Long-Term Effects of Trace NO 2 Addition on the Performance of CANON SBR

: Two parallel CANON Sequencing Bach reactors were started, and 67ppm NO 2 was added into Sequencing Bach Reactor 2 while nothing was added to Sequencing Bach Reactor 1. The total nitrogen removal efficiency of SBR1 was 65.5±5.0% at a removal rate of 0.198±0.023 kgN/m 3 /d. Meanwhile, the SBR2 with NO 2 addition showed a removal efficiency of 67.5±6.2%, with a removal rate of 0.277±0.017 kgN/m 3 /d. The SBR2 had a higher removal efficiency and rate than the SBR1. The continuous addition of trace NO 2 into the CANON Sequencing Bach Reactor allows conventional aerobic ammonia oxidation with O 2 as the electron acceptor and ammonia oxidation of with NO 2 as the electron acceptor to take place simultaneously, thus improving the ammonia oxidation rate and autotrophic nitrogen removal performance.


Introduction
The aerobic ammonia-oxidizing bacteria of Completely Autotrophic Nitrogen Removal Over Nitrite (CANON) would firstly oxidize part of NH 4 + into NO 2 -, while the anaerobic ammonia-oxidizing bacteria then convert NO 2and the rest NH 4 + into N 2 . Compared with traditional nitrification and denitrification, the CANON process can theoretically save 62.5% of O 2 required for nitrification and 100% of organic COD required for denitrification [1] . The CANON process was developed by Sliekers et al. [2] from Delft University of Dutch. Aerobic and anaerobic ammonia-oxidizing bacteria were cultured in SBR with organic matter of low concentration and limited oxygen (< 0.5% of air saturation). The removal rate of total nitrogen in the system reached 0.3kgN/m 3 /d, with most NH 4 + being converted into N 2 (85%) and the rest into NO 3 -(15%). The autotrophic nitrogen removal involving aerobic ammonia oxidation and anaerobic ammonia oxidation is given by [2] : In CANON SBR, aerobic ammonia-oxidizing bacteria and anaerobic ammonia-oxidizing bacteria co-exist, and trace NO 2 has great influence on autotrophic nitrogen removal. Schmidt et al. [3] inoculated aerobic ammonia-oxidizing bacteria and anaerobic ammonia-oxidizing bacteria, and used NO 2 as electron acceptor to study the effect of NO 2 on aerobic and anaerobic ammonia-oxidizing bacteria. In the presence of excessive NH 4 + in the system, aerobic and anaerobic ammonia-oxidizing bacteria would work together, with the former using NO 2 as electron acceptor to oxidize part of NH 4 + in the system into NO 2for anaerobic ammonia-oxidizing bacteria, which consumes NO [4,5] formed by the reduction of NO 2 and relieve the toxicity of NO to cells. In the pure culture of aerobic ammonia-oxidizing bacteria, the microbial activity was inhibited at 50ppm NO 2 . However, the CANON system didn't show significant inhibitory effect even with a NO 2 concentration of up to 250 ppm. With limited NH 4 + , aerobic ammonia-oxidizing bacteria are in a dominant position in the competition with anaerobic ammonia-oxidizing bacteria for NH 4 + , most of which is oxidized to NO 2 -, thus inhibiting anaerobic ammonia-oxidizing bacteria. The accumulation of NO 2leads to a decline in the total nitrogen removal rate.
Ren Hongyang [6] found that the ammonia nitrogen degradation rate of NO2 ammonia oxidation peaked at 5.36mg/(gꞏh) under anaerobic conditions when NO 2 concentration was 4.475mmol/m 3 (100ppm). The introduction of trace NO 2 can improve the activity of aerobic ammonia-oxidizing bacteria. The activity of aerobic ammonia-oxidizing bacteria peaked at 161.21mg/(gꞏh) with a DO concentration of 1.5-2.0mg/L and NO 2 concentration of 4.475mmol/m 3 (100ppm). The enhanced kinetic model of aerobic ammonia oxidation can be regarded as the coupling of conventional aerobic ammonia oxidation and NO 2 ammonia oxidation. The activity of anaerobic ammonia oxidation peaked at 4.10mg/(gꞏh) with the NO 2 concentration of 2.23 mmol/m 3 (50ppm). The enhancement effect of NO 2 on anaerobic ammonia oxidation can be described by the modified Andrews model. According to the kinetic characteristics of NO 2 -enhanced aerobic and anaerobic ammonia oxidation, a kinetic model of NO 2 -enhanced CANON was established, the simulation results of which showed that the total nitrogen removal rate peaked with a DO of 0.6mg/L and NO 2 concentration of 3.0mmol/m 3 (67ppm). In this paper, two parallel SBR reactors were inoculated with anaerobic ammonium oxidation sludge to enable quick start of CANON. This paper observed the long-term effects of trace NO 2 on reactor with or without trace NO 2 (67ppm), and investigated the enhancement effect of NO 2 on CANON SBR reactor.

