Entropy generation model to estimate battery ageing

Determining battery ageing is a key issue in predicting the available charge in battery-operated systems. Beyond the common use of voltage or current measurements, which have proved unsuccessful as general methods, we aim to determine battery ageing in terms of irreversible entropy production. On the one hand, entropy describes transformations, such as ageing, and on the other hand, entropy combines the benefits of voltage, current and temperature effects in a single magnitude. We develop a thermoelectrochemical model based on phenomenological equations to compute entropy generation rates as a function of internal parameters such as concentration, electrode porosity and conduction in NiMH batteries. We simulate ageing by changing internal parameters during battery cycling and we found a threshold in the entropy generation rate of 0.12 mW K-1 cm(-2) and a normalized cumulated entropy of 40% for a capacity decrease of 20% in the studied cell. We conclude that entropy characterization is a promising tool to describe battery ageing because it allows a simultaneous approach to capacity and power fade during ageing.