Modelling, Simulation and Experimental Characterisation of Hydrogen-Bromine Flow Batteries

Kurzfassung

Large-scale energy storage systems are considered one of the key components of power grids based on renewable energy resources. The electricity supply and demand are normally in balance owing to the grid inertia of e.g. rotational energy of steam turbines. With increasing shares of renewables in the energy mix which provide electricity intermittently, the supply and demand fall out of equilibrium and, to prevent blackouts and brownouts or frequency variations, the deployment of large-scale energy storage systems is necessary. Batteries represent just one category within the realm of energy storage systems, with lithium-ion technology reigning as the most prevalent choice. However, due to lithium-ion batteries’ finite cycle lifespan, safety concerns, reliance on limited natural resources, and constraints on energy density, there are many incentives to seek for alternative approaches. Flow batteries address the lithium-ion technology’s issues as they store the energy in the form of liquid electrolyte in separate tanks. One of the promising flow battery systems is the hydrogen-bromine flow battery (HBFB) system which combines high power densities with low cost due to inexpensive constituents of the electrolyte: hydrogen gas, bromine and hydrobromic acid. To better understand the working principles, determine the desired operating conditions and to optimise the HBFB, modelling and simulation tools are utilised. However, a proper parametrisation of such models remains a key challenge on the way to accurate performance prediction. In this doctoral thesis, three particular topics pertaining to the HBFB system modelling and characterisation are addressed in detail: the thermodynamics of the bromine-based electrolytes, mass transport properties of electroactive species in concentrated, aqueous solutions and modelling and simulation of porous electrodes. In the first part, a considerable effort is devoted to the modelling and experi�mental validation of the open-circuit potential (OCP) of the bromine electrode and the open-circuit voltage (OCV) of the HBFB cell. It is shown how solution thermodynamic activity and formation of polybromides affect the predicted OCP values and how different they are from the ones modelled with a basic Nernst equation. An improved, physics-based OCV model is described and validated experimentally in a specially designed setup, offering a maximum prediction error reduction from 20% to less than 1% compared to the basic Nernst equation. The second part deals with mass transport by diffusion and migration of electroactive species in the HBFB posolyte: dissolved bromine and bromide ions. The impact of elevated electrolyte concentrations of more than 8 moles per litre in HBr on the effective diffusion rates is determined empirically with the use of ultramicroelectrodes in a dedicated experimental setup, and theoretical equations to model the limiting current densities of bromide and bromine are proposed which showed satisfactory agreement. Lastly, the problem of upscaling of theoretical equations used to model the performance of porous electrodes (PE) is described. Pore-scale models are able to capture local variations of the variables of interest, however they suffer from high computational cost. Here, a mathematical upscaling method (the volume averaging method, VAM) is used to computationally analyse several different PE geometries and to elaborate effective transport parameters. As shown in the thesis, they can be then used in macroscopic flow battery models to save on computational effort while still capturing the impact of pore-scale effects such as pore geometry or the impact of electrochemical reaction on the effective diffusion coefficient. Here, the novelty is the upscaling of coupled diffusion-advection equation with a full Butler-Volmer-type heterogeneous reaction at the solid-liquid interface.