Reduced order model for isothermal pre-reduction of chromite pellets with hydrogen

Pre-reducing chromite with hydrogen offers significant benefits, including reduced energy consumption in downstream smelting and lower carbon dioxide emissions. To leverage these advantages, it is essential to develop computational models that facilitate the design, optimisation, and control of the process. Full scale models such as 3D computational fluid dynamics models are often used to investigate the multi-scale flow physics in heterogeneous reactors. However, these models impose prohibitive computational expense in the modelling of heterogeneous reacting flow systems. The high computational expense prevents the use of high dimensional systems in real time control and online optimisation of reacting flow systems. In order to facilitate incorporation of predictive models in the heterogenous reacting flow systems control loop, computationally efficient reduced order models are required. Owing to these requirements, a reduced order model for pre-reduction of chromite with hydrogen was developed with the view to rapidly predict the performance indicators for pre-reduction of chromite with hydrogen. The Thiele modulus method was used to develop the reduced order model, and the model results were compared to isothermal pre-reduction experimental results at a range of pre-reduction temperatures from 1200 oC to 1400 oC. The model predicts the degree of reduction, diffusion of gases through the product layer, unreacted core radius of the grain radius, and local and global conversion. The final conversion across all temperatures as predicted by the reduced order model deviate from the experimentally calculated conversion by a maximum relative error of 2.7% while the initial conversions differ significantly, reaching a maximum error of 65%.