Effect of influent COD/N ratio on performance and N₂O emission of partial nitrification treating high-strength nitrogen wastewater

Abstract

The effect of influent chemical oxygen demand/nitrogen (COD/N) ratio on nitrogen removal and nitrous oxide (N₂O) emission during partial nitrification treating high-strength nitrogen wastewater was investigated. Three laboratory scale anoxic/aerobic sequencing batch reactors (SBRs) were operated under different influent COD/N ratios (0.5, 1.0, and 2.5, respectively). Stable and high nitrite accumulation ratios around 90% were obtained at COD/N ratios of 0.5 and 1.0. A temporary high nitrite accumulation ratio (89.9 ± 1.04%) was observed at a COD/N of 2.5 (days 20–65), however, the main nitrogen species in the effluents changed from nitrite to nitrate afterwards. Despite the many benefits of the partial nitrification process, the significant generation of N₂O, a potent greenhouse gas, remains a problem for this innovative and promising process. Attempts to minimize N₂O emissions during the partial nitrification process by increasing the anoxic/aerobic phase fraction were conducted, and it was found that the N₂O emission quantity was reduced by 50.6% and 46.2% in SBR with influent COD/N of 0.5 and 1.0, respectively. This was because a longer anoxic phase benefited nosZ gene expression and resulted in higher abundance of heterotrophic microorganisms.

---

1. Introduction

Recently, high-strength nitrogen wastewater has received growing attention by researchers due to its potential threat to the environment. This kind of wastewater has a wealth of sources such as ammonia and nitrogenous fertilizer industrial wastewater, leather wastewater and food industry wastewater.¹ It is difficult to deal with this kind of wastewater because of its low chemical oxygen demand/nitrogen (COD/N) ratio. Biological nitrogen removal (BNR) has been widely used as one of the most effective and economical methods for removal of nitrogen from wastewater.² The conventional BNR process generally involves autotrophic nitrification and heterotrophic denitrification: NH₄⁺–N is converted to NO₂⁻–N by both ammonia-oxidizing bacteria (AOB) and ammonia-oxidizing archaea (AOA) and further to NO₃⁻–N by nitrite-oxidizing bacteria (NOB), and NO₃⁻–N is then reduced into nitrogen gas by denitrifying bacteria.³ It was reported that the numbers of AOA amoA genes was correlated negatively to the ammonium levels in wastewater, thus AOB was supposed to be play the dominant role in nitrification process treating high-strength nitrogen wastewater.⁴

To date, novel technologies with reduced operational cost have been developed for the treatment of high-strength nitrogen wastewater to reduce its operating costs, including partial nitrification, anaerobic ammonium oxidation (Anammox), completely autotrophic nitrogen removal over nitrite (Canon), oxygen limited autotrophic nitrification denitrification (Oland) and simultaneous nitrification denitrification (SND) processes.⁵ Among all the technologies mentioned above, partial nitrification is based on the fact that nitrite is an intermediate compound in both nitrification and denitrification steps. Compared to conventional nitrification-denitrification via nitrate, partial nitrification via nitrite saves approximately 25% of oxygen for nitrification and 40% of organic carbon sources for denitrification (eqn (1)–(4)). As a result, nitrite denitrification rate is usually 1.5–2 times higher than that of nitrate.⁶ Until now, many influencing factors have been investigated for achieving the process of partial nitrification, including dissolved oxygen (DO) concentration, temperature, sludge retention time and substrate concentration.⁷

Nitrification:

NH₄⁺ + 1.5O₂ → NO₂⁻ + H₂O + 2H⁺ (partial nitrification) (1)

NH₄⁺ + 2O₂ → NO₃⁻ + H₂O + 2H⁺ (full nitrification) (2)

Denitrification:

6NO₂⁻ + 3CH₃OH + 3CO₂ → 3N₂ + 6HCO₃⁻ + 3H₂O (partial nitrification) (3)

6NO₃⁻ + 5CH₃OH + CO₂ → 3N₂ + 6HCO₃⁻ + 7H₂O (full nitrification) (4)

