the operating parameters of full-scale WWTPs (Ahn et al. 2010; Aboobakar et al. 2013; Rodriguez-Caballero et al. 2014). Therefore, the temporal pattern of the dynamics of N2O emissions and their major control factors in activated sludge WWTPs are of major importance (Rodriguez-Caballero et al. 2014) in order to apply the response of N2O emissions to aeration adequacy and evaluate the efficiency of complete nitrification.

In this context, the main goal of this study was to quantify N2O emissions from activated sludge WWTPs operated with BNR and non-BNR, with different aeration systems, DO profiling and organic shock loading, with major focus on DO control, a key factor in the process performance, operational and maintenance costs, and mitigation of N2O emissions.

WWTP characteristics

The first evaluated WWTP (WWTP-1) is located at a research institution in the city of Rio de Janeiro, Brazil, and serves about 2,500 population equivalents (PE). This plant treats an average daily raw wastewater flow of 44 m3 h−1 (consisting of a mixture of laboratory and domestic wastewaters). The raw wastewater flow is linked to the research institution's work regime, so it is not continuous throughout the day. During the 24-h study period, the flows were provided by the plant operators and no raw wastewater entered the WWTP-1 aeration tank from 2 to 4 PM and from 6 PM to 8 AM. The process applied in this WWTP is extended aeration activated sludge with non-BNR. The volume and surface area of the aeration tank are 440 m3 and 110 m2, respectively. The HRT and SRT were 12 h and 25 days, respectively. During the study period, variations in the air flow rate of 135, 205 and 450 m3 h−1 were applied by manual regulation of the air blower, without a DO-controlling system.

The second evaluated WWTP (WWTP-2) is located in the city of Presidente Prudente, São Paulo State, Brazil, and treats domestic wastewater of 200,000 PE, generating an average daily raw wastewater flow of 1,700 m3 h−1. The raw wastewater flow was continuous during the 24-h study period, with higher values from 5 AM to 4 PM. WWTP-2 operates with extended aeration activated sludge with BNR. This plant carries out BNR by simultaneous nitrification and denitrification (SND) processes. The volume and surface area of the aeration tank (three aeration tanks in total) are 6,940 m3 and 1,734 m2, respectively. HRT and SRT were of 16 h and 12 days, respectively. The plant operates with an aeration automatic system controlled by DO concentrations to promote SND. Continuous DO probes are installed at three sites (inlet, middle and outlet regions) of the aeration tank. The air flow rate per tank is automatically maintained in the range of 6,120 to 10,200 m3 h−1, in order to keep DO concentrations within 0.3–1.0 mg L−1.

The third WWTP (WWTP-3) is located in the metropolitan region of the city of Rio de Janeiro. This WWTP usually treats domestic wastewater of 470,000 PE, with an average daily raw wastewater flow of 6,100 m3 h−1 (Ribeiro et al. 2015). However, during the study period, WWTP-3 also received a contribution of a landfill leachate (only in the morning), corresponding to about 1% of the daily raw wastewater flow. This input of landfill leachate causes increases in the organic matter loading (organic shock loading). The process used at this WWTP is conventional activated sludge with non-BNR. The volume and surface area of the aeration tank (four aeration tanks in total) are 11,550 m3 and 2,175 m2, respectively. The HRT and SRT were 9 h and 10 days, respectively. The aeration tanks are divided into six interlinked aeration zones (ZNs), with different amounts of diffusers. The largest number of diffusers is installed in ZN 2 and ZN 5. The magnitude of the air flow rate is controlled by the number of diffusers per zone. The air flow rate per tank is 10,000 m3 h−1, with 3,350 and 1,980 m3 h−1 at ZN 2 and ZN 5, respectively.

Sampling and analysis

Measurements of N2O emissions at the WWTP-1 aeration tank were performed for 24 h on July 8th and 24th and August 9th, 2014, at the same spot, under different experimental aeration conditions (AC) (Figure 1(a)). Chronologically, the air flow rates were set to 135 m3 h−1 (AC 1), 205 m3 h−1 (AC 2) and 450 m3 h−1 (AC 3). The stabilization period for each experimental AC was two weeks. In all conditions, the pH and DO concentrations were logged every minute using an HI9828 multiparameter portable meter (Hanna).

Figure 1

Designs and sampling locations of the aeration tanks at (a) WWTP-1, (b) WWTP-2, (c) WWTP-3. The infrared (IR) N2O analyzer was linked to PVC gas collection floating chambers.

Figure 1

Designs and sampling locations of the aeration tanks at (a) WWTP-1, (b) WWTP-2, (c) WWTP-3. The infrared (IR) N2O analyzer was linked to PVC gas collection floating chambers.

Close modal

At WWTP-2, N2O emissions were continuously measured for 24 h on January 22nd, 2015 at a location close to the outlet end of the aeration tank near the continuous DO probe (Figure 1(b)). The aeration control system was automatically triggered in this aeration tank whenever DO dropped below 0.3 mg L−1, and was suspended automatically when DO concentrations reached levels higher than 1.0 mg L−1. During the study period, DO concentrations were logged every hour using a Luminescent DO sensor (Hach).

