At the royal institute of technology (KTH) in the department of applied process metallurgy, a novel modelling approach has been developed which allows a dynamic coupling between the commercial thermodynamic software Thermo-Calc and the commercial computational fluid dynamic (CFD) software Ansys Fluent, only referred to as Fluent in the study. The dynamic coupling approach is used to provide numerical CFD-models with thermodynamic data for the thermo-physical properties and for the fluid-fluid chemical reactions occurring in metallurgical processes. The main assumption forthe dynamic coupling approach is the existence of local equilibrium in each computational cell. By assuming local equilibrium in each computational cell it is possible to use thermodynamic data from thermodynamic databases instead of kinetic data to numerically simulate chemical reactions. The dynamic coupling approach has been used by previous studies to numerically simulate chemical reactions in metallurgical processes with good results. In order to validate the dynamic coupling approach further, experimental data is required regarding surface reactions. In this study, a graphiteand metallurgical coke oxidation experimental setup was suggested in order to provide the needed experimental data. With the experimental data, the ability of the dynamic couplings approach to numerically predict the outcome of surface reactions can be tested.By reviewing the literature, the main experimental apparatus suggested for the oxidationexperiments was a thermo-gravimetric analyzer (TGA). The TGA can provide experimental data regarding the reaction rate, kinetic parameters and mass loss as a function of both temperature and time. An experimental setup and procedure were also suggested.In order to test the ability of Fluent to numerically predict the outcome of surface reactions, without any implementation of thermodynamic data from Thermo-Calc, a benchmarking has been conducted. Fluent is benchmarked against graphite oxidation experiments conducted by Kim and No from the Korean advanced institute of science and technology (KAIST). The experimental graphite oxidation rates were compared with the numerically calculated graphite oxidation rates obtained from Fluent. A good match between the experimental graphite oxidation rates and the numerically calculated graphite oxidation rates were obtained. A parameter study was also conducted in order to study the effect of mass diffusion, gas flow rate and the kinetic parameters on the numerically calculated graphite oxidation rate. The results of the parameter study were partially supported by previous graphite oxidation studies. Thus, Fluent proved to be a sufficient numerical tool for numerically predicting the outcome of surface reactions regarding graphite oxidation at zero burn-off degree.