Assessment and surface modeling of COD removal from simulated dairy wastewater in a continuous advanced oxidation reactor

The paper herein shows modeling and optimization of a continuous-flow ultraviolet/hydrogen peroxide (UV/H2O2) AOP to degrade chemical oxygen demand (COD) in simulated dairy wastewater. The common conventional treatment procedures, such as biological and physicochemical processes, may be limited by the high amount of sludge formation, high costs, and poor removal of recalcitrant organic contaminants that form. This research, therefore, aims to design a more effective photochemical treatment alternative by conducting experimental studies and developing a more accurate model. A tailor-made cubic photoreactor with double 6 W low-pressure UV lamps (254 nm) was used to investigate the effect of three parameters: the flow rate (10,20, and 30 mL/min), hydraulic retention time (HRT; 10 to 240 min), and initial COD concentration (250 to 1000 mg/L). The studies used 1.2 mL/L H2O2, pH 7.5, at a temperature of 25 ºC. Box-Behnken Design (BBD) and Response Surface Methodology (RSM) were employed, and their optimization of the process and predictive modeling yielded a second-order polynomial regression model. The regression model displayed a high degree of statistical significance (R² = 0.954, Adjusted R² = 0.917; p < 0.0001), while the low root-mean-square error (RMSE = 5.41%) further supports the model’s precision and predictive robustness, accounting for 95.4% of the total variance. At optimal conditions (flow rate 10 mL/min, HRT 240 min, and initial COD 250 mg/L), there was a maximum COD removal efficiency of 82.8% as compared to the standard deviation of 0.37. This increased performance was attributed to the prolonged exposure to oxidation and the elevated availability of hydroxyl radicals in the reduced, low-flow, high-retention flow regimes. The system's operational stability was further indicated by the fact that the values had a low standard deviation during steady-state phases.This work demonstrates that it is viable to use UV/H2O2-based AOPs in continuous reactor systems, which offer a sludge-free, scalable, and energy-efficient approach to treating high-strength industrial effluents. The validated model can be effectively used to predict and optimize real-life processes.