Identification of Overpotentials in Vanadium Redox Flow Battery with Reference Electrodes and Determination of Apparent Electrochemical Rate Constants

Abstract

Vanadium redox flow battery (VRFB) is an amateur technology that passed pilot stages in fulfilling the stationary applications at the energy storage market with a high share of renewable power generation. The commercial applications of VRFBs require the optimization of their internal losses and developing the optimal control strategies for in-grid operation. This calls for a deep understanding of the internal VRFB processes, which greatly benefit from both the mathematical modeling and experimental testing for the model validation and refinement.However, the available experimental data for VRFBs is usually limited by open-circuit voltage (OCV) and potential (OCP), charge/discharge voltage profiles at constant load current, full cell polarization curves at the fixed state of charge (SOC), and internal stack ohmic resistances. This also complicates the decoupling and accessing the voltage loss sources, with the prevailing ones usually classified as ohmic, concentration, and activation overpotentials. As such, Aaron et al. (1) showed that decoupling of overpotentials at positive and negative half-cell requires the half-cell potentials measurements with reference electrodes (RE) in addition to full cell data. Different configurations of RE incorporation (2–4) were suggested with a limited agreement with full cell experimental data. The alternative approach with extrapolation of parameters presented in the literature is also challenging due to their strong dependence on an experimental protocol (e.g., the electrode material and its pre-treatment procedure, a membrane, and a setup geometry). As an illustration, the comprehensive review Roznyatsovskaya et al. (5) highlights the four orders of magnitude difference in reported surface area-normalized rate constants. Therefore, the development of a self-consistent experimental setup is crucial for understanding and decoupling internal losses for flow batteries. Combined with VRFB modeling, this may extend accurate predictions of its operation at extreme SOCs and load currents, where voltage losses hamper the efficiency of the battery with the dominating ones usually attributed to concentration (mass-transport in solutions) and activation (sluggish kinetics of reactions on electrodes) overpotentials.Herein, we developed the experimental setup adopting a 5cm2 single flow cell equipped with reference electrodes at positive and negative sides. The experiments were performed in symmetric and full cell designs. The polarization curves were measured at different SOCs [10 – 90 %], electrolyte flow rates [5-30 ml min-1], and load currents. The additional effort was made to compare the results of the experiments in symmetric and full cell designs and extract the apparent kinetic rate constants for positive and negative half-cells. The latter was performed within the 0D modeling approach, which is based on general mass and charge conservation principles. The key feature of the model is the approach for the determination of the integral mass-transfer coefficient on the electrode-electrolyte interface that allows describing non-linear polarization behaviour at high current densities. The effective thermodynamically consistent single-step reaction mechanisms were considered at both half cells, thus decreasing the number of free kinetic parameters. Overall, the proposed approach demonstrated the growing impact of the membrane losses at high current densities. The extracted apparent rate constants under conditions of a real flow battery provide important insights for developing reliable and efficient VRFBs applicable for energy storage systems.1 D. Aaron, Z. Tang, A. B. Papandrew and T. A. Zawodzinski, J. Appl. Electrochem., 2011, 41, 1175–1182.2 H. Lim, J. S. Yi and D. Lee, J. Power Sources, 2019, 422, 65–72.3 S. Ressel, A. Laube, S. Fischer, A. Chica, T. Flower and T. Struckmann, J. Power Sources, 2017, 355, 199–205.4 M. Cecchetti, A. Casalegno and M. Zago, J. Power Sources, 2018, 400, 218–224.5 N. Roznyatovskaya, J. Noack, K. Pinkwart and J. Tübke, Curr. Opin. Electrochem., 2020, 19, 42–48.