Investigation and optimization of a novel MAHB reactor for COD and lignin removal and methane production using response surface methodology (RSM) and artificial neural network (ANN)

Abstract

In this study, response surface methodology (RSM) and artificial neural network (ANN) were used to investigate and optimize COD and lignin removal and methane production rate in a novel modified anaerobic hybrid baffled (MAHB) reactor treating recycled paper mill effluent (RPME). Both feeding COD concentration and hydraulic residence time (HRT) are recognized as the two most important factors that affect COD and lignin removal and methane production rate. RSM analysis gives an optimum condition with HRT of 3.93 days and feeding COD concentration of 3020.88 mg L-1 that yield COD removal efficiency of 97.42 %, lignin removal efficiency of 59.59 % and methane production rate of 8.07 L CH4 day-1 with desirability value of 0.897. From the analysis using ANN, results show a good agreement between experimental and ANN outputs for COD removal, lignin removal and methane production rate with R2 values of 0.970, 0.9906 and 0.9545, respectively. These demonstrated that RSM and ANN were effective to assess and optimize the MAHB reactor system for COD and lignin removal and methane production, which provides a promising guide to further improvement of the system for potential applications. © 2020 Bulgarian Academy of Sciences, Union of Chemists in Bulgaria