Realistic lifetime prediction approach for Li-ion batteries

A novel methodology for lifetime prognosis of Li-ion cells is presented, covering the validity and accuracy assessment of the predictions. It is especially focused on the procedures for the evaluation of semi-empirical ageing model development process. Combined calendar and cycle ageing is investigated with dynamic and realistic complex operation profiles, as part of the stepwise validation methodology that is proposed. The dynamic validation approach allows identifying the sources of model errors by analysing how the individual stress factors are adjusted under non-constant profiles. The results correspond to a LFP-based 26650-size cell. The presented model is suitable for different applications, such as elevator or vehicle combined with vehicle-to-grid (V2G) use, with 1.4% root-mean-square error accuracy.