The supply of hydrogen peroxide (H2O2) controlling the amount of highly oxidant hydroxyl radicals is the most critical operational issue for the photo-Fenton process. Accordingly, this study addresses the development of a model for photo-Fenton processes including a flexible H2O2 supply given as a function of time. The model is aimed at its future exploitation in treatment optimization and the determination of the optimal profile for H2O2 supply. The work has adopted a photo Fenton model previously reported that includes the inlet flow (fed-batch) and describes the system dynamics under a flexible dosage. Thus, model global sensitivity analysis (GSA) and parameter estimation were performed using Simulink® to examine the behavior of the model under flexible H2O2 dosage. GSA was carried out using partial rank correlation methods and the Latin hypercube sampling to assess to which extent variations of the model parameters affect each measured response (H2O2, total organic carbon, TOC, and dissolved oxygen, O2 –the experimentally available information). Hence, the model is discussed in regard of its hypothesis and chances for reducing its complexity. This resulted in the rejection of two reactions of the initial model. Next, a set of 12 kinetic, stoichiometric and operative parameters was estimated through the fitting of H2O2, TOC, and O2 profiles. Discussion on model fitting includes computational issues, the role of initial values for the estimation process, the goodness of fit criteria, and the sampling method. The model was fit to experimental data with assorted H2O2 supply profiles and validated, and Root Mean Square Error (RMSE) below 0.009 mM, 0.42 mM, and 0.127 mM were obtained for TOC, H2O2 and O2, respectively. Therefore, this work contributes a practical model aimed at providing model-based optimization for the H2O2 dosage profile of the photo-Fenton process.
