To understand the root cause(s) of a field accident in lithium-ion battery product is important because we always need to know what to result in the unexpected failure in order to solve the safety problem. However, to identify the root cause and failure mechanism behind failure(s) in lithium-ion batteries is always a challenge due to two reasons. First of all, the evidence may be burned or destroyed while the failure occurs. And secondly, there are usually multiple contributing factors coming together to result in a battery failure, so the failure mechanism could be very complex to be identified and simulated exactly. To address the complicated situation, the methodology with a series of forensic analysis techniques could be implemented to explore all potential root causes that are related to the weakness or issues in battery design, quality and aging effects.

A practical strategy to identify the potential causes of a battery failure is given in Figure 1. To analyze the failure, all of the detailed background information, such as the environmental conditions, system operation status while failure occurs and all data recorded by the battery management system, will need to be collected as the basis to logistically make the basic hypothesis. To validate the assumptions, the forensic analysis methodology will be the best way to find out the scientific evidences. Furthermore, the simulation of all conditions to duplicate the field event will sometimes be the strong evidence to provide a more solid conclusion regarding the failure mechanisms in a real battery accident. A good example of investigation project using the forensic analysis techniques to study the battery field event is the Boeing 787 battery accident that the aft battery caught fire in January 2013 at Logan International Airport. The cell design was investigated thoroughly by reverse engineering, analytical techniques, non-destructive and destructive analysis, and simulation of abuse conditions to correlate between the root causes and safety behaviors under various operating conditions. All of the key findings and potential root causes are summarized in Figure 2. In this investigation project, the weakness points and potential concerns in the product design, from materials and components to cell and battery levels, were all studied and analyzed because the failure can most likely to be triggered due to the weakness point in the product design or incompatible items between the battery performance and application scope. For example, the battery is expected to be used under low temperature down to -18^oC and high temperature up to 70^oC; however, the conductivity of the electrolyte will reduce 30% due to the increasing in electrolyte viscosity under -18^oC. This will cause the polarization and induce heating effect under extremely high rate application such as the APU engine start in 787 airplanes. In addition, the accelerating rate calorimetry (ARC) study also shows the self-heating in cell can potentially be triggered at low temperature range 60-70^oC. This is also the item that cannot match with the application scope exactly. Besides the issues in the application temperature range, there are also other safety concerns in the battery design, such as the inappropriate design in cell terminals to result in localized hot point, and the uneven stress in compressed jelly-rolls to cause the deformation in jelly-rolls and unbalancing current density distribution. After all potential contributing factors to the battery failure were identified, some simulation work in battery system level, including nail penetration test to study the cell-to-cell thermal runaway path and simulation of APU start cycles to explore the local heating effect in terminals, were also implemented and the outcome of the study can help to successfully find out the potential failure mechanisms and possible root causes in the 787 battery incident.

In this study, we focus on the investigation of the battery failures based upon the analysis in weakness points in battery designs, effects of variation or anomaly in battery quality and the compatibility issues between the battery characteristics and application scope. The key forensic analysis techniques will be introduced and the rationales to link the correlations between failure mechanism and battery failure modes will also be presented. Not only the lesson learned from the Boeing 787 battery, but also some more case studies of failure analysis will further be demonstrated to provide more insights in the safety behaviors for lithium-ion batteries.