periodic N2 flushing did not significantly affect the membrane
rejection, indicating that the membrane cleaning method has little
negative effect on the AnMBR performance. It should be noted that
although significant pollutants reduction is achieved, the effluent
of AnMBR still contains about 1500 mg/L COD and 160 mg/L
NH3–N, which is unable to discharge directly. As a solution, a further biological treatment is still required especially for the removal
of nitrogen. It was reported that anoxic/oxic (A/O) process (Chen
et al., 2010) and sequencing batch reactor (SBR) (Wu et al., 2013)
had been applied as the subsequent treatments of the anaerobic
techniques in treating BIWW, and both the final effluents had successfully met the integrated wastewater discharge standard in China (COD 6 100 mg/L, NH3–N 6 15 mg/L).
The system was operated separately under two modes, with
and without cleaning, at each HRT. As observed in Fig. 6, the experiments in which the N2 cleaning was performed suffered a more
moderate TMP increase compared to those conducted without
cleaning. However, the cleaning effect decreased with an increasing HRT. Under an HRT of 10 d, the differences in TMP between
the 2 modes were slight. The experiment was conducted under
an extreme operating condition when the HRT was decreased to
2 d; under this condition, the TMP increased sharply to 50 kPa