At WWTP-3, N2O emissions were measured at ZN 5 from 9 to 12 AM on June 30th, 2014 (Figure 1(c)). ZN 5 has the second highest number of diffusers, as well as air flow rate, and was chosen as a function of the DO levels provided by a previous study (Ribeiro et al. 2015). This zone exhibits the highest DO concentrations, since most of the organic matter is oxidized as the sewage flows between ZN 1 and ZN 4. N2O emission measurements performed only in the morning were due to input of landfill leachate which took place only at this time of the day. In addition, WWTP-3 is surrounded by marginalized populations, so no safe conditions for a 24-h study period were possible, unlike the other evaluated WWTPs. DO concentrations were logged every 10 seconds using an HI9828 multiparameter portable meter (Hanna).

A floating gas collection chamber (surface area = 0.05 m2) was used for N2O sampling with the off-gas from the headspace directed to an N2O/CO-23d infrared gas analyzer (at 0.5 L min−1) (Los Gatos Research Inc., USA), where the N2O concentration was logged each second (Figure 1). The infrared gas analyzer provides measurements with a precision of 0.05 ppb, sufficient for detecting even minor variations in ambient N2O concentrations. The chamber is made of polyvinyl chloride (PVC) equipped with a float. The N2O emission flux (F) was calculated by multiplying the ΔN2O concentration, which is the difference between the chamber headspace and the atmospheric N2O concentrations, by the emerging air flow rate (Qfloating chamber) and divided by floating chamber area (Afloating chamber) as displayed in Equation (1). The latter was measured at the surface of the aeration tanks, using the PVC floating chamber and a digital rotameter. The air flow rates for the entire aeration tanks (Qair) were extrapolated from the Qfloating chamber values. Analytical precision was ±1% and the quantification limit for the floating chamber technique was 0.05 mg N m−2 h−1.
formula
(1)
Liquid samples of raw and treated wastewaters from WWTP-1 and WWTP-2 were regularly collected, manually and with the aid of automatic refrigerated samplers, respectively, to determine the COD, total Kjeldahl N (TKN) and TN. In addition, dissolved inorganic nitrogen (, and ) was also determined in samples collected from the WWTP-2 and WWTP-3 aeration tanks (Table 1).
Table 1

Sampling strategy for the three studied WWTPs

SamplesWWTP-1WWTP-2WWTP-3
Raw wastewater (COD, TKN and TN)a (COD, TKN and TN)b – 
Aeration tank (N2O emission, DO and pH)c (N2O emission, DO and DIN)c (N2O emission, DO and DIN)d 
Treated wastewater (COD, TKN and TN)a (COD, TKN and TN)b – 
SamplesWWTP-1WWTP-2WWTP-3
Raw wastewater (COD, TKN and TN)a (COD, TKN and TN)b – 
Aeration tank (N2O emission, DO and pH)c (N2O emission, DO and DIN)c (N2O emission, DO and DIN)d 
Treated wastewater (COD, TKN and TN)a (COD, TKN and TN)b – 

aHourly sampling (between 8 AM and 1 PM) during the three experimental ACs.

bHourly sampling during the 24-h study period.

c24-h online sampling.

dSampling period from 9 to 12 AM.

All analytical methods followed standard APHA (2012) protocols. COD, TKN and TN were determined using the closed reflux colorimetric method, block digestion method and direct persulfate digestion method, respectively. concentrations were measured by an ion-selective electrode method using a Star 5 pH-meter (Orion). The and concentrations were determined by ion chromatography with chemical suppression of eluent conductivity, model 790 Personal IC (Metrohm). The limits of quantification were 0.03 mg L−1 for , 0.1 mg L−1 for and , and 10 mg N L−1 for TKN and TN. The analytical precisions for the analyses, performed in triplicate, were within ±5%.

WWTP-1: COD, TKN and TN removal efficiencies

The influent COD and TKN loads, measured from 8 AM to 1 PM, varied from 12 to 42 kg h−1 and 1.2 to 4.2 kg h−1, respectively, with a sharp increase from 8 to 9 AM and little subsequent variability (Table 2). The hourly values were similar in all of the three experimental ACs (AC 1, AC 2 and AC 3) (Figure 2). The sharp increase of organic load observed is related to the dynamics of the diurnal pattern of raw wastewater entering WWTP-1's aeration tank, as described previously. Despite the large variation applied to the air flow rate, COD removal efficiencies were high and similar in the three experimental ACs. Conversely, TN removal efficiencies were unsurprisingly low, since WWTP-1 was not designed for N removal. However, TKN (the fraction of reduced N converted to oxidized N forms) removal efficiencies in AC 2 and AC 3 were much higher than in AC 1. Despite the lack of and data for the treated effluent, TKN removal probably occurred by nitrification, suggested by the significant drop in pH as aeration rates increased (Table 2). The first nitrification stage releases two H+ per mol of oxidized (Wrage et al. 2001; von Sperling 2002). Thus, the variation of the air flow rate in the study range did not significantly affect COD removal efficiency, although significant variability of the removal efficiencies of TKN were observed.

