Prediction of electrocoagulation treatment of tannery wastewater using multiple linear regression based ANN: Comparative study on plane and punched electrodes

Abstract

This study investigated the electrocoagulation (EC) treatment of tannery wastewater using plane and punched aluminum and iron electrodes at the optimum condition of pH 9, voltage 20 V, electrode distance of 1 cm and 90 min electrolysis duration. The efficiency of the EC process was determined by evaluating the levels of biochemical oxygen demand (BOD), chemical oxygen demand (COD), and Chromium (Cr) in the treated effluents. The experiment utilized both linear regression and Artificial Neural Network (ANN) models for modeling, with the ANN model validating the predicted model from the experimental design with 95 % confidence. The use of plane aluminum electrodes resulted in an optimum removal efficiency of BOD (89.66 %), COD (96.21 %), Cr (96.05 %), and TDS (95.77 %). On the other hand, the punched electrodes achieved a removal efficiency of 90.86 % (BOD), 98.62 % (COD), 96.94 % (Cr), and 96.92 % (TDS). Similarly, when using plane iron electrodes, the removal efficiency of BOD, COD, Cr and TDS was 87.57 %, 94.77 % 93.42 % and 93.08 %, respectively, while punched iron electrodes removed 89.01 % of BOD, 96.59 % of COD, 94.66 % of Cr and 95 % of TDS. The results demonstrate that the proposed ANN effectively predicts effluent BOD, COD, Cr and TDS, addressing economic and environmental sustainability concerns.