Electrochemical treatment of agro-industrial effluent: integrating response surface methodology and neural networks for sustainable solutions

The intensification of agro-industrial processes generates complex effluents rich in contaminants, which can cause significant environmental impacts. This study investigated the efficacy of electrochemical oxidation in treating synthetic agro-industrial effluent using a closed-loop automated flow system powered by renewable energy. The experimental design, conducted through Central Composite Design (CCD), evaluated the effects of volumetric flow rate (Q_v), current density (j), and NaCl concentration on the removal of chemical oxygen demand (COD) and total phenols (TP). The ideal conditions identified (55 L h⁻¹, 15 mAcm⁻² and 3 gL⁻¹) resulted in 60 % COD removal and 98 % TP removal. The integration of regression models and response surface methodology (RSM) demonstrated that increasing current density significantly enhances the treatment efficacy, particularly when combined with higher NaCl concentrations, due to intensified generation of oxidizing species, such as active chlorine species. Analysis of variance (ANOVA) confirmed the validity of the models, with R² values of 95.26 % for COD and 85.07 % for phenols, indicating good fit and statistical significance. Artificial Neural Networks (ANNs) were applied to model and predict the efficiency of the system, incorporating pH, dissolved oxygen, conductivity, and temperature, which were monitored in real-time during electrolysis. The ANNs exhibited excellent generalization capability and reinforced the statistical validation provided by CCD. These findings highlight the importance of operational optimization and continuous monitoring of agro-industrial effluents. The application of RSM and ANNs validates the efficacy of treatment and offers a robust tool for managing complex effluents, contributing to the protection of aquatic ecosystems.