Speaker
Description
Hydrogen-based direct reduction (DRI) is a key route to eliminating CO$_2$ emissions from iron and steel production. Reactor-scale models of hydrogen DRI rely on effective transport properties such as permeability, pressure drop, and heat and mass transfer coefficients that emerge from complex flow through packed beds of porous iron ore pellets. To better understand and parameterize these pellet-scale transport mechanisms, detailed CFD simulations of hydrogen flow through idealized pellet-scale unit cells were performed.
In this work, iron ore pellets are represented as porous bodies with an internal porosity of 0.22, embedded in a periodic computational cell. Hydrogen flow, at 1200K, is driven by a prescribed pressure jump across the cell, and the resulting velocity and pressure fields are solved using a finite-volume CFD solver. Two idealized pellet arrangements are compared using periodic unit cells: a body-centered cubic (BCC) configuration with a central pellet and a face-centered cubic (FCC) configuration with four pellets surrounding a central void.
Despite the geometric simplicity, the simulations reveal that the flow does not distribute uniformly through the pore space. Instead, hydrogen organizes into a few dominant high-velocity channels that connect the inlet and outlet across the periodic cell, while other regions remain weakly flushed. In the BCC configuration, the central porous pellet increases resistance along the cell centerline and diverts most of the flow into side channels. In the FCC configuration, the absence of a central pellet creates a more open vertical pathway through the four-pellet junction, resulting in a narrower but more intense high-velocity core and a higher overall pressure drop for the same pressure driving force.
These results highlight that pellet-scale flow in hydrogen DRI beds is governed by the topology of connected flow channels rather than by local gap width alone. The emergence of a small number of preferential high-velocity channels has direct implications for upscaling: these channels are expected to dominate both the effective permeability and once coupled with energy and species transport, the pellet-scale heat and mass transfer rates. The present pellet-scale CFD framework thus provides a physically resolved basis for calibrating Ergun-type correlations and effective transport coefficients used in reactor-scale models of hydrogen-based DRI. In the next step, non-idealized pellet bed configuration will be studied, and it would also capture the heat transfer between hydrogen flow and iron-ore pellets.
| Country | Sweden |
|---|---|
| Acceptance of the Terms & Conditions | Click here to agree |