It is generally accepted that biological wastewater treatment is one of important sources of nitrous oxide (N₂O) emissions.⁸ N₂O is a potent greenhouse gas with a lifetime of 120 years, which has a serious impact on the environment. The global warming potential of N₂O is about three hundred times that of carbon dioxide (CO₂).⁹ In addition, N₂O can also contribute to the destruction of ozone layer by reacting with oxygen (O₂) to produce of nitric oxide (NO) in the stratosphere.¹⁰ N₂O can be an intermediate or end product in the metabolism of both nitrification and denitrification processes. Three sources have been reported to cause N₂O generation and emission from BNR processes: incomplete oxidation of hydroxylamine and nitroxyl (NH₂OH and NOH), nitrifier denitrification and heterotrophic denitrification.^11,12 To date, key factors leading to N₂O emission during nitrogen removal from wastewater were reported in the literature as low dissolved oxygen (DO),¹³ high ammonia concentration,¹⁴ nitrite concentration,¹⁵ low COD/N ratio,¹⁶ as well as organic shock loads.¹⁷ More detailed, in nitrification process, when nitrite is accumulated under low dissolved oxygen, AOB denitrification occurs in which nitrite was reduced to N₂O and N₂.¹⁸ In denitrification process, high nitrite concentrations and low COD/N can lead to a lower denitrification rate and accumulation of NO and N₂O.¹⁹ Partial nitrification expressed great potential in engineering applications, the significant generation of N₂O during this process were not be ignored.²⁰ Therefore, it is desirable to research the N₂O emission in partial nitrification process in order to better understand the sources and mechanisms of N₂O production. Yang et al.²¹ reported that the generation of N₂O was reduced during nitrogen removal via nitrite from domestic wastewater by 50% using the step-feed technique. Until now, limited research has been conducted on N₂O production during nitrogen removal via nitrite from high-strength nitrogen wastewater.

The COD/N ratio is one of the most critical parameters for the successful and sustainable operation of partial nitrification process treating high-strength nitrogen wastewater, and it has effect not only on performance of bioreactors operation by but also on emissions of N₂O in biological nitrogen removal process.²² So far, there is a lack of research to evaluate the effect of COD/N ratio on the performance and N₂O emissions during partial nitrification process.

Thus, the main objective of this study was to investigate the effect of different influent COD/N ratios on performance and N₂O emissions of partial nitrification from high-strength biological nitrogen removal processes. Moreover, as an initial attempt, minimization of N₂O emissions from partial nitrification process was achieved through increasing the anoxic/aerobic phase fraction, and the corresponding mechanism is elucidated by microbial analysis.

2 Materials and methods

2.1 Experiment set-up and operation

Three laboratory-scale SBRs (R1, R2, R3), each with effective volume of 3 L, were operated with three different influent COD/N ratios, i.e., 0.5, 1.0, and 2.5, respectively. The schematic diagram of the experimental reactor is shown in Fig. 1. The study period could be divided into two stages: start-up operation and partial nitrification process (stage I, days 1–200), N₂O emission reduction process (stage II, days 200–260).

Fig. 1 Schematic diagram of SBR: (1) – air (2) – gas flow meter (3) gas sampling pump (4) – gas sampling bag (5) – pH meter (6) – DO meter (7) – liquid sampling port (8) – electromagnetic valve (9) – effluent port (10) – peristaltic pump (11) – water tank (12) – electromagnetic stirrers (13) – rotor (14) – bioreactor.

Influent wastewater was prepared in a storage tank (25 L) and 1.5 L of wastewater was introduced to each reactor using a water pump. Dissolved oxygen (DO) was supplied by an air pump through an air diffuser at the bottom of reactor. The SBRs were operated at room temperature (25 ± 2 °C). The seeding sludge was obtained from the Second Wastewater Treatment Plant of Everbright Water (Jinan) Ltd (Jinan, China). After 3 months of operation, the effluent contaminant concentration tended to be stable and the SBRs were in steady state. The mixed liquor suspended solid (MLSS) of reactors were maintained at approximately 2300–2500 mg L⁻¹.

The reactors were operated by alternating anoxic and aerobic phases. In order to evaluate the influent organic matter on the emission of N₂O and further explore the possible strategy to reduce its emission in reactors, two main stages were consisted in this study. Table 1 shows the detailed experimental conditions during different operational stages of the reactors.

Table 1 Detailed experimental conditions during different operational stages of SBRs

Stage	Time (day)	SBRs	Anoxic stage (min)		Aerobic stage (min)	Anoxic/aerobic phase fraction	Setting (min)	Effluent (min)	Idle (min)	pH	HRT (h)
			Influent filling	Pre-anoxic stage
Ⅰ	1–200	R1 R2 R3	5	25	360	0.07	20	10	60	7.5–8.5	16
Ⅱ	200–260	R1 R2	5	85	300	0.28	20	10	60	7.5–8.5	16

---

2.2 Synthetic nitrogen-rich wastewater

Three SBRs were fed with synthetic nitrogen-rich wastewater containing different concentrations of COD. The compositions of synthetic nitrogen – rich wastewater was as follows (per liter): CH₃COONa, 100–500 mg; NH₄Cl, 200 mg; K₂HPO₄, 112 mg; CaCl₂, 40 mg; MgSO₄·2H₂O, 20 mg; FeSO₄·2H₂O, 20 mg and trace element solution 1.0 mL. The influent COD concentration of R1, R2 and R3 were 105.13 ± 5.34, 196.78 ± 11.45 and 496.63 ± 15.71 mg L⁻¹, respectively. The influent pH values were adjusted to 7.5 – 8.5 by adding NaHCO₃. All the chemicals were purchased from Tianjin Damao chemical reagent factory (China) and used of analytical reagents grade.

