A single sequencing batch reactor (SBR) with an operating mode of anaerobic/oxic/anoxic (A/O/A) was developed to determine a simpler process to recover nitrous oxide (N2O) from synthetic wastewater containing ammonia and glucose. This SBR system was initiated in A/O mode to implement nitritation (ammonia to nitrite) and then switched to A/O/A mode. Using measurements of the dissolved N2O concentration and release rate, the total production and conversion rate of N2O were calculated to reveal the potential of producing and recovering N2O in the extended anoxic phase. Results showed that the A/O/A SBR could convert the majority of the nitrite available in the system into N2O by heterotrophic denitritation over longer anoxic periods, and a conversion rate of 77% could be achieved. As a consequence, the A/O/A SBR presents potential ability to produce and recover N2O from wastewater containing ammonia and organic carbon.
Nitrous oxide (N2O), a potent greenhouse gas, can be emitted during wastewater treatment and significantly contributes to the greenhouse gas footprint (IPCC 2007). N2O has been demonstrated to be produced and emitted in substantial amounts during nitrogen removal. A report provided by the United States Environmental Protection Agency revealed that the amount of N2O emitted from wastewater treatment plants is 3% of the total global emissions of N2O (USEPA 2006). In China, the amount of N2O emitted from wastewater treatment plants is about 1,260 t(N) per year (Wang et al. 2014b). In recent years, a large amount of research has focused on the emission and control of N2O during biological wastewater treatment (Kampschreur et al. 2009).
During biological wastewater treatment, processes such as autotrophic nitrification and heterotrophic denitrification, have been identified as major sources of N2O (Wang et al. 2014a, b). Heterotrophic denitrification is well-known to play an important role in N2O production from nitritation (ammonia to nitrite) reactors treating anaerobic sludge digestion liquor (Wang et al. 2014a). N2O production and emission increase when oxygen is limited (Wunderlin et al. 2012), under low organic carbon concentrations (Schalk-Otte et al. 2000; Itokawa et al. 2001) and when the intracellular organic carbon PHB (poly-β-hydroxybutyrate) is used as the carbon sources for denitrification (Beun et al. 2000; Schneider et al. 2011; Jia et al. 2012).
Despite the environmental risks it produces, N2O is also a type of resource that can be used as a powerful oxidant in combustion reactions in many situations (Scherson et al. 2013). If N2O is recovered from biological sewage treatment plants, the release of N2O to the atmosphere will be mitigated and the energy utilization efficiency of the wastewater treatment plant can be improved.
A N2O recovery process called coupled aerobic–anoxic nitrous decomposition operation (CANDO) was reported to recover N2O successfully through PHB denitrification (Scherson et al. 2013, 2014). The CANDO process is a two-stage sequencing batch reactor (SBR) process. Ammonia is converted into nitrite in the first SBR, and N2O is recovered by PHB denitrification in the second SBR. N2O can be used to improve combustion of methane, a product of anaerobic sludge digestion in biological wastewater treatment plants. In the second SBR, the conversion rates of nitrite to N2O are 65% and 75–80% with the carbon sources of acetate and sludge hydrolysis-acidification liquor, respectively. Using N2O to improve combustion, the output power of biogas power generation can increase by 5.7–7.3% (Scherson et al. 2013, 2014).
In China, sludge outputs from domestic wastewater treatment plants exceed 30 million tons per year. Over 50 plants use biogas produced in anaerobic sludge digestion for power generation or combustion. Recovery of N2O from sludge liquor, digestion liquor and wastewater containing chemical oxygen demand (COD) and ammonia to improve biogas combustion is thus an interesting endeavour. In this paper, an anaerobic/oxic/anoxic (A/O/A) SBR process was developed to study the potential of N2O recovery from synthetic wastewater containing COD and ammonia based on previous investigations of the impacts of heterotrophic denitrification on nitritation. In contrast to the CANDO system proposed by Scherson et al. (2013, 2014), the proposed system allows N2O recovery from a single SBR through nitritation and denitritation. The results obtained in this study help to improve the current understanding of the impacts of heterotrophic denitrification on N2O production and further the development of a novel process for N2O recovery from wastewater containing ammonia and COD.
