A novel methodology for modeling and analyzing thermal runaway propagation in lithium-ion battery modules using probability functions

To address the limitations of existing numerical simulation methods in accurately capturing the probabilistic changes in thermal runaway (TR) triggering temperatures in lithium-ion batteries (LIBs), this paper proposes a novel methodology for modeling and analyzing TR propagation using probability functions. By conducting a statistical analysis of the TR temperature ranges and frequency distributions from a large dataset of LIBs, the probability of TR triggering within each temperature range is computed. This probabilistic approach is integrated into the simulation process, with triggering probabilities determined based on random number-generated distributions. The proposed method is validated through experimental data, and a probabilistic trigger model (PTM) is developed to conduct numerical simulations of 18650-type LIBs from a probabilistic perspective. The TR propagation paths and probabilities in LIB modules, including configurations of four batteries in different arrangements, are examined. Notably, a new TR propagation mode, termed jumping propagation, is identified and simulated for the first time. Further analysis of an 8 x 8 module configuration shows that the TR propagation speed in the PTM closely matches that of the deterministic trigger model (DTM) when the fixed triggering temperature is set to 170 degrees C. Additionally, the use of probabilistic functions introduces fluctuations at the TR propagation front, resulting in an irregular square-shaped progression. The observed jumping propagation significantly reduces the TR time between adjacent cells. This methodology enhances the consistency between simulated and actual TR propagation, offering an effective tool for studying the probabilistic nature of TR processes in LIBs.