Abstract:
Recent trends in the development of membrane bioreactor (MBR) have proven MBR as an advanced wastewater treatment technology, but still membrane fouling is a major limitation in the commercialization of MBR. In this study, MBRs fed with real wastewater were operated at different filtration, relaxation and chemical enhanced backwashing (CEB) durations. Box-Behnken response surface methodology was used to find the optimized filtration mode, which generated different combinations of input values, i.e. 15 combinations of independent variables (filtration, relaxation, and CEB). Membrane fouling; fouling duration, transmembrane pressure (TMP) rate, sludge characteristics; average mixed liquor suspended solids (MLSS), capillary suction time (CST), mean particle size (MPS) and effluent quality; average chemical oxygen demand (COD), ammonium-nitrogen (NH4+1-N), total phosphorous (TP) removal efficiencies were analyzed as a function of independent factors. Regression models were generated for the prediction of all the responses and explained statistically by analysis of variance (ANOVA). Multiple correlation coefficient values for all the models were close to one showing a good relation between predicted and experimental values. An optimized run was generated in the numerical optimization portion of RSM i.e. Run (11 1 1.5) with 11 mins filtration, 1 min relaxation and 1.5 mins CEB. The predicted values of responses under optimized operating conditions were 15.13 days fouling duration, 0.57 kPa/day TMP rate, 9.10 g/L avg. MLSS, 23.88 secs avg. CST, 15.85 μm avg. MPS along with 90.45, 61.54 and 40.56% of COD, NH4+1-N and TP avg. removal efficiencies, respectively. A sustainable run performed in addition to optimized run validated the stability of MLSS which was affected by the use of sodium hypochlorite (NaOCl) in backwash.