Table 2

Aeration tank air and influent wastewater flow rates, aeration tank pH, COD and TKN mass loading rates, and COD, TKN and TN removal efficiencies for the three ACs evaluated during the 24-h sampling period

ACQair (m3 h1)Qwastewater (m3 h1)pHLoad
Removal
COD (kg h1)TKN (kg h1)COD (%)TKN (%)TN (%)
AC 1 135 44 6–7 12–42 1.2–4.2 90 27 12 
AC 2 205 92 82 15 
AC 3 450 4–5 95 95 20 
ACQair (m3 h1)Qwastewater (m3 h1)pHLoad
Removal
COD (kg h1)TKN (kg h1)COD (%)TKN (%)TN (%)
AC 1 135 44 6–7 12–42 1.2–4.2 90 27 12 
AC 2 205 92 82 15 
AC 3 450 4–5 95 95 20 
Figure 2

Influent COD and TKN loadings measured during from 8 AM to 1 PM at the WWTP-1 aeration tank. Data are displayed as means and standard deviations for the three experimental ACs.

Figure 2

Influent COD and TKN loadings measured during from 8 AM to 1 PM at the WWTP-1 aeration tank. Data are displayed as means and standard deviations for the three experimental ACs.

Close modal

WWTP-1: N2O emissions

Figure 3 shows the magnitude of the N2O emissions and DO concentrations under different aeration rates. In AC 1 (Qair = 135 m3 h−1), N2O emissions were negligible (<0.09 mg N m−2 h−1) over the 24-h period (Figure 3(a)). In this experimental AC, DO concentrations (DO < 0.6 mg L−1) (Figure 3(a)) and TKN removal efficiency were low (27%). Low DO concentrations in the aeration tank were probably not sufficient to efficiently oxidize to , leading to low TKN removal and undetectable N2O emissions. These findings are in agreement with other studies (Ahn et al. 2010; Yu et al. 2010; Aboobakar et al. 2013; Rodriguez-Caballero et al. 2014). Ruiz et al. (2003), in a laboratory-scale activated sludge reactor operated only with nitrification process, observed accumulation due to nitrification inhibition when DO levels reached values below 0.5 mg L−1.

Figure 3

24-hour N2O emission fluxes and DO concentrations in the aeration tank of the WWTP-1 under different ACs: (a) AC 1 (Qair = 135 m3 h−1); (b) AC 2 (Qair = 205 m3 h−1); (c) AC 3 (Qair = 450 m3 h−1). Sampling dates: AC 1 = July 08, 2014, AC 2 = July 24, 2014 and AC 3 = August 09, 2014.

Figure 3

24-hour N2O emission fluxes and DO concentrations in the aeration tank of the WWTP-1 under different ACs: (a) AC 1 (Qair = 135 m3 h−1); (b) AC 2 (Qair = 205 m3 h−1); (c) AC 3 (Qair = 450 m3 h−1). Sampling dates: AC 1 = July 08, 2014, AC 2 = July 24, 2014 and AC 3 = August 09, 2014.

Close modal

In AC 2 (Qair = 205 m3 h−1), N2O emissions were very low (<9 mg N m−2 h−1) during the first 10 h of the 24-h measurement, since no raw wastewater entered the WWTP-1 aeration tank until 8 AM (Figure 3(b)). Thereafter, two N2O peaks, the lowest around 12 AM and the highest around 6 PM, occurred. Between the occurrence of these peaks (from 2 to 4 PM), no raw wastewater entered the aeration tank. In AC 2, DO concentrations were fairly constant over the 24-h period, varying from 1 to 3 mg L−1 (Figure 3(b)). During late morning, the response of N2O emissions (between 10 AM and 1 PM) to influent TKN, which rose sharply between 8 and 9 AM (Figure 2), was smaller compared to those observed in the late afternoon (between 4 and 7 PM). Lotito et al. (2012), in an experiment run in a pilot reactor with sludge activated process, reported higher N2O emissions alongside higher TKN loads. Unlike AC 1, DO concentrations in AC 2 were probably sufficient to oxidize organic matter and promote nitrification, as indicated by the 82% TKN removal (Table 2). Ruiz et al. (2003) found complete oxidation of to when DO concentrations were approximately 2 mg L−1 or above in a laboratory-scale activated sludge reactor operated only with nitrification processes.

In AC3, between 9 to 10 AM, N2O emission rates rose promptly in response to the rapid increase of influent TKN (associated with raw wastewater) which took place between 8 and 9 AM (Figure 3(c)). This suggests that the DO concentrations were enough to simultaneously result in the oxidation of the organic matter and lead to nitrification. DO concentrations varied from 3 to 8 mg L−1, with two pronounced peaks preceding N2O peaks during 1 h (Figure 3(c)), with rapid DO consumption during periods of raw wastewater input (8 AM to 2 PM and 4 to 6 PM). The input of raw wastewater with reduced N compounds coupled to the higher aeration rate is the most plausible explanation for the sudden elevated N2O emissions observed from 9 to 10 AM. de Mello et al. (2013) reported very similar behavior of steep N2O emissions as soon as the aeration period (60-min) began functioning in an activated sludge WWTP with intermittent aeration system.

