Prospect of modeling industrial scale flow batteries – From experimental data to accurate overpotential identification

A vast literature exists on modelling of small-scale single-cell experiments for flow batteries, but very few reports have been published on large stacks, consisting of tens of cells, each with an active area of hundred square centimeters. In this report, a large set of measurements taken on a kW-class vanadium test facility is used to develop an accurate ad-hoc physical model. Experimental data consist of polarization curves at a broad range of states of charge and electrolyte flow rates, as well as electrochemical impedance spectra. The model is capable to decouple three sources of voltage losses: activation, ohmic, and concentration overpotentials. In addition, a new numerical approach for identifying the main parameters of electrochemical kinetics and mass-transport has been proposed. To the best of our knowledge, this is the first time when a model for voltage losses analyses was developed and fitted to a data from large-scale flow battery, being validated with a sensitivity analysis study. Investigations showed that activation losses have a sophisticated nature in combination with mass-transport limitations and play an important role in a wide range of loading current densities. As a result, they must be included in a reliable model able to reflect a non-linear voltage behavior of large stacks. This work also highlights that, in the investigated case, activation losses are likely to be attributed to the positive electrode rather than to the negative one. The obtained results could facilitate the development of advanced simulation and design of large-scale flow battery stacks.