Process performance and multi-kinetic modeling of a membrane bioreactor treating actual oil refinery wastewater

The performance of a membrane bioreactor treating oily wastewater generated by a petroleum refinery was investigated at mixed liquor suspended solids concentrations of 6.5 and 8.5 g/l with hydraulic retention times ranging from 12 to 24 h. An evolutionary polynomial regression technique and multi-kinetic models consisting of the first-order, Grau second-order, and modified Stover-Kincannon were applied for predicting the performance of the bioreactor. The results showed that a maximum chemical oxygen demand reduction of 97% is achievable at the mixed liquor suspended solids of 8.5 g/l and hydraulic retention time of 24 h. The results also revealed positive influence of higher mixed liquor suspended solids and hydraulic retention time values on the bioreactor performance. In addition, the high accuracy of the evolutionary polynomial regression evidenced its capability in investigating the membrane bioreactor. Based on the analysis performed by the resulted model from the evolutionary polynomial regression, the optimum operating conditions (hydraulic retention time of 21 h; mixed liquor suspended solids of 8.2 g/l) were recommended. According to the kinetic evaluation, the Grau second-order and modified Stover-Kincannon kinetics seemed to be the best approach in describing the chemical oxygen demand reduction in the membrane bioreactor.