The ever-growing societal move towards sustainable forms of transportation, such as electric vehicles, motivates the research behind safe, efficient, and cost-effective use of lithium-ion batteries. Consequently, careful modelling of the electric, thermal, and ageing behaviour is paramount for all real-world applications.
In the last decades, modelling of batteries using lumped equivalent electric circuit models has gained prominence in the research community. This can be attributed to intuitive, explicable results, and relatively accessible measurement procedures based on electrochemical impedance spectroscopy (EIS). State of charge (SOC), state of health (SOH) and thermal factors such as the effect of the ambient temperature, heat-up during operation, and resulting uneven spatial temperature distributions, called temperature gradients, can also be parametrically factored into an electrical model. Resulting electrical models are used to rapidly simulate or control lithium-ion batteries during operation via battery management systems.
Despite the causes and effects of the ageing phenomena being relatively well understood in the literature, little light has been shed on their effect on the electrical circuit model parameters under various ageing conditions and temperature gradients. The aim of this work has therefore been to carry out a quantitative analysis on the behaviour of select electrical model parameters under said conditions.
This work uses a physics-motivated impedance-based electrical battery model. It contains twelve elements representing lumped circuit elements corresponding to electrochemical properties and processes. The selected parameters for the analysis are the series resistance of the battery and charge transfer resistances of the anode and the cathode. The studied electrical model is parametrized based on an EIS measurement campaign consisting of five discrete temperature gradients and an approximate ageing down to 80% of initial capacity.
Preliminary results show physically interpreteble behaviour of electrical model parameters over equivalent full cycles. This behaviour is analysed over a three dimensional parameter field consisting of various SOCs, degrading SOH, and discrete temperature gradients of up to 30 K. Observed divergent ageing behaviour of the anode and the cathode is supported by a post-mortem analysis of the cell. Moreover, a generalized method for electrical model parameter behaviour evaluation is presented and ageing functions best fitting the each parameter are empirically selected, evaluated and analysed. Finally, areas of further research are identified and an optimized measurement procedure for the likes of this work’s analysis is put forward.