2.3 Real-time quantitative PCR

DNA was extracted using PowerSoil DNA Isolation Kit (12888, MOBIO, USA) from sludge of three SBRs and store at −20 °C. Quantitative detection of the nosZ gene was achieved by quantitative PCR using Roche LC-480 (USA). Reaction mixtures (20 µL) contained SYBR Green Mix (10 µL, Roche), RNAase-free water (7 µL), primer (0.5 µL for each) and template DNA (2 µL). nosZ2F and nosZ2R (20 µM) were used as primers.²³ Thermal cycling conditions for the nosZ2F primers were as follows: an initial cycle of 95 °C for 15 min; 6 cycles of 95 °C for 15 s, 67 °C for 30 s with a touchdown of −1 °C by cycle, 72 °C for 30 s, and 80 °C for 15 s (acquisition data step); 40 cycles of 95 °C for 15 s and 62 °C for 15 s, 72 °C for 30 s, and 80 °C for 15 s; and 1 cycle at 95 °C for 15 s and 60 °C for 15 s, to 95 °C for 15 s. The thermal cycling conditions for the nosZ2R primers were similar except for the annealing temperature starting at 65 °C for the first 6 cycles and 60 °C for the 40 cycles. The final quantitative PCR data was generated using Abs Quant/2nd Derivative Max by Roche LC-480 install.

2.4 Analytical methods

To monitor the performance of reactor, NH₄⁺–N, NO₂⁻–N, NO₃⁻–N, COD, MLSS and MLVSS were tested using the standard methods.²⁴ The pH values and DO concentrations were determined using a pH meter (SG2, METTLER TOLEDO, Switzerland) and a DO meter (HQ40d53LDO™, HACH, USA), respectively.

The nitrification activity of the activated sludge was characterized by specific oxygen uptake rate (SOUR), which was measured according to the method of Joanna et al.²⁵ Denitrification rate was measured by using the method reported by Christensen et al.²⁶ The denitrification rates of three SBRs were achieved by addition of 35–40 mg NO₂–N/L as electron acceptor.

The N₂O concentration in gas was measured using a gas chromatography (GC) (SP-3410, Beifen, China) with an electron capture detector (ECD) and a Poropak Q column. The off-gas N₂O was collected and calculated according to the methods of Hu et al.²⁷

2.5 Calculations

The nitrite accumulation ratio (NAR) was calculated according to the following equation:

The stability of nitrite accumulation can be realized by controlling free ammonia (FA; NH₃) concentration and free nitrous acid (FNA; HNO₂), which were the potential compounds for inhibiting the activity of NOB. FA and FNA concentrations in the reactor could be estimated by the following expression (eqn (6) and (7)) proposed by Ford et al.²⁸:

The emission rate and quantity of N₂O were calculated by the following expression (eqn (8) and (9)) proposed by Hu et al.:²⁹

where N₂O–N is total N₂O emission quantity (mg); n is the number of sampling points; w_N2O,n is the N₂O emission rate (mg min⁻¹ g_MLSS⁻¹) at the sampling point n; Δt is the time interval between each sampling point (min); V_L is the volume of the bioreactor (L); MLSS is the mixed liquor suspended solid of the cycle (g L⁻¹).

The specific oxygen uptake rate (SOUR) and R_N (denitrification rate) were calculated according to the following expression (eqn (10) and (11))³⁰:

where SOUR is oxygen uptake rate (mg (g min)⁻¹); DO₁ − DO₂ is the dissolved oxygen interval between each sampling point (mg L⁻¹); R_N is denitrification rate (mg (g min)⁻¹); [NO₂⁻–N]₁–[NO₂⁻–N]₂ is the nitrite concentrations interval between each sampling point (mg L⁻¹); MLSS is the mixed liquor suspended solid of the cycle (g L⁻¹); t₂ − t₁ is the time interval between each sampling point (h).

3 Results and discussion

3.1 Effect of influent COD/N ratio on the performance of partial nitrification process

Fig. 2 shows the ammonium removal efficiencies and nitrite accumulation ratio of SBRs under different COD/N ratios in stage I. Different NH₄⁺–N removal efficiencies were observed during stable stage (days 160–200, Fig. 2A). The NH₄⁺–N removal efficiencies of the three SBRs were maintained at 73.84 ± 0.38%, 83.09 ± 0.56% and 96.13 ± 0.24%, respectively, indicating that influent COD/N ratio had a significant effect on the nitrification capacity of the reactors.

