Understanding Lithium-Ion Battery Degradation through P2D Model Parameters Evolution

Abstract

Despite its commercial level, lithium-ion battery ageing is still a hot topic in the research community. Indeed, the relation between ageing and the operating conditions and loads in a real-life operation is under investigation and the corresponding residual life prediction is still a challenging task.In this work, a combined approach with experiments and modeling is applied. A wide driving cycle campaign is carried out, involving 12 different profiles to test different combinations of temperature, C-rate, SoC and DoD, whose values were defined after a literature review about the most common values in a real application, and highlight their effect on battery ageing. Two calendar ageing campaigns at the same temperatures are also conducted to quantify the acceleration of the degradation rate due to cycling and analyse if the degradation path is different.For the interpretation of the results, the P2D model is employed. No explicit description of the main degradation mechanisms is applied. Indeed, the degradation is described through the variation of the physical parameters of the usual model implementation, like a reduction of lithium inventory or active electrode material or a variation of reaction rates and diffusion constants. A specific sequence of tests has been developed as check-up procedure. As detailed in 1, it is a trade-off between testing time and the operating conditions that maximise the identifiability of the parameters of the model.Therefore, the parameter identification problem is performed periodically upon check-ups and parameters trend over time/cycles is derived. Since it is a physical model, every parameter has a physical meaning and its evolution can be associated to a certain range of degradation phenomena. In this way, the model can support the identification of the most likely phenomena that are ongoing in the battery and the value of the parameters constitutes a new concept of state of health, more comprehensive than the residual capacity value.Finally, the model interpretation is proved in two ways. First, a validation dataset is provided, to evaluate the capabilities of the calibrated model in reproducing the behaviour of an aged cell under different operating conditions with respect to the check-up procedure. Then, some cells undergo post-mortem analyses to verify the origin of degradation and measure the values of the physical parameters, like optical and chemical analyses, q-OCP discharges of harvested electrodes in coin cell setup for thermodynamic parameters or EIS for kinetic parameters.This approach, though not suitable on-board, provides a novel physics-based method for a reliable analysis of lithium-ion batteries degradation. 1 G. Sordi, C. Rabissi, A. Casalegno, Reliable thermal-physical modeling of lithium-ion batteries: consistency between high-frequency impedance and ion transport. Under Review