Following the sudden increase, N2O emissions decreased exponentially until 1 PM, even with continuous raw wastewater loading conditions and constant aeration rate. This behaviour could be attributed to the changing conditions of the process, such as shock loads. Significant alterations in process conditions, such as sudden inputs of organic loading (organic shock loading) following a period with no raw wastewater discharge, can lead to increased N2O emission. However, the adaptation of bacterial metabolism requires time and would probably result in decreased N2O emissions. Kampschreur et al. (2009) suggested that rapidly changing operational conditions could result in N2O emission; however, the bacterial populations subjected to repeatedly varying operational conditions can reduce N2O emissions by adapting. Future studies should be performed to better understand the relationship between N2O emissions and the adaptation of bacterial metabolism during rapidly changing process conditions.

In AC 3, as in AC 2, two N2O peaks were observed (Figure 3(c)). However, they differ markedly in magnitude, especially the morning peak. During this period (8 AM to 1 PM), the maximum N2O fluxes were 34 mg N m−2 h−1 and 391 mg N m−2 h−1 for AC 2 and AC 3, respectively. In addition, the increasing aeration rate, from 205 m3 h−1 (AC 2) to 450 (AC 3) m3 h−1, resulted in a two-fold increase of N2O emission flux (the 24-h integrated). The 24-h integrated N2O emission fluxes were, respectively, of 32 and 66 mg N m−2 h−1. These outcomes demonstrate the negative effect of over-aeration in the overall N2O ERs of a WWTP, as highlighted by Castro-Barros et al. (2015).

The overall experiment consisted of varying the air flow rate in the aeration tank of an activated sludge WWTP operating a non-BNR system, with the aim of using N2O emission responses for operational adjustments regarding the adequacy of the AC. AC 2 was shown to be the best operating condition for WWTP-1, adequate for achieving both organic matter oxidation processes and nitrification. This AC provided adequate performance in relation to favourable COD (92%) and TKN (82%) removal efficiencies, with low N2O emissions when compared to those found in AC 3. However, in AC 2, the N2O EF based on PE was 19 g N2O person−1 year−1, 6-fold higher than the limit proposed by IPCC (2006).

Prior to this study, WWTP-1 operated with an air aeration rate similar to that in AC 3. Currently, it applies the same air flow rate as applied in AC 2, maintaining DO concentrations in the range of 1 to 3 mg L−1. However, in order to reduce effluent TN levels, the design and operation of this plant must be altered, i.e. the establishment of an anoxic zone, and internal recirculation systems must be added, to carry out the denitrification process, which is the key process for the effective removal of TN as N2 (complete denitrification) and reducing N2O emissions (Foley et al. 2010). This is the first study to demonstrate the actual application of a 24-h online sampling strategy to determine short-term N2O emissions in order to improve the operating condition of a full-scale system, with major focus on the aeration rate, an important factor in the process performance, operational and maintenance costs, and mitigation of N2O emissions.

WWTP-2: COD, TKN and TN removal efficiencies

The 24-h study demonstrated that the influent COD and TN loads began rising in the early morning, with two major peaks observed, one at noon and the other in the early evening (Figure 4(a) and 4(b)). The influent COD and TN loads varied within the range of 318 to 1,964 kg h−1 and 28 to 259 kg h−1, respectively. This 24-h variability did not affect COD, TN and TKN removals, whose mean efficiencies (hourly range values) were 90% (80–96%), 83% (60–94%) (Figure 4(c)) and 86% (61–97%), respectively. These results were expected, given that WWTP-2 operates with BNR by SND coupled to an automatic aeration system controlled by DO concentrations. This system is set up to maintain DO levels in the range of 0.3 to 1.0 mg L−1, even during sudden organic loading variations.

Figure 4

WWTP-2 (a) influent COD load; (b) influent TN load and (c) COD and TN removal efficiencies (%).

Figure 4

WWTP-2 (a) influent COD load; (b) influent TN load and (c) COD and TN removal efficiencies (%).

Close modal

WWTP-2: N2O emissions

Figure 5(a) displays two major accumulation periods, from 1.5 to 4.3 mg N L−1 (between 3 and 5 AM) and from 2.5 to 7.5 mg N L−1 (between 8 and 10 AM), near the outlet end of the aeration tank operated with a 16-h HRT. These periods were associated with higher influent TN loads (Figure 4(b)). In parallel, and, especially, concentrations were low (Figure 5(b) and 5(c)).

Figure 5

N2O emission fluxes (mg N m−2 h−1) and (a) ; (b) ; (c) and (d) DO concentrations measured in the WWTP-2 aeration tank.

Figure 5

N2O emission fluxes (mg N m−2 h−1) and (a) ; (b) ; (c) and (d) DO concentrations measured in the WWTP-2 aeration tank.