Fig. 2 Ammonium removal efficiencies and nitrite accumulation ratio of SBRs under different COD/N ratios in stage I: (A) NH₄⁺–N removal efficiencies; (B) nitrite accumulation ratios.

Correspondingly, different nitrite and nitrate concentrations were observed during long-term operation of the three reactors. In detail, R1 and R2 always showed stable and high nitrite accumulation ratios (87.9 ± 0.24% and 90.04 ± 0.97%) during the stable stage (days 160–200), implying that partial nitrification process was achieved at COD/N of 0.5 and 1.0. For R3, a temporary high nitrite accumulation ratio (89.9 ± 1.04%) was observed from days 20 to 65. Afterwards, the main nitrogen species in the effluents changed from nitrite to nitrate, implying that the reactor shifted from partial nitrification to full nitrification.

In order to further explain the effect of influent COD/N ratio on the nitrification and denitrification properties of the sludge, (SOUR)-NH₄ (AOB), (SOUR)-NO₂ (NOB) and denitrification rate in each SBR were tested and calculated, as shown in Table 2. R1 and R2 expressed higher (SOUR)-NH₄ (AOB) than that of R3, suggesting that AOB had low abundances or low activity in R3. On the contrary, (SOUR)-NO₂ (NOB) of R3 was higher than that of R1 and R2, indicating higher abundance or activity of NOB in R3. It also showed that the ratio of (SOUR)-NH₄/(SOUR)-NO₂ in R2 and R3 decreased from 4.06 ± 0.08 to 1.13 ± 0.07, implying AOB is more likely to be inhibited compared with NOB by influent COD/N increasing from 1.0 to 2.5.

Table 2 Variation of autotrophic and heterotrophic activities of activated sludge under different COD/N ratios^a,b

	R1 (COD/N = 0.5)	R2 (COD/N = 1.0)	R3 (COD/N = 2.5)
a (SOUR)-NH4 (AOB) = SOUR (+NaClO3) − SOUR (+ATU). b (SOUR)-NO2 (NOB) = SOUR (control) − SOUR (+NaClO3).
(SOUR)-NH4 (AOB) (mg O2/(g SS h))	44.64 ± 0.08	49.68 ± 0.05	25.56 ± 0.06
(SOUR)-NO2 (NOB) (mg O2/(g SS h))	11.12 ± 0.03	12.24 ± 0.06	22.68 ± 0.04
(SOUR)-NH4/(SOUR)-NO2	4.01 ± 0.06	4.06 ± 0.08	1.13 ± 0.07
Denitrification rate (mg NO2^−–N (g^−1 SS h))	6.97 ± 0.12	11.36 ± 0.09	5.53 ± 0.08

---

The influent COD/N ratio is one of the most critical parameters of the nitrification process, because it directly influences the growth competition between autotrophic and heterotrophic microorganism populations.³¹ With the carbonaceous organic matter increased, the proliferation of type heterotrophic bacteria was promoted and the growth of autotrophic nitrification bacteria was inhibited. However, AOB is more likely to be inhibited compared with NOB (Table 2). The number and diversity of AOB began to reduce by influent COD/N increasing to 2.5. Thus, partial nitrification transformed into full nitrification.

The denitrification rates of the three reactors were 0.1162, 0.1893 and 0.0921 mg (g min)⁻¹, respectively. R2 exhibited higher denitrification rate than that of R1 because of the sufficient carbon source for the former. Thus, R2 shows higher nutrient removal efficiency than R1 for the sufficient denitrification. However, minimum denitrification rates were observed in R3 which had the highest influent COD/N ratio among the three SBRs. This was because denitrification rate was measured by adding nitrite as substrate, while denitrifying bacteria in R3 (full-nitrification reactor) usually received nitrate as electron acceptor instead of nitrite.

3.2 Nitrogen transformation during partial nitrification process with different influent COD/N ratios

Variations of NH₄⁺–N, NO₂⁻–N and NO₃⁻–N concentrations in one cycle were tested every 30 min, in order to further study the effect of influent COD/N ratios on the nitrogen transformation during partial nitrification process. Fig. 3 shows the time course of variations of nitrogen compounds, DO concentration and pH in typical cycles of SBRs under different COD/N ratios.

Fig. 3 Time profiles of nitrogen compounds, DO concentration and pH-during typical cycles of SBRs under different COD/N ratios: (A) N-compounds in R1; (B) pH and DO in R1; (C) N-compounds in R2; (D) pH and DO in R2; (E) N-compounds in R3; (F) pH and DO in R3.