Test equipment, sewage and sludge
The SBR is made of cylindrical plexiglass, with an inner diameter of 11cm, height of 50cm and effective volume of 2.5L. There is a constant temperature water bath layer outside the reactor set at 31±1℃. The exterior is wrapped with black shading material for inhibition of light-induced anaerobic ammonium oxidation bacteria. The temperature of the reactor is controlled at 31±1℃ by constant temperature water bath layer. The pH value is adjusted by adding NaHCO 3 into the inlet water, and is left alone during operation. The addition of NO 2 into the reactor was enabled by a set of gas flow control device, which introduces a proper amount of air and NO 2 with a concentration of 5000ppm (the balanced gas consists of N 2 and CO 2 with a concentration of 300ppm), which obtained NO 2 with a concentration of 67ppm in the SBR. See Figure 1 for the diagram of enhanced SBR with trace NO 2 . The SBR reactor without trace NO 2 showed the same performance as previously reported [7] .
The inflow of SBR was synthetic wastewater consisting of (NH 4 ) 2 SO 4 as the substrate and NaHCO 3 as inorganic carbon source with a concentration of 1.5mg /L. Based on existing literatures, the trace elements include two types (trace element liquid I and II) [8,9]

Experiment Plan
The anaerobic ammonia oxidation sludge was inoculated and two parallel CANON SBRs (SBR1 and SBR2) were started synchronously. The operation cycle of SBR lasts for 240min (4h), including 12min of water inflow (a volume of about 1L), 138min of oxygen-limited aeration, 66min of anaerobic digestion, 12min of sedimentation, and 12min of discharge. In the first step, the ammonia nitrogen and aeration in the influent gradually increased, with the aeration rate rising from 100mL/min on the 16th day to 150mL/min (on the 46th day), which remained constant till the end. On the 16th day of operation, a small amount of NO 2 was introduced into SBR2, while nothing was done to SBR1. Meanwhile, the other operating conditions were kept the same. Then, the paper studied the long-term effect of NO 2 on CANON SBR.

Method for analysis
The pH value was determined by online pH electrode (pH2100e, METTLER TOLEDO). The concentration of NH 4 + -N was determined by Nessler's reagent spectrophotometry [10] , while that of the NO 2 --N was tested by N-(1-naphthyl)-ethylenediamine dihydrochloride spectrophotometric method [10] . Phenol disulfonic acid was used to measure the concentration of NO 3 --N [10] , and gravimetric analysis was adopted to determine the MLSS and MLVSS of sludge [10] .

Long-term effect of trace NO2 on the operation of reactors
See Figure 2 for the start-up of CANON SBR with or without trace NO 2 . The operation of the reactor (SBR1) without adding trace NO 2 are shown in Figure 2a and 2b. The operation of the gas reactor (SBR2) with 67ppm NO 2 added after 16 days of operation is shown in Figure  2c and 2d. The NO 2 -enhanced SBR2 had a removal efficiency of total nitrogen of 67.5±6.2% and a removal rate of total nitrogen of 0.277±0.017kgN/m 3 /d (The average values from 32d to 64d are shown in Figure 2d). Meanwhile, SBR1 without NO 2 had a removal efficiency of total nitrogen of 65.5±5.0% and a removal rate of total nitrogen of 0.198±0.023kgN/m 3 /d (The average values from 39d to 52d are shown in Figure 2b), which were lower than the former. According to Figure 2a and 2b, when the aeration rate increased to 150mL/min in the late operation (54d-64d) of SBR1 without NO 2 , the concentration of effluent nitrate nitrogen rocketed along with a sharp fall in the total nitrogen removal rate and total nitrogen removal efficiency. However, the NO 2 -enhanced reactor didn't have the problem in the same period (54d-64d).