Close modal

From 5 to 8 AM, decreased from 4.3 to 2.5 mg N L−1 (40% reduction). Simultaneously, a slight increase in N2O fluxes was observed from 0.5 to 1.2 mg N m−2 h−1. A different profile was observed from 10 AM to 7 PM, when decreased from 7.5 to 1.9 mg N L−1 (75% reduction) with simultaneous increases of (ranging from 0 to 0.3 mg N L−1) and (ranging from 0.7 to 4.2 mg N L−1) concentrations. The N2O fluxes during this period increased from 1.2 to 8.7 mg N m−2 h−1, reaching their maximum during the maximum and concentrations (Figure 5(a)5(c)). Daelman et al. (2015) reported that and peaks were positively correlated to N2O emissions from the aeration tank of a full-scale WWTP. After this period (between 7 and 8 PM), and concentrations decreased in response to the limited DO levels (≤0.2 mg L−1) (Figure 5(d)), which can indicate denitrifying activity. In addition, N2O emission fluxes began to decrease even though reduction of oxidized N forms ( and ) was still ensuing (Figure 5).

In summary, the highest N2O emissions occurred when DO concentrations were very low (≤0.3 mg L−1) (Figure 5(d)) and were correlated to and peaks (Figure 5(b) and 5(c)). Previous studies have reported the effect of low DO concentrations (<1 mg L−1) on N2O emissions during nitrification (Tallec et al. 2006; Wunderlin et al. 2012; Aboobakar et al. 2013). Aboobakar et al. (2013) reported a negative and direct correlation between DO concentrations and N2O emissions at a full scale WWTP, with the highest N2O emissions occurring during lower DO ranges (<1 mg L−1). Tallec et al. (2006) found higher N2O emissions at low DO levels (from 0.1 to 1 mg L−1) from a laboratory-scale batch experiments under various oxygenation conditions. Clearly, these outcomes indicate that, even at low DO concentrations, the nitrification process was the main driving force behind the highest N2O emissions from the WWTP-2 aeration tank, as reported by other authors evaluating full-scale WWTPs (Ahn et al. 2010; Aboobakar et al. 2013; Rodriguez-Caballero et al. 2014).

In contrast to WWTP-2, the AC 1 experiment at WWTP-1 showed negligible N2O emissions (<0.09 mg N m−2 h−1) with low DO concentrations (<0.6 mg L−1) and within the range found for WWTP-2 (between 0.3 and 1 mg L−1). However, WWTP-1 operated with low TKN removal efficiency (27%) (Table 2), unlike WWTP-2 (61–97%). Thus, WWTP-1 requires DO concentrations within the range of 1 and 3 mg L−1 to combine both oxidation of organic matter and nitrification processes, as described previously (AC 2). These results could be explained by microorganism adaptation to WWTP-2 operating conditions (limited DO concentrations). Microorganism adaptation to limited DO concentrations has been reported in the literature (Belluci et al. 2011; Wunderlin et al. 2012). Belluci et al. (2011) achieved complete nitrification in a laboratory-scale reactor at DO concentrations as low as 0.5 mg L−1, where high performance resulted from elevated ammonia oxidation bacteria (AOB) levels in the reactor. Thus, WWTP-1 under AC 1 would need to alter certain operating conditions, combining reduced SRT and extended HRT, to favour the AOB nitrifying process. In addition, the denitrification process should be conducted, combined not only with TKN but also TN (as N2), since WWTP-1 was not designed to remove N and was thus subject to elevated N2O emissions (Foley et al. 2010).

Figure 6 displays the N2O fluxes from the WWTP-1 aeration tanks during AC 2 (the best operating condition for the presence of the nitrification process alone) and for WWTP-2 (SND processes) throughout the 24-h study period. In the case of WWTP-2, DO concentrations showed no abrupt change throughout the 24-h monitoring period due to the DO control applied, maintaining DO concentrations within the range of 0.1 to 1.2 mg L−1 (Figure 5(d)), despite COD and TN load variations (Figure 4(a) and 4(b)). Thus, controlling DO concentrations may be the key to mitigating N2O emissions. Dynamic changes in DO concentrations are reported in the literature as being responsible for N2O emission peaks (Yu et al. 2010; Aboobakar et al. 2013; Rodriguez-Caballero et al. 2014).

Figure 6

N2O fluxes in the WWTP-1 aeration tanks during AC 2 (the best operating condition) and in WWTP-2 during the 24-h sampling period.

Figure 6

N2O fluxes in the WWTP-1 aeration tanks during AC 2 (the best operating condition) and in WWTP-2 during the 24-h sampling period.

Close modal

Another controlling factor for low N2O emissions could be associated to higher TN removal efficiencies, as observed for the WWTP-2 (Figure 4(c)) and, consequently, lower effluent TN concentrations (<10 mg N L−1). Foley et al. (2010) reported that a WWTP designed and operated to achieve low effluent TN concentrations (<10 mg N L−1) could result in relatively low N2O emissions as a result of the denitrification process. These outcomes demonstrate the importance of controlling the decreases of TN levels and N2O emissions. These results also clearly demonstrated that non-BNR WWTPs show high risks for N2O emissions, unlike the BNR WWTPs with controlled nitrification and denitrification processes. Therefore, the findings at WWTP-1 and WWTP-2 indicate that N2O emission measurements can be used as indicators of TN removal efficiency, as well as applied for the identification of peaking influent COD and TN loads, with the main goal of DO control.