Fig. 3A and C shows the typical variations of nitrogen compounds when the influent COD/N ratios were 0.5 and 1.0. The influent wastewater was pumped into the reactor, due to dilution process, NH₄⁺–N concentration first decreased sharply to 129.3 mg L⁻¹ and 125.5 mg L⁻¹, respectively. Afterwards, denitrification occurred under the presence of organic matter in the influent and nitrate or nitrite in the residual effluent of the previous cycle during anoxic phase (0–0.5 h). As a result, nitrite and nitrate concentrations in R1 were decreased from 58.0 and 7.3 mg L⁻¹ to 40.5 and 3.9 mg L⁻¹, respectively. Meanwhile, nitrite and nitrate concentrations in R2 were decreased from 56.7 and 5.4 mg L⁻¹ to 29.8 and 2.2 mg L⁻¹, respectively. Compared to R1, R2 has a better denitrification performance. The reason was that a higher COD/N ratio brings more carbon source with more electron donors for denitrification. Thus, a higher COD/N ratio may promote denitrification and nitrite nitrogen removal in theory.³² The NH₄⁺–N concentration gradually decreased under the supply of DO during aeration phase. More precisely, reduced NH₄⁺–N concentration was accompanied with the increase of NO₂⁻–N concentration, and the NO₃⁻–N concentration was always below 10.0 mg L⁻¹. More detailed, for R1 the NO₂⁻–N and NO₃⁻–N concentrations increased from 54.88 and 4.35 mg L⁻¹ to 116.02 and 14.59 mg L⁻¹ during aeration phase, respectively. Meanwhile, for R2 the NO₂⁻–N and NO₃⁻–N concentrations increased from 41.13 and 1.64 mg L⁻¹ to 113.49 and 10.73 mg L⁻¹ during aeration phase, respectively. Nitrate accumulation was obvious during R1 and R2, with NAR of 88.83% and 91.36%, respectively, indicating the fully limitation of NOB activity.

Compared to partial nitrification-reactors (R1 and R2), variations of nitrogen compounds were different in the full nitrification-reactor (i.e., R3) (Fig. 3E). The reduced NH₄⁺–N concentration was accompanied with the increased NO₃⁻–N concentration, and the NO₂⁻–N concentration was always below 6.0 mg L⁻¹ in R3. More detailed, the NO₂⁻–N and NO₃⁻–N concentrations increased from 2.59 and 3.77 mg L⁻¹ to 3.48 and 103.15 mg L⁻¹ during aeration phase, respectively. Nitrate was the main nitrogen species in the effluents, which implied a high activity of NOB in the full nitrification-reactor.³³

Meanwhile, the COD concentrations during typical cycles of R1, R2 and R3 under different COD/N ratios were tested. The results show that effluent COD of all three SBRs was low. The COD removal efficiencies of three SBRs were 94.1% 95.1% and 96.2%, respectively. The results implied that both reactors had good COD removal efficiencies performance and there was no significant difference of three SBRs.

The DO concentration of R1 and R2 was maintained around 0.3 mg L⁻¹ during the eve of aerobic phase which could inhibit the activity of NOB.³⁴ When the nitrification was completed, the DO concentration increased rapidly (Fig. 3B and D). However, R1 and R2 always showed stable and high nitrite accumulation ratios. The result that nitrite accumulation detected during the oxide phase was because of the inhibition of nitrification, caused by the presence of FA and FNA. In order to better improve understanding in the relationship between FA and FNA on the inhibition of nitrifying bacteria, variations of FA and FNA concentrations in SBR under influent COD/N ratios were evaluated, as illustrated in Fig. 4.

Fig. 4 Typical profiles of FA and FNA concentrations with different influent COD/N ratios: (A) 0.5 of COD/N ratio; (B) 1.0 of COD/N ratio; (C) 2.5 of COD/N ratio.

The FA concentrations of R1 and R2 decreased from 52.06 to 0.55 mg L⁻¹ and 70.54 to 0.72 mg L⁻¹ in the first 2.5 h aeration time (30–180 min), respectively, which was mainly attributed to the rapid decrease of NH₄⁺–N concentration and pH values (Fig. 3B and D). However, FNA concentration of R1 and R2 increased from 0.00037 to 0.056 mg L⁻¹ and 0.00015 to 0.031 mg L⁻¹, respectively, which accompanied by FA decrease during aeration process. Anthonisen et al.³⁵ reported that FA at a concentration of 10 to 150 mg L⁻¹ could inhibit the AOB activity and 0.1 to 1.0 mg L⁻¹ inhibit the NOB activity. It was found that the range of FNA concentrations affecting NOB activity was from 0.011 to 0.07 mg L⁻¹, where complete inhibition was observed at 0.026–0.22 mg L⁻¹.³⁶ Moreover, FNA could cause nitrite accumulation as another potential compound for inhibiting the activity of NOB.^37,38 Nitrate accumulation was obvious and NAR was about 90% (Fig. 3A and C), implying the activity of NOB was fully limited in R1 and R2.

