Given their capability of spreading active chemical species and collecting electricity, porous media made of carbon fibers are extensively used as diffusion layers in energy storage systems, such as redox flow batteries. In spite of this, the dispersion dynamics of species inside porous media is still not well understood and often lends itself to different interpretations. Actually, the microscopic design of efficient porous media, which can potentially and effectively improve the performances of flow batteries, is still an open challenge. The present study aims to investigate the effect of fibrous media micro-structure on dispersion, in particular the effect of fiber orientation on drag and dispersion dynamics. Several lattice-Boltzmann simulations of flows through differently oriented fibrous media coupled with Lagrangian simulations of particle tracers have been performed. Results show that orienting fibers preferentially along the streamwise direction minimizes the drag and maximizes the dispersion, which is the most desirable condition for diffusion layers in flow batteries’ applications.

We present lattice Boltzmann pore-scale numerical simulations of solute transport and reaction in porous electrodes at a high Schmidt number, $Sc=10^{2}$ . The three-dimensional geometry of real materials is reconstructed via X-ray computed tomography. We apply a volume-averaging upscaling procedure to characterise the microstructural terms contributing to the homogenised description of the macroscopic advection–reaction–dispersion equation. We firstly focus our analysis on its asymptotic solution, while varying the rate of reaction. The results confirm the presence of two working states of the electrodes: a reaction-limited regime, governed by advective transport, and a mass-transfer-limited regime, where dispersive mechanisms play a pivotal role. For all materials, these regimes depend on a single parameter, the product of the Damköhler number and a microstructural aspect ratio. The macroscopic dispersion is determined by the spatial correlation between solute concentration and flow velocity at the pore scale. This mechanism sustains reaction in the mass-transfer-limited regime due to the spatial rearrangement of the solute transport from low-velocity to high-velocity pores. We then compare the results of pre-asymptotic transport with a macroscopic model based on effective dispersion parameters. Interestingly, the model correctly represents the transport at short characteristic times. At longer times, high reaction rates mitigate the mechanisms of heterogeneous solute transport. In the mass-transfer-limited regime, the significant yet homogeneous dispersion can thus be modelled via an effective dispersion. Finally, we formulate guidelines for the design of porous electrodes based on the microstructural aspect ratio.

Large-scale introduction of electric vehicles (EVs) to the market sets outstanding requirements for battery performance to extend vehicle driving range, prolong battery service life, and reduce battery costs. There is a growing need to accurately and robustly model the performance of both individual cells and their aggregated behavior when integrated into battery packs. This paper presents a novel methodology for Lithium-ion (Li-ion) battery pack simulations under actual operating conditions of an electric mining vehicle. The validated electrochemical-thermal models of Li-ion battery cells are scaled up into battery modules to emulate cell-to-cell variations within the battery pack while considering the random variability of battery cells, as well as electrical topology and thermal management of the pack. The performance of the battery pack model is evaluated using transient experimental data for the pack operating conditions within the mining environment. The simulation results show that the relative root mean square error for the voltage prediction is 0.7–1.7% and for the battery pack temperature 2–12%. The proposed methodology is general and it can be applied to other battery chemistries and electric vehicle types to perform multi-objective optimization to predict the performance of large battery packs.