Multiphysics Modelling of an Alkaline All-Iron, All-Soluble Aqueous Redox Flow Battery

Abstract

The development of redox flow battery (RFB) technologies has attracted considerable attention in recent years. Redox flow batteries are electrochemical energy storage devices that operate as flowing systems. Unlike what is possible in conventional batteries, the ability to size the electrolyte storage tanks and electrodes separately enables the battery energy and power capacities to be decoupled and these important properties to be designed and scaled independently. Such systems are particularly attractive for large-scale grid energy storage, especially in conjunction with intermittent energy generation from renewable sources. As RFBs move from research and development to commercial adoption, the use of mathematical models becomes increasingly important for design and analysis of these systems and is indispensable for ensuring their success. Most RFB modelling to date has focused on the all-vanadium RFB, although novel RFBs are continuously investigated and developed. One such novel RFB is the all-iron all-soluble aqueous RFB that is the focus of the present work. This RFB makes use of iron-cyanide (Fe(II)-CN/Fe(III)-CN) and iron-triethanolamine (Fe(II)-TEOA/Fe(III)-TEOA) redox couples in alkaline aqueous solutions. Both redox couples have fast kinetics and the use of high-pH conditions mitigates the loss of current efficiency due to the hydrogen evolution side reaction.

A model has been developed in the present work for the novel all-iron all-soluble aqueous redox flow battery presented by Gong et al. It is the first model to be developed for this RFB. The transient two-dimensional model considers transport of all redox species in the two electrode compartments using porous electrode theory. The side reaction involving the oxidation of TEOA following its permeation across the ion exchange membrane to the positive side is investigated and incorporated into the model. The hydrogen evolution reaction is also incorporated in the model. Parameter values are obtained from literature where available; the remainder of these values are obtained from fitting of the voltage-time curves for charge and discharge to published experimental data. A simulation of a sequence of repeated charge-discharge cycles is conducted and compared with experimental data. The RFB capacity and current efficiency are stable over this duration, which is consistent with experimental observations in the original study. The model has been shown to fit the available experimental data well and describe the behaviour of the RFB. The electrode potentials and reactant species concentrations are found to remain fairly uniform, indicating facile mass transport within the electrode. Recommendations are also made on future experimental and modelling work that can be conducted for this system.