ANN-Based Modeling of Combined O3/H2O2 Oxidation, and Activated Carbon Adsorption Treatment System: Forest Polluting Site Leachate

Treatment of organic leachate is one the most controversial topics around the world which led this study to assess the efficiency of the combined oxidation and adsorption treatment (COAT) process in the treatment of forest polluting site (FPS) leachate by considering local experiments. The removal of effluent parameters (TDS, COD, BOD) was enhanced by oxidizing the GAC surface as a catalyst with NaOH before the process and by ozone within the procedure as well. comparison between COAT process and ozone-GAC was investigated based on variation in ozone dose, H2O2 and pH. Assessing the interacting effect of operating variables (i.e., ozone concentration, GAC density, reaction time and pH) provides valuable information for optimization. Response surface methodology (RSM) and for better prediction, another modeling tool, artificial neural networks (ANN), were employed. Moreover, the results of both software were compared and ANN showed a higher R2 of 0.908, 0.923, and 0.919 for TDS, COD, and BOD respectively in contrast with RSM (R2: TDS = 0.8538, COD = 0.9105, BOD = 0.8108). The optimized model’s circumstances are the reaction time of 30.77 min, ozone dosage of 141.29 mg/l, pH of 7.2, and the GAC density of 1.29 gr/cm3 with the predicted removal percentage of 51.63