Modeling Capacity and Power Fade of Cylindrical NCA-Si Graphite Battery Cell in Electric Vehicles

Abstract

The energy storage device in most electric vehicles today are lithium-ion batteries. Particularly, Tesla electric vehicles uses the lithium ion chemistry of nickel cobalt aluminum oxide (NCA) and graphite electrodes. Their aging and lifetime are critical issues that must be addressed for a successful large-scale commercialization of electrified vehicles. Therefore, the present work focus on the study of possible aging mechanisms that contribute to the capacity fade and impedance increase of NCA and Si-graphite battery cells in a modeling framework. The batteries are subjected to partial cycling at different state-of-charge (SOC) ranges to imitate different driving behaviors. By predicting the aging parameters from separated electrodes through model, it is possible to understand the responsible aging mechanisms in such applications. The NCA/Si-graphite cylindrical cells were calendar-aged and cycle-aged between SOC of 20-50 %, 35-65 %, 60-90 % and 20-95 % until end of test (EOT) is reached. The cells electrochemical characterization methods include electrochemical impedance spectroscopy (EIS) at 100-cycle interval to measure the impedance behavior and slow cycle discharge to measure the capacity available. We use two models in analyzing the aging behavior through fitting and validation with experimental data – time-based model for capacity fade and frequency-based model for impedance increase. They are physics-based Doyle-Fuller-Newman (DFN) model, which is based on porous electrode theory. The models account for (electrochemical) aging processes within the electrode. Specifically, the solid-electrolyte interface (SEI) build-up, surface and particle cracking; and electrolyte oxidation on the cathode particle interface. These lead to the loss of lithium inventory (LLI) and loss of active material (LAM), contributing to the capacity fade. The same aging mechanisms are represented in the impedance model in the form of charge transfer resistance, interfacial resistance and diffusion time constant. The challenge in modeling the full cell impedance is the superimposition of aging parameters of roughly the same time constants, making it quite hard for the deconvolution of anode and cathode aging. The deduced aging mechanisms from the modeling work are supported by experimental observation, e.g. scanning electron microscope (SEM). For example, micro-cracks and isolation of particles observed from SEM explain the high possibility of increase in LAM and structural change. Overall, modeling of the battery’s capacity and power fade can deepen our understanding of aging processes, allow lifetime prediction, or optimal battery design.