Abstract
Sustainability is a thriving global topic of concern and following the advancement of technological progress and increased standards of living, the demands for energy, fuels, chemicals and other requirements have increased significantly. Methanol is one such chemical which has seen increases in demand due to its importance as a precursor in the development of widely used chemicals such as formaldehyde. In order to gain insight into the reaction mechanisms driving the process, it is beneficial to develop kinetic models that accurately describe the system for several reasons: (i) to develop process understanding; (ii) to facilitate control and optimisation; (iii) to reduce experimental burdens; and (iv) to expedite scale up and scale down of processes. Two commonly used kinetic reaction rate models are the power law and Langmuir-Hinshelwood expressions, however the strong assumptions made when developing such models may limit their predictive performance through the introduction of inductive bias (i.e. model structural uncertainty). A solution to counter these drawbacks is known as hybrid modelling where, the inauguration of a data-driven component within the kinetic modelling framework allows for any complex, less understood kinetics to be instead learnt from historical data by a machine learning model. In order to identify the pros and cons associated with each kinetic and hybrid modelling strategy for chemical reaction network modelling, a thorough comparison was made using syngas methanol production as a case study. It was shown that hybrid models offered increased predictive accuracy, robust uncertainty quantifications, and improved generalisability under limited data availability.