Nitrogen and COD removals under various HRTs

At 9 d HRT, the SBR was operated with influent NH4+–N and COD concentrations of 200 mg/L and 100 mg/L, respectively, corresponding to the NLR of 22.2 g/m³-d and OLR of 11.1 g/m³-d. Table 6 compare the range of nitrogen loading rates used and total nitrogen removal under different nitrogen removal processes. As shown in Table 6, the loading rates of anammox and OLAND processes are lower than the present study, which evidences that autotrophic nitrogen removal can happen in lower loading rates also. Moreover, this is the first stage of SNAD seed sludge acclimation in the SBR; thus, the reactor was operated in moderate loading conditions to avoid substrate inhibition. The operating conditions of the SBR under various HRTs are shown in Table 6. The organic loading rate (OLR) and nitrogen loading rate (NLR) to the SBR under various HRTs were worked out, and are also shown in Table 5. However, the NH4+

-N and COD concentrations were kept constant under all HRTs and the ratio of influent COD/TN was maintained at a constant level (0.5).

30

Table 5. The ranges of loading rate under different nitrogen removal processes Nitrogen Removal

Process

Requirement of O2/COD

Nitrogen loading (Kg N m^-3 reactor d^-1)

Total nitrogen removal (%)

Application status

Common reactor configuration

Reference

Conventional High/Yes 0.3-9 95 Full-scale Activated sludge [23,24]

SND Low/No 1-3.5 100 Lab-scale SBR [43]

ANAMMOX None/No 0.003-20 87 Full/Lab scale FBR,SBR [6,7,12,36]

SHARON Low/No 0.5-1.5 90 Full-scale Activated sludge [2,3]

CANON Low/No 0.04-3 75 Lab-scale SBR, UASB [4,3,5]

OLAND Low/No 0.001-0.1 85 Lab-scale SBR, RBC [5,37]

In this study Yes/Yes 0.022-0.066 95 Lab-scale SBR [9,10,12]

Table 6. Characteristics of synthetic wastewater before and after treatment

aVFR increased by 3 times, and without aerator and water jacket problems

bVFR increased by 3 times, and with aerator and water jacket problems

cTN is the sum of NH4

+-N, NO2

- -N and NO3

- -N;

dVFR represents volumetric flow rate and the increases, i.e.1.5, 2 and 3 times, based on 9 d HRT.

32

The influent and effluent profiles of nitrogenous matter and organics are shown in Fig.

4 and 5, respectively. In the first 40 d of operation, a consistent NH4+–N removal (more than 90%) was observed and small quantities of NO₂^--N and NO₃^--N accumulation were found in the SBR. However, SBR displayed a very poor COD removal efficiency (less than 65%) during this period. In the subsequent days (40-65 d), the removal efficiencies increased gradually and have shown a stable NH4+–N and COD removal efficiencies of 96% and 87%, respectively.

0 30 60 90 120 150 180 210 240 270 300 330

Fig.7. Performance of the concentration of nitrogen compounds and removal efficiency of ammonium and total nitrogen in the SBR at different HRTs. 3d ^a without

aerator and water jacket problems, 3d^b with aerator and water jacket problems.

0 30 60 90 120 150 180 210 240 270 300 330

Fig. 8. Performance of the concentration of COD and removal efficiency of COD in the SBR at different HRTs. 3d ^a without aerator and water jacket problems, 3d^b with aerator and water jacket problems.

In order to find the effect of loading rate on the SNAD process, the NLR and OLR were progressively increased by decreasing the HRT from 9 d to 4.5 d, and operated for 47 d (Table 5). Despite the higher influent NLR and OLR, a stable conversion of NH4+–N, without accumulation of NO2

--N/NO3

--N was observed in the SBR. The increased NLR (44 g/m³-d) and OLR (22 g/m³-d), decreased the COD removal efficiency of the SBR from 87% to 78%, whereas the NH4+–N removal efficiency was maintained in the same level, i.e. 95%. This reveals that the increase NLR and OLR have no significant effect on the SNAD system. Table 5 shows the steady-state concentrations of NH₄⁺–N, NO2

--N, NO₃^--N and COD under various HRTs.

Following to the steady-state condition at 4.5 d HRT, the reactor NLR and OLR were further increased to 66 g/m³-d and 33 g/m³-d, respectively, also the HRT was decreased to 3 d. The decrease in the HRT to 3 d has decreased the NH4+–N and COD

34

removals in the system. An increasing trend in the effluent NH₄⁺–N concentration can be noticed in Fig. 5. This indicates that the increases in NLR and OLR (at 3 d HRT) have produced slight inhibition/toxicity to the partial nitrifiers; as a result, insufficient NO2

Unexpectedly, aerator and water jacket were went out-of-order under this recovery stage, which drastically decreased the reactor performance. During this stage, the DO in the SBR has went down to below 0.2 mg/L, pH drop down to less than 6 and the temperature decreased by 5 to 8C. It can be noticed in Table 5 that only 52% of the NH4+–N was removed in the reactor, and interestingly, around 86% of the COD was removed in the reactor. Under this situation, it is hypothesized that Anammox bacteria might be inactive and the NO2

--N produced as a result of partial nitrification could have been utilized only by denitrifiers. In order not to increase further loading under these circumstances, the reactor NLR and OLR were decreased to 33 g/m³-d and 16 g/m³-d. The reactor started to recover when the HRT was increased from 3 to 6 days.

The effluent concentration of NH4+

-N was decreased from 94 mg/L to 25 mg/L, also by slightly adjusting the DO and pH value back to optimal condition, the removal efficiencies of NH4+–N and TN has come back to 75% and 67%, respectively. These observations and hypothesis indicate that high DO concentrations (>2 mg/L) could result complete nitrification in the SNAD system, whereas low DO concentration (<0.5 mg/L) could reduce the rate of nitrification and overall performance of the reactor. Moreover, these data reveal that SNAD process is more resistant to substrate

shock loading compared to sudden change in aeration rate and temperature.

Despite of stable conversion of ammonium, nitrate accumulation was detected in effluent from the end of HRT 3 d. Nitrate production was related to a possible response of different electron acceptors such as sulfate which supplied from (NH4)2SO4 in the medium instead of nitrite in SNAD process. In many researches, except for nitrite, nitrate and propionate, there might be some other electron acceptors for ammonium oxidation and sulfate is considered to be a suitable selection for its strong oxidization capacity. Polanco et al. (2001) showed the possibility of removing ammonium and sulfate simultaneously. They postulated that the nitrite formation and subsequent Anammox process were responsible for nitrogen removal according to the following equations (Eq. (13), (14), (15) and (16))[44]:

3SO₄^2-+4NH₄⁺→3S^2-+4NO₂^-+4H₂O+8H⁺ (13) 3S^2-+2NO₂^-+8H⁺→N2+3S+4H₂O (14) 2NO₂^-+2NH₄+→2N₂+4H₂O (15) SO₄^2-+2NH₄⁺→N2+S+4H₂O (16)

After disturbance during HRT 3 d, the reactor was in unsteady state, the end product of combining sulfate with ammonium might also produce nitrate as well. On the other hand, the long period acclimation of SNAD system may result to accumulation of sulfide which can be toxic to microorganisms. The sludge might be covered by sulfur which could limit the sufficient contact among reactants. The Anammox activity might also affect by some middle medium, such as nitrite, H2S and sulfur causing

36

nitrate accumulation. With decreasing Anammox activity, further works need to focus on reduction of the released sulfureted hydrogen and collection of sulfur from reaction.