Performance, evaluation, and modeling of an integrated petroleum refinery wastewater treatment system using multi-layer perceptron neural networks

Abstract

ABSTRACT This study investigates the performance of an integrated petroleum refinery wastewater treatment system. The proposed system attempts to improve the performance of an activated sludge system by using immersed vertical rotating biological contactors in the aeration basin of the system. This system is an innovative method of biological petroleum refinery wastewater treatment with a hybrid growth process. A sand filter column was used in the last part of the treatment process. Also, a multi-layer perceptron neural network (MLPNN) was applied to predict pollutants in the effluent. Overall treatment efficiencies of chemical oxygen demand (COD), total suspended solids (TSS), oil, ammonia (NH 3 ), and turbidity were 94%, 90%, 88%, 93%, and 92%, respectively. According to the findings, the removal efficiencies of pollutants in our integrated system were superior to conventional activated sludge systems. Training procedures of all effluent quality parameters were successful for the MLPNN model. The training model showed an almost acceptable match between the experimental and predicted values. For all models predicting effluent COD, TSS, oil, NH 3 , and turbidity, the correlation coefficient was higher than 0.90, and the mean squared error varied from 0.0001 to 0.234 for the measured parameters. The results confirmed the effectiveness of the integrated system in achieving high removal efficiencies.