Variance-Based Global Sensitivity Analysis: A Methodological Framework and Case Study for Microkinetic Modeling

Chemical models are built up from chemical reactions and parameters. Each of these parameters has a degree of uncertainty. Sensitivity analysis has proven to be an important tool to quantify and trace this uncertainty to specific input parameters. In this study, the methodology of a prominent global sensitivity analysis method, that is, Sobol's variance-based method, is presented for chemical modeling with a focus on microkinetic modeling. Sobol's method is developed to be used as an analysis framework, which-once set-up for microkinetic modeling-can easily be used for different models. This analysis framework is successfully demonstrated by means of two case studies from the field of microkinetic modeling: 1) CO oxidation and 2) oxygen evolution reaction (OER) at the photoanode in a photo-electrochemical cell. The results give insight into the influence of each input parameter on the output uncertainty. For CO oxidation, it is found that the temperature and chemisorption energies have most impact on the output. For the OER model, the valence band energy and solvent reorganization energy are most influential. Based on this, a workflow is proposed incorporating the sensitivity analysis into the modeling process, aimed at reducing the output uncertainty and at validating and optimizing the model.