WWTP-3: , and profiles

Figure 7 displays the concentration patterns of DIN forms (, and ) along the six WWTP-3 aeration tank zones on June 30th, 2014, before and after the input of landfill leachate. The concentration patterns during both periods were similar, with the highest concentrations present in the first four zones, followed by a sharp decrease in the last two (Figure 7(a)). As proposed by Ribeiro et al. (2015), DO drives the oxidation of organic matter in the first four zones and therefore nitrification takes place only in the last two.

Figure 7

Spatial variability of (a) , (b) , (c) concentrations throughout the six aeration tank zones on June 30th, 2014; without and with landfill leachate input.

Figure 7

Spatial variability of (a) , (b) , (c) concentrations throughout the six aeration tank zones on June 30th, 2014; without and with landfill leachate input.

Close modal

During the period without any landfill leachate input, decreases in concentrations from 5.5 to 0.3 mg N L−1 and increasing concentrations from 0.1 to 4 mg N L−1 were observed throughout the six aeration tank zones (ZN1-ZN6), with negligible concentrations (Figure 7). The degradation rate was of 3.3 mg N L−1 h−1, and and production rates were 0.2 and 2.5 mg N L−1 h−1, respectively. These results indicate that complete nitrification was achieved, since 75% of the was converted into , with no significant denitrifying activity. A similar profile was observed during the landfill leachate input, when decreased from 29 to 9.2 mg N L−1 with increasing values (ranging from 1 to 9.1 mg N L−1) throughout the six aeration tank zones. In contrast, an accumulation from 0.3 to 1.5 mg L−1 in the last two zones occurred (ZN4-ZN6) (Figure 7(b)). The degradation rate was of 13 mg N L−1 h−1, and the and production rates were 0.9 and 5.2 mg N L−1 h−1, respectively. Around 40% of the was converted into , indicating probable simultaneous nitrifying and denitrifying activities (N loss as N2). These results indicate that organic shock loading (input of landfill leachate) is responsible for the accumulation of , with a loss of conversion to (from 75 to 40%). These results clearly demonstrate that landfill leachate input is an important factor influencing the efficiency of complete nitrification.

WWTP-3: N2O emissions

N2O emission monitoring was carried out only at ZN 5, due to the higher available DO concentrations and the evolution of the nitrification process in ZN4-ZN6 of the aeration tank, as described previously (Figure 7) and reported by Ribeiro et al. (2015). Figure 8 displays the variation of N2O fluxes and DO concentrations at ZN 5 during landfill leachate input. During the first 90 min of the study period (from 9 to 10:30 AM), the N2O emissions fluxes varied from 147 to 561 mg N m−2 h−1 (Figure 8) and DO concentrations varied from 2 to 6 mg L−1 at ZN 5 (Figure 8) characterizing complete nitrification, as described previously. Ribeiro et al. (2015) found a similar N2O flux (250 mg N m−2 h−1) at the same plant, linked to higher aeration rates and DO concentrations (between 2 and 4 mg L−1), favouring complete nitrification. Thus, higher DO concentrations (DO ≥ 2 mg L−1) seem to favour both oxidation of organic matter and complete nitrification in the WWTP-3 aeration tank, as indicated by Ribeiro et al. (2015).

Figure 8

3-h (from 9 to 12 AM) measurements of N2O emission fluxes and DO concentrations at ZN 5 during periods before and after landfill leachate input.

Figure 8

3-h (from 9 to 12 AM) measurements of N2O emission fluxes and DO concentrations at ZN 5 during periods before and after landfill leachate input.

Close modal

During the first 90 min, the maximum N2O emission flux was higher than those measured for other studied WWTPs (WWTP-1 and WWTP-2). Especially for WWTP-1 that operates with non-BNR as WWTP-3, the difference observed in the magnitude of the N2O flux can be associated to air flow rate (Qair) applied in relation to wastewater flow rate (Qwastewater). The Qair to Qwastewater ratio for WWTP-3 was almost two times higher than those for AC 1 and AC 2 at WWTP-1. Conversely, although the Qair to Qwastewater ratio was 1.5 times higher than for WWTP-3, the maximum N2O flux values in AC 3 were close to those observed at WWTP-3. Therefore, WWTP-3 operates with over-aeration, that correlates strongly with N2O emissions, since the excessive air flow rate intensifies N2O transfer from liquor to the atmosphere by air stripping. Thus, the Qair to Qwastewater ratio must be adjusted with the input of landfill leachate.

The last 90 min of the study period (from 10:30 to 12 AM) reflect landfill leachate input. N2O emissions begin to increase and reached a maximum of 1,890 mg N m−2 h−1 (Figure 8). The maximum emission value was three times higher than that observed during the previous period (without landfill leachate). This higher N2O emission is probably due to accumulation (Figure 7(b)). It could be argued that the organic shock loading caused by the landfill leachate input was responsible for reducing DO concentrations from 4 to 0.5 mg L−1 (Figure 8). These results demonstrate that the addition of landfill leachate to the regular treatment of domestic wastewater favored greater N2O production and emission. It is reasoned that higher organic matter content in the landfill leachate also contributes to decreased DO concentrations, which result in accumulation (partial nitrification). Rodriguez-Caballero et al. (2013) reported that, under partial nitrification conditions, higher N2O emissions occur as compared to complete nitrification conditions. In addition, sudden changes, such as increasing loadings, could also lead to peaks and higher N2O emissions, as reported by other authors (Foley et al. 2010; Rodriguez-Caballero et al. 2013; Toor et al. 2015). Therefore, the continuous monitoring of N2O emissions seems to be a good indicator to evaluate nitrification efficiency. In addition, since higher N2O emissions from the aeration tank are correlated with increasing concentrations, they can provide early warning signs for complete nitrification failure, as reported by Burgess et al. (2003) and Butler et al. (2009).

