Applications of integrated response surface methodology statistic techniques and artificial neural network‐based machine learning to optimize residual chlorine production and energy consumption

A multifactor interaction study was performed using the combined response surface methodology and an artificial neural network on the operational parameters and their influence on residual chlorine production. The operating variables, sodium chloride concentration, electrical potential, electrolysis time, and electrode gap, were evaluated over the response, residual chlorine and energy consumption. The results indicated that the optimum value for residual chlorine was 2450 mg/L achieved at an electrical potential of 8.8 V for 25 min in the presence of 25 g/L of sodium chloride and an electrode distance of 1 cm, and the optimum corresponding energy consumption was measured at 21.76 kWh/L. The study reveals that electric potential, sodium chloride concentration, and electrolysis time positively influence residual chlorine production. ANN models showed superior prediction ability compared with RSM models. This suggests electrolysis can be used for active chlorine production from saline solutions, potentially for industrial applications and water disinfection. Highlights The electrode gap was shown to have little effect on the formation of residual chlorine. The electrolysis time and electric potential have a direct impact on energy consumption. Artificial neural network models demonstrated superior capability for process prediction. A maximum of 21.756 kWh/L of energy can be utilized for producing residual chlorine.