The soluble lead flow battery: Image-based modelling of porous carbon electrodes

A novel numerical modelling framework coupling physics-based model equations and image-based input parameters is developed to simulate the behaviour of the soluble lead flow battery when reticulated vitreous carbon (RVC) electrodes are used. Experimental results are presented to validate the model. Open-source software OpenImpala is used to predict the macro-homogeneous properties of RVC from computed tomography scans of various grades of RVC. The process is repeated on manipulated datasets where a voxel dilation technique has been used to estimate the geometry of RVC electrodes with a range of thicknesses of electrodeposited material. The model predicts that with a region of free electrolyte dividing the electrodes, the electrolyte velocity is low within the electrodes. This is exacerbated by a build-up of deposit close to the inlet. By dividing the electrodes with only a porous separator, a deposit build-up is no longer seen, and the concentration within the electrodes is shown to be far more even. Finally, with an applied current density of 50 mA cm−2, the overpotential is predicted to be reduced by over 100 mV when 100 ppi RVC electrodes are used instead of 10 ppi electrodes. An experimentally validated voltage efficiency of over 80% is achieved.