Application of Sensitivity and Flux Analyses for the Reduction of Model Predicting the Photooxidative Degradation of an Azo Dye in Aqueous Media

The prediction of the system behavior is of significant interest when evaluating appropriate technologies for wastewater treatment. The robust prediction could be achieved by empirical mathematical modeling techniques, but they do not include steps in degradation of organic pollutants. Mechanistic models (MM) include chemical/physical phenomena, but may also include numerous reactions resulting with the complicated kinetic expressions with large number of parameters. This modeling approach can be challenging for complex system such as advanced oxidation processes. With the goal to reduce the number of reactions involved in developed MM, keeping the high prediction power, sensitivity, and flux analyses was employed. The results showed that MM describing the degradation of organic dye in water matrix by photooxidation processes can be significantly simplified, by reducing the number of reactions included without affecting the predictive power. The calculated root mean square deviation values between data predicted by MM and reduced MM_R differ insignificantly (≤1.4 %).