Effect of the Particle Size Averaging Method on the Predicted Behavior of Lithium-ion Batteries According to the Pseudo-2D Model

Abstract The pseudo-2D (P2D) model is widely utilized in the modeling of lithium-ion batteries for product design and state monitoring applications. The P2D model relies on an average electrode particle size as an input parameter, but existing literature has been inconsistent as to how this is obtained. In this work, we examined the effect of the particle size averaging method on the behavior of Li-ion batteries as predicted by the P2D model. We first set up a standard P2D model that only considers representative sizes and a modified P2D model that accounts for the entire size distribution. Number average, volume-surface average, and D50 particle sizes were then obtained for graphite anodes and lithium iron phosphate and nickel manganese cobalt oxide cathodes typically found in commercial lithium-ion cells. Lastly, the results generated by monodisperse and polydisperse models were compared. It was found that the volume-surface average and D50 sizes accurately predict the energy density and cell capacity with respect to the polydisperse model, while the number average particle size may result in notable discrepancies. For this reason, the number average particle size, or getting a simple average particle size from scanning electron microscopy images, should be avoided.