Kinetic Modelling of Pure Iron Oxide Powder Reduction with Hydrogen

The aim of this doctoral thesis is to model key phenomena governing the reduction of pure iron oxide powders in the presence of hydrogen based on novel experimental data. This objective is strongly associated with the urgent need to reduce the carbon footprint of metallurgical processes involved in large-scale iron and steel production, which remains one of the industrial sectors most in need of ecological transition.

The main goal was to elucidate and describe the fundamental aspects of the series of reductions that characterize the reactive network through a rigorous mathematical framework based on a dynamic shrinking-core model. The ultimate purpose was to identify the controlling regimes and to estimate and validate the associated kinetic parameters. The results of this study provide a solid basis for the scale-up of more complex reactor systems that may serve as a realistic alternative to current production chains or contribute to their optimization by increasing hydrogen use in existing facilities.

A research strategy was developed that alternated between mathematical modelling and experimental investigation, including a detailed characterization of solids under transient conditions. Such an approach is crucial, since mathematical models must consistently describe emerging phenomena through sound assumptions, while simulations—however preliminary—can support the interpretation, inspire the design of new experimental campaigns, and confirm the key mechanisms of the process.

In the first stage, an extensive review of the scientific literature was conducted to assess the wide experimental and modelling variability associated with hydrogen-based iron-oxide reduction. This survey highlighted the heterogeneity of kinetic parameters reported over the past six decades and the consequent need to establish an experimental methodology capable of generating accurate and reproducible data for reliable regime and kinetic evaluation.

Accordingly, reduction experiments were carried out in a fixed-bed reactor loaded with pure hematite and magnetite powders, using a chemisorption analysis device equipped with a thermal conductivity detector for exhaust-gas analysis. The experimental work was supported by the development of a comprehensive mathematical model capable of describing both the reactive solid-bed dynamics and the reaction of individual particles, including the motion of multiple reaction fronts. To ensure physical consistency, a mathematical constraint was introduced to prevent the overlap of the reaction fronts. The final model, which combines a classical axial-dispersion formulation with the shrinking-core model, was used to describe the solid conversion over time, with its consistency and flexibility evaluated through an in-depth parametric analysis.

A dedicated comparison between the experimental results of thin-layer reduction tests and the corresponding model simulations was then carried out. For this purpose, an innovative numerical-mathematical method was developed to correlate the evolution of the thermal conductivity of the exhaust gases with the overall reduction degree of the solids. By incorporating granulometric data from sample analysis, the model enabled the precise evaluation of the kinetic parameters for each reaction step involved in the reduction sequence.

Finally, the estimated kinetic parameters were successfully validated through an extensive experimental dataset obtained from the reduction of iron-oxide beds under a wide range of operating conditions. The mathematical model thus provided a refined microscopic description of the reduction process, elucidating the implications of the reversible nature of the reactions and offering a robust interpretative framework for future hydrogen-based ironmaking technologies.