In contrast to R1 and R2, significant variations of FA and FNA concentrations could be found in Fig. 4C during the full nitrification-reactors (R3). The FNA concentration was always under low level accompanied by FA decrease during aeration process. Therefore, R3 achieved a full nitrification process without accumulation of nitrite in the system.

3.3 N₂O emissions during partial nitrification process with low influent COD/N ratio

N₂O emissions characteristics in SBRs with different influent COD/N ratios are shown in Fig. 5. It can be seen that N₂O emission characteristics in the partial nitrification-reactors system at influent COD/N of 0.5 and 1.0 showed an identical pattern but with different details. The majority of N₂O in partial nitrification-reactors were detected during the anoxic period. This result was different from that of Itokawa et al.³⁹ The reason was the large amount of NO₂⁻ during all anoxic period (Fig. 3A and C).

Fig. 5 Compared of N₂O emission rate during typical cycles of R1, R2 and R3.

N₂O emission rates in both reactors increased rapidly at the beginning of aeration. The highest N₂O emission rate of R1 and R2 were 13.02 and 13.61 µg min⁻¹ g_MLSS⁻¹ at 90 and 180 min, respectively. Then, the N₂O emission rates of the two reactors were gradually decreased to 0.63 and 0.32 µg min⁻¹ g_MLSS⁻¹ at the end of aerobic phase, implying that the profile of N₂O emission rate in each SBR was consistent with the change of nitrite concentration (Fig. 3A and C), which was also observed in other literature.²⁹ Nitrifier denitrification and heterotrophic denitrification, which were the two main processes responsible for most N₂O emission during high-nitrite PN process, were carried out quickly during this period, leading to the high N₂O emission rate. Additionally, the N₂O emission rate increased rapidly at the beginning of anoxic period in partial nitrification-reactors (R1 and R2) and reached to the peak at 20 and 15 min, respectively. This was because much higher denitrification rate of R2 (Table 2).

In this study, both the reactors were supplied with high ammonia–nitrogen wastewater under low COD/N ratio. Kampschreur et al.⁴⁰ reported that COD limitation and high ammonium concentration could lead to nitrite accumulation during denitrification stage. Moreover, Chung YC et al.⁴¹ reported that lack of biodegradable organic carbon is known to increase N₂O emission during denitrification. Fig. 5 showed that increasing the influent COD/N ratio could lead to a lower production of N₂O in one cycle during the partial nitrification process. This was because the imbalance between synthesis of NO₂⁻ reductase and N₂O reductase during partial nitrification process. Under lower COD/N ratio, the competition of nitrite reductase and nitrous oxide reductase for electron donor resulted in accumulation of N₂O.

Compared to partial nitrification-reactors, variations of N₂O emission were different in the full nitrification-reactor (R3). The N₂O emission rate of R3 was lower than partial nitrification-reactors during the whole cycle which attributed to the nitrite as an important factor causing N₂O production. The result was consistent with the literature reported by Itokawa et al.¹⁶ Thus, it is necessary to reduce N₂O emissions during partial nitrification mode.

In our study, the following experimental strategy (stage II) was designed and implemented selecting R1 and R2 as operational conditions to reduce N₂O emissions during partial nitrification process.

3.4 Attempts to minimize N₂O emissions during partial nitrification process

Contrast to stage I, the anoxic/aerobic phase fraction was changed from 0.07 to 0.28 in the following operational period (days 200–260). After over 2 months of acclimation, R1 and R2 achieved stable nitrogen removal and nitrite accumulation ratios, and then the generation of N₂O in two SBRs was assessed. The N₂O–N emission amount and conversion rate during different operational periods (stage I and II) were calculated and the results are shown in Table 3.

Table 3 A typical cycle of N₂O–N emission quantity during different operational periods

Operation stage		N2O–N emission quantity during anoxic stage (mg)	N2O–N emission quantity during aerobic stage (mg)	Total N2O–N emission quantity (mg)	N2O–N conversion rate^a (%)
a N2O–N conversion rate = N2O–N/N removed × 100%.
Stage I	R1	15.28	4.20	18.67	96.17
	R2	2.62	7.93	10.56	23.93
Stage II	R1′	5.03	3.38	9.23	48.02
	R2′	1.81	3.89	5.70	20.09

---

During stage I, the emission amount during aerobic stage for R1 was negligible compared to the amount during anoxic stage. The reason was that high nitrite concentration and shortage of carbon source during anoxic stage. It was found that N₂O emission amount was reduced with the increase of COD/N ratios. The N₂O–N emission amount per cycle of R2 was 43.44% lower than that of R1. Only 23.93% of removed nitrogen of R2 was converted to N₂O–N, which was also much lower than that of R1.

