Characterization and Analysis of Lithium-Ion Battery Fire Signatures and Debris

Fire is a tool that criminals use for fraudulent activities or to obscure evidence of a crime. Fire forensic reconstruction, however, is complex and is one of the most error-plagued areas of forensic science. Improper use of underlying fire science and often fire "junk" science can result in improper convictions. As Lithium-ion (Li-ion) batteries and devices become more commonplace, arson investigators are encountering these devices at fire scenes with increasing frequency. Currently, arson investigators lack science-based tools to evaluate whether the Li-ion battery started or was the victim of the fire. The findings from our project will increase the scientific knowledge on Li-ion battery fire dynamics and provide much needed experimental case studies to improve hypothesis generation and testing. The project will also develop chemical analysis tools and databases for Li-ion battery fire signatures. The project includes a suite of experimental tests designed to explore Li-ion battery fire characteristics and dynamics. The proposed test plan includes hundreds of fire tests of common Li-ion battery devices including laptops, smartphones and power tool packs. Li-ion fire characteristics, with emphasis on ignition and fire spread capacity to typical residential high heat release rate furnishings, will be quantified. Methods to evaluate physical signatures of fire-damaged Li-ion batteries and resulting fire debris will be explored. These methods include 3D X-Ray Computed Tomography (CT) and various microscopy-based surface material analysis methods. In the latter parts of the project, large-scale compartment fires will be conducted, and forensic methods developed in the first phase of the project will be applied. Results from testing will be collected and organized into a data repository that will be accessible to arson investigators. Because of the large and multidimensional data set that is expected to be generated by this project, statistical data analysis tools for classification, clustering, and regression will be applied to the data and scenarios. The project will seek to identify what types of data are most critical to clustering of scenarios into battery initiated fires. This project contains a research and/or development component, as defined in applicable law, and complies with Part 200 Uniform Requirements - 2 CFR 200.210(a)(14). CA/NCF