Removal of Reactive Black 5 from Polluted Solutions by Electrocoagulation: Modelling Experimental Data Using Artificial Neural Networks

Abstract

The wide range of today's industry increases the diversity of pollutants in the wastewater characteristics. In particular, the wastewater of the textile industry is highly colored. Different techniques are used for color removal of dyes from wastewater. In this work, the removal efficiency of the textile dye (Reactive Black 5) at different current densities (48.5 A/m2, 97.18 A/m2, 194.36 A/m2, 291.5 A/m2, 388.7 A/m2 ) was investigated by electrocoagulation method. The dye concentration of wastewater prepared in the laboratory scale was adjusted to 100 mg/L. Two iron electrodes and 3 g NaCl were used in the electrocoagulation system. The samples which taken periodically were measured after the centrifugal processes with the UV spectrophotometer. The experimental results were also modelled with artificial neural networks (ANNs). As a result of the experiments, approximately 90-100% color removal efficiency was obtained. According to the modelling study, the ANNs can predict the color removal efficiency with coefficient of determination (R2 ) between the experimental and predicted output variable reached up to 0.99.