MATERIALS AND METHODS
The laboratory-scale SBR had a height of 35 cm, inside diameter of 16 cm and working volume of 5,000 cm3 and was operated in the laboratory. Compressed air was supplied to the reactor during the aerobic phase and the air-flow rate was adjusted to 40 L/h. To select a suitable aeration flow rate, rates of 20, 40, 60 and 80 L/h were tested. It was found that lower aeration flow rate correlated with higher nitrite accumulation ratio and longer nitritation duration, whereas higher aeration flow rate correlated with lower nitrite accumulation ratio rate and shorter nitritation duration. Therefore, a moderate aeration rate of 40 L/h was adopted in this study. The contents of the reactor were mixed using a mechanical stirrer at 60 rpm in all phases except for the settling and decanting phases. The mixed liquor temperature was controlled using a water jacket to 27 ± 1 °C. The influent and effluent were pumped and controlled by a microcomputer-controlled system.
The synthetic wastewater was composed of (per litre): 0.500 g C6H12O6, 1.354 g NH4HCO3, 0.330 g KH2PO4, 0.050 g MgSO4·7H2O, 0.016 g CaCl2, 0.1500 g NaHCO3 and 1 mL of a trace element stock solution (Lovley & Phillips 1988) to achieve influent COD and NH4+-N concentrations of 530 mg/L and 240 mg/L, respectively.
The SBR was operated over one cycle of 8 h consisting of 5 min of anaerobic feeding, 5 h of anaerobic stirring, about 2 h of aerating (controlled by dissolved oxygen (DO) level) and 55 min of settling and decanting. No excess sludge discharged in the start-up period. In each cycle, 2 L of synthetic wastewater was fed to the reactor, resulting in a hydraulic retention time of 20 h.
The reactor was seeded with sludge from a domestic wastewater treatment plant in Xi'an, China. The preliminary concentration of the mixed-liquor suspended solids (MLSS) was 2,800 mg/L. After 20 d of nitrifying culture, 95% of the NH4+ was converted into NO2−; at this point, successful completion of startup was achieved.
Operating mode for N2O recovery
After completion of SBR start-up, 4 months of stable nitritation operation was carried out. The MLSS concentration reached 5,470 ± 28 mg/L. The operation mode of the SBR was then changed to N2O recovery mode with one cycle of 24 h consisting of 5 min of anaerobic feeding, 3 h of anaerobic stirring, about 2 h of aerating (controlled by DO level), 18 h of anoxic stirring, 5 min of aerating to release N2O and 50 min of settling and decanting. An extended anoxic period was designed to study the potential of N2O production and recovery in the new operating mode.
Ammonium (NH4+-N), nitrite (NO2−-N), nitrate (NO3−-N) and MLSS were analysed according to Standard Methods (APHA 1998). The pH in the reactor was measured with a pH meter (pH S10, Sichuan, China). DO concentrations in the reactor were monitored with a DO meter (Hach-HQ30d, USA). N2O concentrations in the liquid phase were measured using a N2O microsensor (Unisense, Denmark) with data logged every 10 s.
Calculations of N2O production
N2O release rate was defined as the amount of N2O released per minute per litre of liquid. Total N2O yield was defined as the total amount of N2O produced per minute per litre of liquid. Zhang et al. (2012) and Quan et al. (2012) have shown that the profiles of air stripping rates of N2O can be described using linear equations with respect to the dissolved N2O concentration in the liquid phase, as shown in Equation (1):
where re is the N2O release rate in mg N2O-N/(L·min), CN2O is the soluble N2O concentration in mg N2O-N /L, and k is the N2O release coefficient in 1/min.
From re, the accumulated amount of N2O released (mg N2O-N/L) can be calculated by integrating re with the running time. Total N2O production (mg N2O-N/L) can be calculated by adding the accumulated amount of N2O released with liquid concentrations of N2O. The total N2O yield, mg N2O-N /(L·min) can be calculated by adding re to the variation rate of N2O concentration in the liquid; the latter can be calculated from the measured values (Quan et al. 2012; Zhang et al. 2012).
