LAUSR

Abstract This paper reports thermal (burner) and mechanical (blunt trauma and nail penetration) abuse experiments on electric vehicle lithium ion modules comprising eight 56.3Ah lithium nickel manganese cobalt (NMC) pouch cells. The aim of project part of which is described in this paper was to study the problem of thermal runaway in lithium ion batteries under different abuse conditions and at different SOC and to bridge the current gap in the literature between cell level studies and research at pack and system level. These experiments were part of an ongoing research programme leading up to studies at pack and system level. The responses of the cells to the various forms of abuse were monitored with optical and thermal cameras, thermocouples and by measuring cell voltage. Draeger gas sensors were also employed where possible. The nail penetration experiments were carried out at (nominally) 96.5%, 75% and 50% SOC, and at 96.5% SOC as a function of penetration location: the experiments strongly suggest that low SOC is as hazardous as high SOC, in contrast to a general perception in the literature, as the likely hazards are simply different and include the possibility of violent vapour cloud explosion. Thus, in all experiments, the first obvious indication of thermal runaway was the ejection of white vapour: if this ignited, the obvious hazard was that of fire. If, however, the vapour did not ignite, it posed an entirely different hazard in terms of high toxicity and the potential for a violent vapour cloud explosion: this is the first mention of such a phenomenon linked to lithium ion batteries in the academic literature. The experiments showed that cell voltage cannot be employed as a reliable warning of thermal runaway. Finally, the data obtained support a wholly novel theory, yet to be adopted across the community, in which thermal runaway can involve the direct solid-state electrochemical reaction between anode and cathode at temperatures ≥ 250 ˚C following venting of the electrolyte.