Fig 2 Performance of CANON SBR with/without trace NO2 addition
To further study the influence of trace NO 2 on autotrophic nitrogen removal, a typical operation cycle of 14th day (no trace NO 2 was added to both reactors) and 48th day (trace NO 2 was added to SBR2) was selected to investigate changes in the concentration of ammonia nitrogen, nitrite nitrogen and nitrate nitrogen in SBR1 and SBR2. The results are shown in Figure 3. On the 14th day of operation before trace NO 2 addition (see Figure 3a and 3b for operation on the 14th day), nitrite accumulation was found in both reactors, indicating both SBR reactors were in the initial stage of CANON dominated by nitrosation. SBR1 had a removal rate of total nitrogen of 0.113 kgN/m 3 /d and an efficiency rate of 48.4%, while the number was 0.086kgN/m 3 /d and 37.0% for SBR2 respectively. The total nitrogen removal rate and efficiency as well as effluent nitrite accumulation concentration of SBR1 were higher than those of SBR2, while the effluent nitrate nitrogen concentration was lower than that of the latter, indicating that SBR1 had a slightly higher efficiency than that of SBR2. The operation of both reactors on day 48 with trace NO 2 addition to SBR2 (67ppm) are shown in Figures 3c and  3d, and no nitrite accumulation was found in the two reactors. The ammonia nitrogen degradation rate of SBR2 was obviously higher than that of SBR1, while the effluent nitrate nitrogen concentration was lower than that of the latter. The removal rate of total nitrogen in reactor SBR1 was 0.216 kgN/m 3 /d, with a removal efficiency of total nitrogen of 55.8%. The number for SBR2 was 0.299kgN/m 3 /d and 68.6% respectively. The removal rate and efficiency of total nitrogen in SBR1 were lower than those of SBR2, indicating that long-term addition of trace NO 2 can effectively enhance autotrophic nitrogen removal.

Principle of enhanced autotrophic nitrogen removal by trace NO2 addition
NO 2 can enhance aerobic ammonia oxidation. The addition of trace NO 2 in the aerobic ammonia oxidation allows conventional aerobic ammonia oxidation with O 2 as electron acceptor and NO 2 ammonia oxidation with NO 2 as electron acceptor to take place simultaneously, thus significantly improving the ammonia oxidation rate [7,11] . Zhang et al. [11] studied the effect of trace NO 2 addition on aerobic ammonia-oxidizing bacteria through batch experiments, and found that the addition of 100ppm NO 2 into the mixed gas of 21% O 2 could maximize the ammonia oxidation rate (139.11mgN/COD/h) and nitrogen removal rate (34.19%). Kartal et al. [12] studied the long-term effect of high concentration NO 2 (4000ppm) on the expression of Nitrosomonas eutropha C91 gene and protein under aerobic and anaerobic conditions. They argued that NO 2 may have short-term effects on ammonia-oxidizing bacteria under anaerobic conditions. Under aerobic and anaerobic conditions, the addition of high concentration NO 2 resulted in a decrease in protein related to N or C assimilation storage and an increase in protein related to energy metabolism (such as ammonia monooxygenase AmoCAB), while some protein related to nitrogen oxide metabolism did not change. This explains the previously reported enhancing effect of NO 2 on ammonia oxidation rate [11] : in the presence of NO 2 , the energy generated by ammonia oxidation is more used for the basic metabolism of microorganisms than the nutrients storage.

Conclusion
Two parallel CANON SBRs were quickly started by inoculating anaerobic ammonium oxidation sludge. The total nitrogen removal efficiency of SBR1 without adding trace NO 2 was 65.5±5.0%, and the total nitrogen removal rate was 0.198±0.023 kgN/m 3 /d. The removal efficiency of total nitrogen in SBR2 with continuous NO 2 addition was 67.5±6.2%, and the removal rate of total nitrogen was 0.277±0.017 kgN/m 3 /d. Both indicators of SBR2 were higher than those of SBR1, which means the continuous NO 2 addition into CANON SBR could improve the ammonia oxidation rate and enhance the autotrophic nitrogen removal performance by allowing conventional aerobic ammonia oxidation with O 2 as electron acceptor and NO 2 ammonia oxidation with NO 2 as electron acceptor to take place simultaneously.