N2O release tests were conducted to determine the mass transfer coefficient, k, of Equation (1). The results revealed that the N2O release rate is linearly correlated with liquid N2O concentration with k being 0.144 /min (R2 = 0.9395) for aeration of 40 L/h (including air stripping and diffusion by stirring at 60 rpm in aeration periods) and 0.0057 /min (R2 = 0.9858) for stirring only (diffusion by stirring at 60 rpm during non-aeration periods).
RESULTS AND DISCUSSION
N2O production in the nitritation process
During nitritation, aeration flow of 40 L/h was supplied continuously until the DO concentration exceeded 1.0 mg/L, which signified the end of the nitritation process. Most DO concentration values were 0.3–0.4 mg/L within the 1.8 h aeration period. pH dropped gradually from 8.1 at the beginning to 7.6 at the end. A NO2−-N concentration of 71.0 mg/L was produced and about 82% of the ammonia nitrogen (86.4 mg/L) was converted. A total N2O production of 9.6 mg/L was obtained, and the maximum total yield was 0.18 mg N2O-N/(L·min) (Figure 1(a)). The N2O conversion rates were about 86% of the total nitrogen removed and 11% of the ammonia nitrogen load. The vast majority of N2O produced during nitritation was released as a gas into the air; only a small amount of this compound remained in the liquid as a dissolved material.
Role of heterotrophic denitritation in N2O production
The test results shown in Figure 2 reveal that the liquid N2O concentration initially increases and then gradually decreases when aeration is stopped at the rise period. The early and brief increase in liquid N2O concentration might be brought about by the decrease in air stripping. The subsequent gradual drop may be attributed to the concentrations of COD in the rise period remaining high enough to be used as the carbon source for reducing N2O to N2 by heterotrophic microbes. During this process, COD, NH4+ and NO2− concentrations changed from 108 mg/L, 84.3 mg/L and 0.6 mg/L to 93 mg/L, 72.0 mg/L and 12.0 mg/L, respectively. The negative value of total N2O yield in Figure 2 demonstrates that N2O was consumed. The reduction rates of NO and N2O are greater than those of NO3− and NO2− during heterotrophic denitrification under normal conditions (Pan et al. 2012). Therefore, N2O may be quickly restored into N2 and appears to resist accumulation or release when the organic carbon was sufficient (Wang et al. 2014a, b).
Figure 3 reveals that liquid N2O concentrations continuously increase when aeration is stopped at the highest point (point B in Figure 1(a)). The early, rapid and short increase in liquid N2O concentration might be brought about by a decrease in air stripping. The subsequent continuous rise may be attributed to lower concentrations of COD at the highest point and intracellular carbon source used as the electron donor for reducing NO2− to N2O by heterotrophic microbes; here COD, NH4+ and NO2− concentrations of 58 mg/L, 40 mg/L and 39 mg/L, respectively, were obtained. N2O has been reported to accumulate in liquid when organic carbon is limited because of weak competition of N2O reductase for electrons (Li et al. 2013). Similarly, when PHB is used as the carbon source, N2O emission increases because of slow PHB degradation (Jia et al. 2012).
In Figure 3, when aeration is stopped at the highest point (point B in Figure 1(a)), the total yield of N2O is approximately 0.055 mg/(L·min); which is about 30% of the maximum value of the total yield during nitritation (Figure 1(a)). This finding demonstrates that the function of heterotrophic denitritation in N2O production in the nitritation reactor is considerable, especially under lower DO concentrations and exhaustion of the available organic carbon.
Production of N2O by heterotrophic denitritation via the A/O/A SBR
Test results showed that nearly all of the nitrite accumulated in the nitritation phase, 71.0 mg/L, was reduced after about 18 h in the anoxic phase with a total N2O production of 54.6 mg/L, which is much greater than that for the nitritation phase (9.6 mg/L). The conversion rate of nitrite to N2O obtained was 77%, higher than that obtained with acetate as the carbon source (65%) by Scherson et al. (2013, 2014). During the 2 weeks of continuous tests, MLSS of the reactor decreased slightly from 5,470 to 4,888 mg/L. The average specific N2O production rate was 0.58 mg-N per g MLSS per h in the later anoxic phase. However, the pH always increased from the beginning to the end of the later anoxic phase in one cycle, e.g. 7.60 to 8.83 as shown in Figure 5.
