A correlated multi-observable assessment for vanadium redox flow battery state of charge estimation : long-term stability of methods

Abstract

Accurate half cell specific state of charge (SOC) monitoring is needed for vanadium redox flow batteries (VRFB) to quantify electrolyte imbalances and to increase operation efficiency by appropriately timed re-balancing. However, electrolyte crossover shifts the vanadium concentrations and can impact the SOC estimation in combination with side reactions in the long run. As a result, the long-term stability of any SOC estimation is crucial for its accuracy, while a rigorous cross-evaluation of SOC monitoring methods for VRFBs is still missing in the literature. In this study, we investigate the long-term stability for SOC monitoring based on half cell electrolyte potentials, electrolyte densities, electrolyte volumes, electrolyte viscosity related pressure drops, pH related potentials and UV/Vis absorbances. We conduct 500 h long-term charge/discharge cycling for a membrane with 100 µm thickness and 300 h for a 50 µm thin membrane. For each cycle, we determine reference SOCs validated by potentiometric titration, which enables us to calculate calibration parameter for each SOC estimation method in each cycle. We analyze the variation and stability of these parameters and assess the long-term accuracy of each method. The highest long-term stability is shown by the UV/Vis absorbance measured at 400 nm in the negative half cell with estimation errors below 2 % and the electrolyte half cell potentials with errors below 2.5 %. The UV/Vis absorbance measured at 580 nm in the positive half cell exhibits a moderate stability with estimation errors lower than 5 %. The electrolyte densities and pressure drops measured in the negative half cell show moderate stabilities with estimation errors below 7.5 % which could allow for accurate SOC estimation with more frequent re-calibrations.