Identification and quantification of degradation modes in lithium-ion battery cells under dynamic load conditions using equivalent circuit and physics-based models

Identification and quantification of degradation modes (DMs) in lithium-ion batteries are essential for preventing thermal runaway. As the battery ages, losses start to accumulate; however, the three interdependent DMs can be evaluated using non-invasive diagnostic techniques. The DMs have been extensively analyzed under the perspective of a constant current (CC) discharge protocol, whereas aging under real driving conditions remains less explored. This study investigates DMs utilizing an open-source cyclic aging dataset comprising NMC chemistry cells aged under urban driving conditions. The cyclic aging is approximated by fitting both a 2RC equivalent circuit model (2RC-ECM) and a Physics-based model (PBM) to electrochemical impedance spectroscopy (EIS) spectra. The component values estimated from the 2RC-ECM are used to identify and quantify the DMs. Whereas the parameters estimated from the PBM correspond to the changes in the material parameter, allowing for the determination of the changes in material properties resulting from cyclic aging. The results revealed 5.91% conduction losses, 28.16% lithium inventory losses, and 32.1% active material losses. Estimated material parameter changes included a 64% reduction in exchange current density, a 38% increase in double-layer capacitance, and increases of 10%, 20%, and 48% in film resistance, active-material particle radius, and active-material volume fraction, respectively.