As shown in Figure 4, to investigate the potential of N2O production in the anoxic phase by heterotrophic denitritation further, 100 mg/L NO2−-N was added to the reactor when the liquid N2O concentration was near zero (marked by the arrow in Figure 4). The liquid N2O concentration, release rate, total yield and total production increased when additional nitrite nitrogen was supplied. These results indicate that the proposed operation mode could reduce most of the nitrite nitrogen remaining in the reactor after the aeration phase and convert it into N2O. The high concentration N2O in the liquid was easily recovered with brief additional aeration.
Impacts of influent COD concentrations to N2O production by the A/O/A SBR
Theoretical reduction of NO2− to N2O without cell synthesis required a COD/N ratio of 1.14. At lower COD/N ratios, insufficient organics was converted into PHB in the anaerobic and aerobic phases, and PHB was insufficient for reduction of nitrite; residual NO2− remained in solution at the end of the anoxic phase. In the study of Scherson et al. (2013, 2014), COD/N ratios of 1.5 and 3.0 were adopted. Therefore, besides the influent COD of 530 mg/L (COD/N is 2.2) mentioned above, two other concentrations of influent COD, namely 265 (COD/N is 1.1) and 1,060 mg/L (COD/N is 4.4), were selected to investigate their effects on N2O production by heterotrophic denitritation in the later anoxic phase of the A/O/A SBR. Results showed that, when influent COD concentration was changed from 530 to 1,060 mg/L, the complete nitritation process needed double time to complete. However, the time for nitrite conversion to N2O in the later anoxic phase was shortened for 2 h with a slight increase in total N2O yield. Variation of pH and DO during the later anoxic phase were similar to that of COD, which was 530 mg/L. However, MLSS increased from the initial value of 4,888 to 5,166 mg/L at the end of the 1 week of continuous tests with an average specific N2O production rate of 0.72 mg-N/g MLSS/h in the later anoxic phase. When influent COD concentration was switched to 265 mg/L, corresponding time for complete nitritation process were shortened only by about 15 min. The nitrite still remained in the reactor after 18 h of anoxic stirring, resulting in only 23.6 mg/L of nitrite being converted to N2O. The variation of pH and DO followed the same trends as that of COD, which were 530 and 1,060 mg/L. However, MLSS decreased from the initial value of 5,166 to 4,785 mg/L at the end of the 1 week test. As a result, COD/NH4+-N was greater than 2.2 in the influent, and might be necessary for completely denitrifying nitrite to N2O in the later anoxic phase.
Potential of energy recovery
N2O can act as a powerful oxidant in combustion reactions. When used to oxidize methane, N2O increases the heat of reaction by −329 kJ/mol compared with O2 (Scherson et al. 2013). Injections of N2O into a biogas-fed engine at flow rates simulating a full-scale system increased power output by 5.7 − 7.3% (Scherson et al. 2014). Scherson et al. (2014) found that conversion of NO2− to N2O was essentially complete within 24 h with an initial NO2− concentration of 75 − 85 mg-N/L at the beginning of the anoxic phase. In contrast, initial NO2− concentration of 65–76 mg-N/L at the beginning of the anoxic phase was essentially completely reduced in about 18 h of anoxic phase in our study. This indicates that a similar amount of potential energy as in Scherson et al. (2014) could be recovered and used for improving methane combustion. Thus, A/O/A SBR has the potential for production and recovery of N2O from wastewater containing ammonia and organic carbon, such as reject water.
When the operating mode of the single SBR was A/O/A, the majority of the nitrite transformed from ammonia nitrogen during nitritation process could be converted into N2O in the later anoxic period by heterotrophic denitritation with a conversion rate of 77%. As a consequence, the proposed mode can potentially recover N2O from synthetic wastewater containing ammonia and glucose at influent ammonia nitrogen concentration of 240 mg/L and COD concentration between 503 and 1,060 mg/L, corresponding to COD to NH4+-N ratios greater than 2.2.
This work was supported by the Fundamental Research Funds for the Central Universities (Grant No. 2013G3292017) and the Shaanxi Province Science and Technology Development Program (Grant No. 2014K15-03-02).