Vanadium Flow Battery Electrochemistry and Fluid Dynamics Model with in-Situ Current Distribution Validation

Abstract

It has been estimated that the high penetration of renewable electricity generation (68%-73% from solar and wind) could be possible in the United States by 2050. To achieve this goal, large-scale energy storage technologies will play important role. Energy storage technologies can help to improve electricity stability, flexibility, reliability and resilience on the grid. Among different energy storage methods vanadium redox flow batteries (VRFBs) have been a recent focus of interest due to distinct advantages, i.e., flexibility and scalability, high columbic efficiency, and long cycle life. However, high system cost is a main challenge for widespread commercialization of VRFBs. Although the component costs (membrane cost and chemical cost) tend to dominate the overall system costs, it is possible to decrease costs by improving electrolyte utilization and system efficiency. This can be achieved through increasing vanadium solubility in the solvent, modifying membrane and minimizing crossover, facilitating electrochemical kinetics, improving mass transport and reducing parasitic pump losses. However, VRFB is a complex system with multiple interrelated parameters (mass transport and electrochemical reactions) affecting electrochemical cell performance; these interrelated parameters are difficult to disentangle experimentally. A comprehensive and suitably validated mathematical simulations can both help to understand complex phenomena inside the VRFB cell and provide practical knowledge for controlling and optimizing VRFB systems. In this study, three dimensional, steady-state multi-physics model is developed for VRFB strip cell architecture under the dilute solution theory assumption. The simulation domain includes membrane, electrodes, flow plates and current collectors. Continuum equations: conservation of mass, momentum, species and charge coupled with Butler-Volmer kinetics are employed. Polarization curve analysis and fully segmented, printed circuit board (PCB)-based, localized current distribution measurements are employed to validate the mathematical model. Good agreement is achieved between the model predictions and the experimental measurements. The model successfully predicts not only the charge-discharge polarization curve but also the current distribution along the channel. Figure 1