Data-based optimization approach for the automatic operation of solar photo-Fenton plants

This work showcases the main design guidelines for implementing a data-based decision-making approach for the automatic operation of solar photo-Fenton plants. The case study explores the removal of microcontaminants (MCs) from an urban wastewater treatment plant secondary effluent using a 100-m² raceway pond reactor (RPR). Automatic system design was based on the computational simulation of the plant, using a kinetic model of the photo-Fenton process at acidic pH. The model was validated for the continuous flow operation of the 100-m² photoreactor. Afterwards, a simulation study was conducted to correlate MC removal yield, residual H₂O₂ concentration and operating cost with the main experimental factors using polynomial models. These models were then incorporated into an optimization routine to determine optimal plant configuration for different environmental scenarios (UVA irradiances and RPR temperatures). As a result, two data-based self-automatic procedures were obtained for 80-% and 90-% MC removal targets, operating cost minimization being the optimization target. Plant performance was evaluated in manual and automatic mode for both clear and cloudy days. The simulation results showed the robustness and reliability of the automatic solution under adverse environmental conditions, achieving stable treated water quality throughout the operation. Furthermore, improvements in cost efficiencies and water treatment capacities of up to 34.4 % and 61.6 % were attained, respectively, leading to reductions in the operating cost of up to 26.2 %. The results delivered highlight the potential of a reliable and easy-to-implement automatic operation strategy for immediate application in existing solar photo-Fenton plants. This approach paves the way for treatment automation, promoting further research on its commercial scale-up.