Foundational Multiphysics Investigation of Loss Effectsin Redox Flow Battery Stacks

Modeling redox flow battery (RFB) stacks is crucial for understanding the interplay of physical, chemical, and electrochemical processes that govern their performance. While the majority of multiphysics models has been focused on materials and small-scale devices [1] only basic lumped circuit or semi-analytical models have been proposed for modelling flow batteries stacks at industrial scale level [2,3]. Beside their simplicity, they often fail to capture the spatially resolved flux distributions and coupled transport phenomena critical for stack-level optimization, notably species crossover and shunt currents, which are irrelevant in single cell investigations. These elusive effects occurring in flow battery stacks received partial attention despite being a major cause of internal losses, directly affecting efficiency and operability [4]. This work presents a foundational analysis of the charge carriers moving in the fluid electrolytes because of the electric potential differences among homologous electrodes and membranes. Taking the vanadium chemistry as a study case, the conductive, diffusive and convective motions of ions V2+, V3+, VO2+, VO2+, H+, HSO4–, SO42– were analyzed resorting to Navier-Stokes, Nernst-Planck and conservation equations in both 3D and 2D numerical models, with the finite element method in the COMSOL Multiphysics® environment. Such strategy allowed the explicit and fine evaluation of interactions between mass transport, charge transport, and electrochemical kinetics, as well as evaluating the effects of design parameters such as channel geometry, electrode porosity, and membrane properties on the overall stack performance. The simulations provided a full operative picture of the device in both steady state and transient conditions extended over some charge-discharge cycles. Key performance indicators such as battery capacity, state of charge, and efficiency were computed. Shunt current and crossover impacts were investigated in stacks of different size and under different loads, using as benchmarks industrial scale cells designed and tested at EESCoLab of the university of Padova [4]. Results indicate that the power losses due to shunt currents range from less than 1% in a 5-cell stack to up to 7% in a 40-cell stack, being more pronounced at lower load currents. Similarly, species crossover effects were higher in larger stacks, significantly impacting on the coulombic efficiency. These results align with experimental data obtained from real stack tests [4]. Additionally, the methodology graded the key factors affecting shunt currents and species crossover, such as membrane permeability and conductivity, electrode porosity and kinetics, and flow channel design. The method here presented paves the way for stack design optimization, bridging the gap between laboratory research and industrial-scale implementation. By enhancing the understanding of intricated interactions, our approach can contribute to the efficiently address scalability and cost-effectiveness of next-generation flow batteries.