In short, the operation of WWTP-3 must be altered, in order to minimize N2O emissions, since there is a propensity for WWTPs in developing countries to receive landfill leachate into their wastewater systems. Organic shock loading should be avoided through influent equalization during landfill leachate input.

The continuous measurement of N2O emissions in activated sludge WWTPs operated with BNR and non-BNR was evaluated and the main conclusions are as follows:

  • Nitrification process is the main driving force behind N2O emission peaks.

  • Variation of air flow rates influence significantly N2O emissions; high emissions denote over-ACs or incomplete nitrification, with accumulation of concentrations.

  • Continuous measurements of N2O emissions can provide information on aeration adequacy and the efficiency of complete nitrification.

  • Adequate control of DO concentrations is a key factor to mitigate N2O emissions.

  • Sudden input of organic loads (e.g. due to landfill leachate input) to the aeration tank of an activated sludge process can lead to substantial increase of N2O emissions.

  • Non-BNR WWTPs are subject to high N2O emissions, unlike the BNR WWTPs, with controlled nitrification and denitrification processes.

Aboobakar
,
A.
,
Cartmell
,
E.
,
Stephenson
,
T.
,
Jones
,
M.
,
Vale
,
P.
&
Dotro
,
G.
2013
Nitrous oxide emissions and dissolved oxygen profiling in a full-scale nitrifying activated sludge treatment plant
.
Water Res.
47
,
524
534
.
Ahn
,
J. O.
,
Kim
,
S.
,
Park
,
H.
,
Rahm
,
B.
,
Pagilla
,
K.
&
Chandran
,
K.
2010
N2o emissions from activated sludge processes, 2008–2009: results of a national monitoring survey in the United States
.
Environ. Sci. Technol.
44
,
4505
4511
.
APHA
2012
Standard Methods for the Examination of Water and Wastewater
, 22nd edn.
American Public Health Association
,
Washington, DC, USA
.
Belluci
,
M.
,
Ofiteru
,
I. D.
,
Graham
,
D. W.
,
Head
,
I. M.
&
Curtis
,
T. P.
2011
Low-dissolved-oxygen nitrifying systems exploit ammonia-oxidizing bacteria with unusually high yields
.
Appl. Environ. Microbiol.
77
(
21
),
7787
7796
.
Brotto
,
A. C.
,
de Mello
,
W. Z.
,
Kligerman
,
D. C.
&
Piccoli
,
A. S.
2010
Nitrous oxide emissions from activated sludge wastewater treatment plant with prolonged aeration process
.
Quim. Nova
33
,
618
623
(in Portuguese)
.
Brotto
,
A. C.
,
Kligerman
,
D. C.
,
Andrade
,
S. A.
,
Ribeiro
,
R. P.
,
Oliveira
,
J. L. M.
,
Chandran
,
K.
&
de Mello
,
W. Z.
2015
Factors controlling nitrous oxide emissions from a full-scale activated sludge system in the tropics
.
Environ. Sci. Pollut. Res. Int.
22
,
11840
11849
.
Burgess
,
J. E.
,
Stuetz
,
R. M.
,
Morton
,
S.
&
Stephenson
,
T.
2003
Dinitrogen oxide detection for process failure early warning systems
.
Water Sci. Tech.
45
(
4–5
),
247
254
.
Butler
,
M. D.
,
Wang
,
Y. Y.
,
Cartmell
,
E.
&
Stephenson
,
T.
2009
Nitrous oxide emissions for early warning of biological nitrification failure in activated sludge
.
Water Res.
43
,
1265
1272
.
Castro-Barros
,
C. M.
,
Daelman
,
M. R. J.
,
Mampaey
,
K. E.
,
van Loosdrecht
,
M. C. M.
&
Volcke
,
E. I. P.
2015
Effect of aeration regime on N2O emission from partial nitritation-anammox in a full-scale granular sludge reactor
.
Water Res.
68
,
793
803
.
Czepiel
,
P.
,
Crill
,
P.
&
Harriss
,
R.
1995
Nitrous oxide emissions from municipal wastewater treatment
.
Environ. Sci. Technol.
29
,
2352
2356
.
Daelman
,
M. R. J.
,
de Baets
,
B.
,
van Loosdrecht
,
M. C.
&
Volcke
,
E. I.
2013
Influence of sampling strategies on the estimated nitrous oxide emission from wastewater treatment plants
.
Water Res.
47
,
3120
3130
.
Daelman
,
M. R. J.
,
van Voorthuizen
,
E. M.
,
van Dongen
,
U. G. J. M.
