Fast & Reliable End-of-Life Evaluation of Li-ion Batteries in Electric Vehicles: A Dual Metric Approach for Ageing Heterogeneity Assessment

Abstract As electric vehicles (EVs) are expected to gradually replace internal combustion engine vehicles, managing end-of-life (EoL) batteries is crucial. This work proposes a two-step framework for state-of-health (SoH) diagnosis of retired EV lithium-ion batteries, determining whether they should be recycled or refurbished for second-life applications. The framework uses an impedance-based sorting algorithm along with a purposeful use of an incremental capacity (IC)-derived method. A standardised protocol was applied to 46 battery elements from two retired Nissan LEAF Gen.1 packs. Unexpected results were observed: elements with high remaining capacity exhibited extreme impedance signatures. Additionally, IC analysis showed a close linear correlation with remaining capacity across several restricted voltage ranges, especially outside the classically studied peak zones. Using impedance features between 1 Hz and 1 kHz, a quantile-based threshold quickly sorts unsuitable batteries for reuse. Linear regression is then used to estimate SoH with an accuracy of about 99% using IC data. Finally, this study underscores the non-equivalence between impedance growth and capacity loss, highlighting the necessity to consider these features as distinct yet complementary indicators of SoH. Additionally, the application of IC beyond the commonly studied peak zones presents a novel approach that may offer new perspectives.