During stage II, it was found that the N₂O emission quantity during anoxic and aerobic stages was decreased with the increase of the anoxic/aerobic phase fraction. Table 3 shows the N₂O–N emission quantity of R1 and R2 during anoxic stage was significantly reduced from 15.28 and 2.62 mg to 5.03 and 1.81 mg, resulting in the N₂O–N conversion rate reduced from 96.17% and 23.93% to 48.02% and 20.09%, respectively. Meanwhile, the reduction of N₂O–N emission quantity during aerobic stage of R1 and R2 were 19.52% and 50.94%, respectively.

In order to further explain the causes of low N₂O generation in stage II, quantitative real-time PCR assay targeting the nosZ gene was conducted. It is reported that nosZ gene encoding the catalytic subunit of the nitrous oxide reductase and has recently been used for the detection of denitrifier-specific DNA.⁴² Abundance of nosZ genes in activated sludge under different operational stages is shown in Fig. 6.

Fig. 6 Abundance of nosZ genes in activated sludge under different operational stages. Error bar indicate standard errors of the two independent PCRs of the three replicate DNA extractions.

As shown in Fig. 6, higher nosZ copies density was observed in R2 (COD/N = 1.0) during both stages. The reason was because higher carbon source concentration in the influent of R2 could improve the quantity of heterotrophic microorganisms.⁴³ Contrast to stage I, the density of nosZ gene copies increased significantly in stage II, mainly because higher anoxic/aerobic phase fraction (i.e., longer anoxic phase) accelerated the expression of nosZ gene. Henry et al.⁴⁴ found that higher density of nosZ gene copies was correspondence with more bacteria capable of reducing N₂O to N₂. Therefore, the generation of nitrous oxide in R1 was much lower than in R2 and of that in stage II was much lower than in stage I.

4. Conclusions

This study aimed to evaluate the effect of influent COD/N ratio on performance and N₂O emissions of partial nitrification treating high-strength nitrogen wastewater. The key findings of this study can be drawn as follows:

(1) The influent COD/N ratio is one of the most critical parameters for the successful and sustainable operation of partial nitrification process. Long term operation under high COD/N ratio (>2.5) could mitigate NO₂⁻ accumulation and resulting in destruction of partial nitrification.

(2) N₂O emission was highly dependent on the COD/N ratio, which is one of the most significant parameters for N₂O emission during partial nitrification process. Although partial nitrification process was stably achieved at COD/N below 1.0, a large quantity of N₂O emission was detected. And COD/N ratio has a significant effect on N₂O emission during partial nitrification process.

(3) Reduced N₂O emission was achieved through the increase of anoxic/aerobic phase fraction, which could improve the nosZ gene expression and lead to higher abundance of heterotrophic microorganisms.

Acknowledgements

This work was supported by National Natural Science Foundation of China (No. 21307076 & No. 21177075), Fundamental Research Funds of Shandong University (No. 2014TB003), and Independent Innovation Foundation of Shandong University (2014JC023), Fundamental Research Funds of Shandong University (No. 2014TB003 & 2015JC056).

