Abstract:Abstract: Waste-water is very difficult to treat, especially the waste water from swine farms with high contents of organic matters. Addressing the environmental problems incurred by waste-water is important for the swine farms for their sustainable development. The raw waste-water from swine farm in this study was treated by submerged membrane bioreactor (MBR). Three levels of dissolved oxygen concentrations (DO: <1.5 mg/L, 1.5~3.0 mg/L, and >3.0 mg/L), three levels of hydraulic retention times (HRT: 0.75 d, 1.5 d and 3.0 d) and three levels of reflux ratios (RR, 200%, 300%, and 400%) were used to form nine combined experiments treatments using the orthogonal design. The nine treatments were conducted in three batches in a swine farm with manual manure dry-collection method. The farm was located in Henan province. The MBR was designed with an available volume of 30 L, and the inside temperatures of bioreactors were controlled at (25±5)℃ by heating rods, and the pH values were maintained around 7~8 by addition of acid or alkali. During the experiments, waste-water was pumped to MBR automatically, and sludge retention time (SRT) was controlled at 25-30 d. Each batch of the experiments lasted for 50 d, including 20 d of acclimatization and 30 d for the trial. The results showed that the average concentration of chemical oxygen demand (COD), ammonia nitrogen (NH4+-N), total nitrogen (TN) and total phosphorus (TP) of the MBR effluents were( 202±201) mg/L, (56.6±54.0)mg/L, (91.6±69.1) mg/L and (19.2±10.0) mg/L when the influent having COD, NH4+-N, TN and TP concentrations of (3277±1192) mg/L, (203.8±51.2) mg/L, (361.0±133.3) mg/L, and (65.0±23.1) mg/L, respectively. The corresponding removal rates of MBR were 93.6%±9.5%, 70.0%±27.2%, 70.7%±20.7% and 68.3%±17.4%, respectively for each of these measured parameters. MBR could also remove the fecal coliforms from waste water efficiently. A removal rate of (99.9±0.08)% was observed, and the amount of fecal coliforms in effluents ranged from 30 to 19,350 count/L. About 86.4% of the effluent could meet the hygiene requirements of the national standard. The range analysis results indicated that the importance order of operational parameters on the removal of COD and NH4+-N was DO>HRT>RR, while the order for the TP removal was HRT> DO>RR. When DO increased from <1.5 mg/L to 1.5-3.0 mg/L, the removal rates of COD, NH4+-N, and TN increased significantly (P<0.01). However, when DO was further increased to >3.0 mg/L, significant difference (P<0.01) was only observed in the removal rate of NH4+-N. In comparison with HRT of 0.75 d, the removal rates of COD NH4+-N and TP increased significantly (P<0.01) under HRT of 1.5 d. As for reflux rate (RR), the removal rates of NH4+-N and TN at RR of 300% were significantly higher (P<0.05) than those at RR of 200%, but significant differences (P<0.05) in the removal rates of COD and TP were observed between RR of 300% and RR of 400%. With reference to range analysis results, the optimized operational conditions of MBR were DO of 1.5~3.0 mg/L, HRT of 3.0 d and RR of 300%, which corresponded to the Treatment 4 in the experiments, in which the COD, NH4+-N, TP concentrations were (69.3±48.7) mg/L, (10.0±8.2) mg/L and (14.0±9.9)mg/L, respectively, and the corresponding removal rates were 97.8%±1.5%, 93.8%±5.0% and 81.5%±14.2%, respectively. The effluent from MBR could be disinfected using ultraviolet to ensure biosecurity if the disinfected effluent was recycled for on farm flushing, which implicates the water resource conservation in swine production.