Computational Modeling of Iron-oxide Pellets Reduction Using H2 in a Fixed Bed

Abstract

Full-fledged computational modeling of Direct Reduction (DR) reactors encompasses single-pellets models and the step-wise scaling-up to industrial-scale reactors. The specific focus lies here on the scale-up from a single iron ore pellet to a fixed-bed reactor model. However, challenges include the generation of synthetic packed-bed structures due to cost limitations, difficulties in generating good quality mesh for multi-pellet beds, and the cost-intensiveness of scaling up to a Computational Fluid Dynamics (CFD) environment. Furthermore, the correct modeling of transport and kinetics-related processes for a single pellet is a prerequisite for a meaningful scale-up. This has not yet been demonstrated. Therefore, in this work, the chemistry and transport data for the reduction of single iron oxide pellets with hydrogen (H2), obtained from a previously developed 1D solid porous model, will be used. The purposes of this article are (1) Proposing a CFD model that reproduces single-pellet reduction experiments with H2 for a wide range of experimental conditions in a 3D-CFD environment. (2) Computationally generating a random packing of 212 industrial pellets (0.5 kg) by applying the discrete element method (DEM) to simulate a lab-scale fixed-bed reactor. (3) Creating a 3D-domain, based on the particle position data from the previous step, and meshing the pellets and the voids among them in different refinement levels. (4) Reproducing a multipellet fixed-bed experiment with pure H2 from the literature. (5) Investigating the effects of temperature variations in the bed. In this way, the concept of scaling up to multi-pellet fixed-bed simulations with H2 are demonstrated successfully.