Balance between model detail and experimental information in steam methane reforming over a Ni/MgO‐SiO2 catalyst

The optimization and full understanding of chemical reactions is aided by the construction of an adequate kinetic model. The development of such a kinetic model remains a challenging task. To tackle this challenge in the most efficient way, an iterative, systematic methodology, originally demonstrated for n‐hexane hydroisomerization, is now extended aiming at finding the balance between the envisaged model detail and available information, often originating from time‐consuming and expensive experiments. Steam methane reforming on the Ni/MgO‐SiO2 case study is used for this purpose, that is, the construction of a kinetic model that embeds a maximum amount of information contained in the dataset. The kinetic model is expanded stepwise from a power law model over a model with reactant adsorption toward a Langmuir–Hinshelwood–Hougen–Watson model. The performance of the initially underparameterized model improved significantly by adding reactant adsorption, yet including product adsorption led to overparameterization rather than enhanced model performance. © 2019 American Institute of Chemical Engineers AIChE J, 65: 1222–1233, 2019