,
Volcke
,
E. I. P.
&
van Loosdrecht
,
M. C. M.
2015
Seasonal and diurnal variability of N2O emissions from a full-scale municipal wastewater treatment plant
.
Sci. Total Environ.
536
,
1
11
.
de Mello
,
W. Z.
,
Ribeiro
,
R. P.
,
Brotto
,
A. C.
,
Kligerman
,
D. C.
,
Piccoli
,
A. S.
&
Oliveira
,
J. L. M.
2013
Nitrous oxide emission from an intermittent aeration activated sludge system of an urban wastewater treatment plant
.
Quim. Nova
36
,
16
20
.
Intergovernmental Panel on Climate Change, IPCC
2006
IPCC Guidelines for National Greenhouse Gas Inventories, v. 5, Kanagawa
.
Intergovernmental Panel on Climate Change, IPCC
2013
Climate change 2013: the physical science basis
. In:
Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
(
Stocker
,
T. F.
,
Qin
,
D.
,
Plattner
,
G.-K.
,
Tignor
,
M.
,
Allen
,
S. K.
,
Boschung
,
J.
,
Nauels
,
A.
,
Xia
,
Y.
,
Bex
,
V.
&
Midley
,
P. M.
, eds).
Cambridge University Press
,
New York, NY
,
USA
.
Kampschreur
,
M. J.
,
Temmink
,
H.
,
Kleerebezem
,
R.
,
Jetten
,
M. S. M.
&
van Loosdrecht
,
M. C. M.
2009
Nitrous oxide emission during wastewater treatment
.
Water Res.
43
,
4093
4103
.
Law
,
Y.
,
Ye
,
L.
,
Pan
,
Y.
&
Yuan
,
Z.
2012
Nitrous oxide emissions from wastewater treatment processes
.
Philos. T Roy. Soc. B.
367
,
1265
1277
.
Lotito
,
A. M.
,
Wunderlin
,
P.
,
Joss
,
A.
,
Kipf
,
M.
&
Siegrist
,
H.
2012
Nitrous oxide emissions from the oxidation tank of a pilot activated sludge plant
.
Water Res.
46
,
3563
3573
.
Ravishankara
,
A. R.
,
Daniel
,
J. S.
&
Portman
,
R. W.
2009
Nitrous oxide (N2O): the dominant ozone depleting substance emitted in the 21st century
.
Science
326
,
123
125
.
Ribeiro
,
R. P.
,
de Mello
,
W. Z.
,
Kligerman
,
D. C.
,
Oliveira
,
J. L. M.
,
Alvim
,
R. B.
,
Andrade
,
A. S.
,
Brotto
,
A. C.
&
de Almeida
,
P. A.
2013
Emissões de óxido nitroso do tanque de aeração de uma estação de tratamento de esgotos com sistema de lodos ativados convencional
.
Quim. Nova
36
,
998
1003
(in Portuguese)
.
Ribeiro
,
R. P.
,
Alvim
,
R. B.
,
Andrade
,
S. A.
,
Kligerman
,
D. C.
,
de Almeida
,
P. A.
,
Oliveira
,
J. L. M.
&
de Mello
,
W. Z.
2015
Controlling factors of nitrous oxide emissions from a conventional activated sludge wastewater treatment plant
.
Engevista
17
(
3
),
375
384
.
Rodriguez-Caballero
,
A.
,
Ribera
,
A.
,
Balcázar
,
J. L.
&
Pijuan
,
M.
2013
Nitritation versus full nitrification of ammonium-rich wastewater: comparison in terms of nitrous and nitric oxides emissions
.
Bioresour. Technol.
139
,
195
202
.
Rodriguez-Caballero
,
A.
,
Aymerich
,
I.
,
Poch
,
M.
&
Pijuan
,
M.
2014
Evaluation of process conditions triggering emissions of green-house gases from a biological wastewater treatment system
.
Sci. Total Environ.
493
,
384
391
.
Strokal
,
M.
&
Kroeze
,
C.
2014
Nitrous oxide (N2O) emissions from human waste in 1970–2050
.
Curr. Opin. Environ. Sustain.
9–10
,
108
121
.
von Sperling
,
M.
2002
Activated Sludge: Principles of the Biological Treatment of Sewage
.
Universidade Federal de Minas Gerais
,
Belo Horizonte, Brazil
(in Portuguese)
.
Wrage
,
N.
,
Velthof
,
G. L.
,
van Beusichem
,
M. L.
&
Oenema
,
O.
2001
Role of nitrifier denitrification in the production of nitrous oxide
.
Soil. Biol. Biochem.
33
,
1723
1732
.
Wunderlin
,
P.
,
Mohn
,
J.
,
Joss
,
A.
,
Emmenegger
,
L.
&
Siegrist
,
H.
2012
Mechanisms of N2O production in biological wastewater treatment under nitrifying and denitrifying conditions
.
Water Res.
46
(
4
),
1027
1037
.
Yu
,
R.
,
Kampschreur
,
M. J.
,
van Loosdrecht
,
M. C. M.
&
Chandran
,
K.
2010
Mechanisms and specific directionality of autotrophic nitrous oxide and nitric oxide generation during transient anoxia
.
Environ. Sci. Technol.
44
(
4
),
1313
1319
.