Thermal Runaway in Battery Cells: Guiding Principles for Continuum Scale Modeling
Zixin Ping,Kyle Pietrzyk,Ilenia Battiato
DOI: https://doi.org/10.1149/ma2024-01452524mtgabs
2024-08-09
ECS Meeting Abstracts
Abstract:Lithium-ion batteries are sensitive to temperature. Low-temperature or high-temperature operation may lead to lithium plating and dendrites, SEI decomposition, and separator collapse. Such phenomena may cause internal short circuits, rapid heat generation and, subsequently, may trigger thermal runaway, resulting in fire or explosion [1][2]. Hence, accurate modeling and prediction of the thermal response of battery cells inside the pack is critical to preventing safety hazards as well as optimizing battery performance. Battery thermal management system (BTMS) plays a significant role in maintaining an optimal operation temperature and ensuring thermal safety for electric vehicles [3]. However, the large number of battery cells inside one module makes it challenging for BTMS algorithms to realize real-time simulations based on fully resolved physics-based models. The conventional lumped capacitance (LC) thermal model achieves reduced computation costs under the hypothesis that temperature is spatially uniform. Nevertheless, such a model works better for small Biot number conditions, usually under low C-rate operation and moderate cooling scenarios, and it lacks justification under a broad range of battery operating conditions and materials [4]. To overcome such limitations, rigorous multiscale models, based on mathematical coarse-graining theories, have been developed to predict the thermal response of batteries at the module scale [5] while not compromising on computational cost. In this work, we investigate the accuracy of the classical LC thermal model for twenty potential battery modules constructed with commercial batteries and packing materials. We show that the classical LC thermal model may be inaccurate under a broad range of design conditions. In the context of multiscale modeling, we identify which effective models should be instead used, and analyze the impact of SOC and the relative strength of heat insulation and heat dissipation on model selection and accuracy. This work provides guidance, based on first principles and rigorous multiscale theory, on best practices for model selection in practical applications involving thermal behavior and thermal runaway in battery cells at the module scale. References: [1] Qin, Y., Xu, Z., Wu, Y., Lu, L., Han, X., Du, J., & Ouyang, M. (2022). Temperature distribution of lithium ion battery module with inconsistent cells under pulsed heating method. Applied thermal engineering , 212 , 118529. [2] Feng, X., Ren, D., He, X., & Ouyang, M. (2020). Mitigating thermal runaway of lithium-ion batteries. Joule , 4 (4), 743-770. [3] Noelle, D. J., Wang, M., Le, A. V., Shi, Y., & Qiao, Y. (2018). Internal resistance and polarization dynamics of lithium-ion batteries upon internal shorting. Applied energy , 212 , 796-808. [4] Mahamud, R., & Park, C. (2013). Spatial-resolution, lumped-capacitance thermal model for cylindrical Li-ion batteries under high Biot number conditions. Applied Mathematical Modelling , 37 (5), 2787-2801. [5] Pietrzyk, K., Bucci, G., Behandish, M., & Battiato, I. (2023). Automated upscaling via symbolic computing for thermal runaway analysis in Li-ion battery modules. Journal of Computational Science , 74 , 102134.