Anisotropic and homogeneous model of heat transfer for self-heating ignition of large ensembles of lithium-ion batteries during storage. (October 2021)

• Record Type: Journal Article
• Authors: Hu, Zhenwen; He, Xuanze; Restuccia, Francesco; Rein, Guillermo
• Is Part Of: Applied thermal engineering. Volume 197(2021)
• Publication Date: 2021-10
• Subjects: Self-heating -- Lithium-ion battery -- Thermal runaway -- Safety -- Heat transfer; Heat engineering -- Periodicals; Heating -- Equipment and supplies -- Periodicals; Periodicals; 621.40205
• DOI: 10.1016/j.applthermaleng.2021.117301
• Languages: English
• ISSNs: 1359-4311

Abstract

Highlights: A transient 3D anisotropic and homogeneous heat transfer model is developed to study the self-heating ignition behaviour of Lithium-ion battery ensembles. The computational efficiency of the resulting model is 10 times higher than a comparable heterogeneous model. The model predicts that the critical ambient temperature to trigger self-heating ignition drops significantly with size, which indicates a fire hazard for warehouse. The model can find safer storage conditions of storage size and packing design as a function of environmental temperature. Abstract: Self-heating ignition is a fire hazard in warehouses when stacking large quantities of reactive materials for storage, including lithium-ion batteries. Due to the heavy costs and dangerous fire risks, the thermal behaviour of large-scale LIB ensembles is usually studied by numerical methods. The state-of-the-art self-heating models on LIBs are either too computationally expensive to be applied to the predictions of large LIB ensembles, or capable of large ensemble predictions but missing important heat transfer characteristics like insulation in packaging. Based on four-step kinetics from the literature (Solid electrolyte interphase decomposition, negative-electrolyte reaction, positive-electrolyte reaction, and electrolyte decomposition), we have developed a 3D anisotropic homogeneous (Ani-Hom) transient heat transfer model that can incorporate complex packaging and is numerically affordable for large ensemble predictions based on COMSOL Multiphysics. The effect of packaging insulation is considered by using weight-averaged thermophysical properties and directional thermal conductivities. Lithium Cobalt batteries (LCO) are used as a case study. This Ani-Hom model was verified by comparing a box-scale simulation against an isotropic heterogeneous (Iso-Het) model from the literature. Both the predictions of temperature evolution and the heat generation agreed to within 5%, while the computational time of the Ani-Hom model is one order of magnitude lower than the Iso-Het model. The Ani-Hom model is then applied to LIB ensembles in four possible storage sizes, ranging from a single cell to a rack with around 10 million cells, with different packing configurations and spacing between cells. The model predicts that the presence of packaging insulation promotes self-heating ignition. A rack of this LCO LIBs is predicted to self-ignite at an ambient temperature of 45℃, which indicates that LIBs in a warehouse are vulnerable to fire hazards in warm environments. The presence of defects or abuse will even lower this critical ambient temperature. This work provides insights into the effects of complex insulation and spacing on self-heating ignition of LIBs during storage and contributes to a better understanding which can help mitigate such fires.