Particle-resolved computational modeling of hydrogen-based direct reduction of iron ore pellets in a fixed bed. Part II: Influence of the pellet sizes and shapes

Full-fledged computational modeling of Direct Reduction reactors encompasses single-pellets models and the step-wise scaling up to industrial-scale reactors. This study delves into the particle-resolved computational modeling of hydrogen-based direct reduction of 0.5 kg iron ore fixed-beds, as a scale-bridging step and focusing on the influence of pellet sizes and shapes. To investigate the effect of particle sizes, two beds with particles sized 10.0–12.5 and 12.5–16.0 mm were characterized and numerically reconstructed using the discrete element method. While to assess the effect of particle shapes, a third numerical bed was generated directly upon a high-resolution CT-scan of a real bed. Through reactive CFD simulations, we investigated the reduction process in these three beds using hydrogen as the reducing gas. Our findings reveal that the bed structure significantly impacts the reduction efficiency and overall conversion degree. This study emphasizes the importance of accurate bed reconstruction and provides critical insights for optimizing the hydrogen-based direct reduction process in industrial applications. • Three CFD simulations with varying pellet sizes and comparing spherical vs. real pellets. • Bed with larger pellets reduce slower than small-pellet beds of equivalent weight. • Real bed structure is more compact and rearranged compared to DEM-generated beds. • Shapes considerations impact flow fields locally, but global conversion is similar. • DEM-generated beds are representative enough for modeling the reduction.