The storage of large quantities of Li-ion batteries is necessary in various industries and warehouses around the world. However, an increasing number of fire incidents is continuing to raise questions on the risk and safety of storage configurations of Li-ion batteries, because the fires have the potential to spread rapidly with minor chance for firefighting to quickly control the large-scale battery fires. In the consequence, these fires can release great amounts of toxic products that impact the environment and lead to major economic losses for the affected industry. Thus, a research project was conducted in Germany to investigate the controllability of large destructive fires in factory buildings with battery systems and its fire spread models (project title: BEGIN-HVS, German name: “Beherrschbarkeit von Großschadensfeuern in Industriehallen mit dem Gefahrgut Hochvoltspeicher und deren Ausbreitungsmodelle”). The project coordinator is the City of Munich – Fire Department with the TU Braunschweig conducting the experimental testing and BAM running the numerical investigations.
In the research project, the burning behaviour of Li-ion BEV batteries in the storage situation was investigated experimentally in three stages. In the first test series, BEV modules (3-9 kWh) were triggered into thermal runaway and the burning behaviour was investigated in a calorimeter test stand. In the second and third test series, up to eight full-size BEV battery systems (50-110 kWh) were placed in different storage configurations inside of a large-scale calorimeter. Again, the fire was triggered in these batteries and the burning behaviour of the individual full-size BEV-battery systems as well as the fire spread between multiple full-size BEV-battery systems was recorded.
A key characteristic in the investigation of fires is the heat released from the combustion. However, in large-scale investigations, heat release cannot be measured directly as the test setup is much to large to fit in an adiabatic calorimeter. Form other types of fire load characterisation, OCC and CDG methods are established to calculate heat release from exhaust gas concentrations (e. g. in Willstrand et. al., 2024, DOI: 10.1016/j.firesaf.2023.104078). All combustion gases must be collected and evaluated for a complete mass balance.
Oxygen consumption calorimetry (OCC)
Most widespread method for measurement of heat release with good accuracy for unknown fuel compositions is the OCC-method. The accuracy and uncertainty for battery fires is a current research topic. The general principle is a continuous measurement of the Reduction of the oxygen concentration in exhaust gas during the fire. This oxygen consumption is proportional to the heat release for the complete combustion. Other components like CO2, CO, water vapor can be considered as well.
Carbon-dioxide generation calorimetry (CDG)
Similar to OCC-method, the CDG-method estimates heat release from exhaust gas concentrations. However, here the increase in concentration of CO2 in the exhaust gas is correlated with the heat release. A wider range of E-factors for CO2 and CO (=heat release per kilogram of CO2 or CO production) must be mentioned as a limiting factor for the general application of this method when compared to the OCC-method.
The poster presents two experimental battery fire data sets analysed with both the OCC and CDG-method.
The research project “BEGIN-HVS” was funded from 2023-2025 by the German Federal Ministry of Research, Technology and Space (BMFTR) under the program “Forschung für die zivile Sicherheit” (www.sifo.de).