References

1. J. Carrera, T. Vicent and J. Lafuente, Process Biochem., 2004, 39, 2035–2041.
2. N. Frison, A. Chiumenti, E. Katsou, S. Malamis, D. Bolzonella and F. Fatone, J. Cleaner Prod., 2015, 93, 126–133.
3. D. Wei, B. Du, J. Zhang, Z. Hu, S. Liang and Y. Li, Bioresour. Technol., 2015, 190, 474–479.
4. Y. Bai, Q. Sun, D. Wen and X. Tang, FEMS Microbiol. Ecol., 2012, 80, 323–330.
5. J. Lin, P. Zhang, J. Yin, X. Zhao and J. Li, Int. Biodeterior. Biodegrad., 2015, 102, 49–55.
6. W. Jianlong and Y. Ning, Process Biochem., 2004, 39, 1223–1229.
7. J. Guo, Y. Peng, S. Wang, Y. Zheng, H. Huang and Z. Wang, Bioresour. Technol., 2009, 100, 2796–2802.
8. P. Wunderlin, J. Mohn, A. Joss, L. Emmenegger and H. Siegrist, Water Res., 2012, 46, 1027–1037.
9. IPCC, Climate Change 2001: The Scientific Basis, Cambridge University Press, Cambridge, 2001.
10. A. R. Ravishankara, J. S. Daniel and R. W. Portmann, Science, 2009, 326, 123–125.
11. H. Itokawa, K. Hanaki and T. Matsuo, Water Res., 2001, 35, 657–664.
12. Z. Hu, J. Zhang, S. Li, J. Wang and T. Zhang, Bioresour. Technol., 2011, 102, 5486–5491.
13. Z. Hu, J. W. Lee, K. Chandran, S. Kim and S. K. Khanal, Environ. Sci. Technol., 2012, 46, 6470–6480.
14. D. Wei, L. Shi, G. Zhang, Y. Wang, S. Shi, Q. Wei and B. Du, Bioresour. Technol., 2014, 171, 487–490.
15. Y. Zhou, M. Pijuan, J. R. Zeng and Z. G. Yuan, Environ. Sci. Technol., 2008, 42, 8260–8265.
16. H. Itokawa, K. Hanaki and T. Matsuo, Water Res., 2001, 35, 657–664.
17. C. Li, T. Wang, N. Zheng, J. Zhang, H. H. Ngo, W. Guo and S. Liang, Bioresour. Technol., 2013, 141, 160–166.
18. S. Kim, M. Miyahara, S. Fushinobu, T. Wakagi and H. Shoun, Bioresour. Technol., 2010, 101, 3958–3963.
19. R. V. Schulthess, M. Kühni and W. Gujer, Water Res., 1995, 29, 215–226.
20. H. Itokawa, K. Hanaki and T. Matsuo, Water Res., 2001, 35, 657–664.
21. Q. Yang, X. Liu, C. Peng, S. Wang, H. Sun and Y. Peng, Environ. Sci. Technol., 2009, 43, 9400–9406.
22. D. Wei, Z. Qiao, Y. Zhang, L. Hao, W. Si, B. Du and Q. Wei, Appl. Microbiol. Biotechnol., 2013, 97, 1745–1753.
23. S. Henry, D. Bru, B. Stres, S. Hallet and L. Philippot, Appl. Environ. Microbiol., 2006, 72, 5181–5189.
24. APHA, AWWA and WEF, Standard Methods for the Examination of Water and Wastewater, Washington, D.C, 21st edn, 2005.
25. J. Surmacz-Gorska, K. Gernaey, C. Demuynck, P. Vanrolleghem and W. Verstraete, Water Res., 1996, 30, 1228–1236.
26. M. H. Christensen and P. Harremoës, Prog. Water Technol., 1977, 8, 509–555.
27. Z. Hu, J. Zhang, S. Li and H. Xie, Environ. Sci. Pollut. Res., 2013, 20, 1059–1069.
28. D. L. Ford, R. L. Churchwell and J. W. Kachtick, J. – Water Pollut. Control Fed., 1980, 52, 2726–2746.
29. Z. Hu, J. Zhang, S. Li, H. Xie, J. Wang, T. Zhang, Y. Li and H. Zhang, J. Biosci. Bioeng., 2010, 109, 487–491.
30. J. L. Wang, L. B. Wu, X. Qi and Y. Qian, Acta Sci. Circumstantiae, 1999, 19, 225–229 , in Chinese.
31. S. Cheng, Water Sci. Technol., 1994, 30, 131–142.
32. Y. Zhao, Y. Zhang, Z. Ge, C. Hu and H. Zhang, Environ. Sci.: Processes Impacts, 2014, 16, 567–575.
33. W. Jia, S. Liang, H. H. Ngo, W. Guo, J. Zhang, R. Wang and Y. Zou, Bioresour. Technol., 2013, 141, 123–130.
34. J. Guo, Y. Peng, S. Wang, Y. Zheng, H. Huang and Z. Wang, Bioresour. Technol., 2009, 100, 2796–2802.
35. Y. Zhou, A. Oehmen, M. Lim, V. Vadivelu and W. J. Ng, Water Res., 2011, 45, 4672–4682.
36. J. Wu, J. Zhang, W. Jia, H. Xie, R. R. Gu, C. Li and B. Gao, Bioresour. Technol., 2009, 100, 2910–2917.
37. D. Wei, X. Xue, L. Yan, M. Sun, G. Zhang, L. Shi and B. Du, Chem. Eng. J., 2014, 235, 19–26.
38. Y. Chung and M. Chung, Water Sci. Technol., 2000, 42, 23–27.
39. H. Itokawa, K. Hanaki and T. Matsuo, Water Res., 2001, 35, 657–664.
40. M. J. Kampschreur, H. Temmink and R. Kleerebezem, Water Res., 2009, 43, 4093–4103.
41. Y. Chung and M. Chung, Water Sci. Technol., 2000, 42, 23–27.
42. D. J. Scala and L. J. Kerkhof, FEMS Microbiol. Lett., 1998, 162, 61–68.
43. S. L. Henderson, C. E. Dandie, C. L. Patten, B. J. Zebarth, D. L. Burton, J. T. Trevors and C. Goyer, Appl. Environ. Microbiol., 2010, 76, 2155–2164.
44. S. Henry, D. Bru, B. Stres, S. Hallet and L. Philippot, Appl. Environ. Microbiol., 2006, 72, 5181–5189.

---