FKFS Veranstaltungen

2025 Sustainable Energy & Powertrains

25 - 26 November 2025 | Stuttgart

Session: 5-Minute-Pitches | | 13:00-14:00

Multiscale Analysis and Modeling of Thermal Runaway in Lithium-Ion Batteries

Deniz Ceylan, Porsche AG

Thermal runaway (TR) in lithium-ion batteries presents a critical safety concern, particularly in high-voltage battery (HVB) systems used in electric vehicles and stationary energy storage. Understanding and modeling the complex chain of events that lead to and propagate TR is essential for developing effective safety strategies and mitigation technologies. This study presents a comprehensive multi-scale analysis of the thermal runaway process, systematically describing its evolution across four key hierarchical levels: electrode, single cell, module, and the HVB system.

At the electrode scale, the initiation of TR is traced back to electrochemical and chemical degradation processes, including solid electrolyte interphase (SEI) decomposition, separator failure, and internal short circuits. These localized reactions trigger exothermic events, which subsequently drive gas generation, pressure accumulation, and venting at the cell level. As TR progresses to the module scale, heat transfer and fire propagation mechanisms—both horizontal and vertical—enable the cascading failure of neighboring cells. Ultimately, system-level effects at the HVB scale are influenced by cell arrangement, thermal management design, and enclosure properties.

To capture the interplay of these phenomena, we propose a semi-empirical modeling framework that integrates thermal, electrical, and chemical sub-models. The model incorporates key parameters such as conduction, ohmic heating, electrolyte decomposition, and gas-phase reactions, enabling the simulation of both intra-cellular reactions and inter-cell propagation dynamics. A conjugate heat transfer approach is employed to couple internal battery behavior with external heat and fluid flow processes, facilitating accurate predictions of fire behavior and energy release patterns.

This modeling approach provides a versatile platform for investigating TR propagation under various abuse conditions and system configurations, with potential applications in battery design optimization and risk assessment.