Lithium-Ion Battery Thermal Behavior and Safety in Simulated Scenarios

Abstract

Lithium-ion batteries offer high specific energy and power but can demonstrate thermal instabilities that lead to safety issues with large modules. During off-nominal conditions such as over-charge or short circuit, individual cells may reach elevated temperatures where various exothermic side reactions such as solid-electrolyte interphase decomposition, anode/cathode reactions with the solvent, and the vaporization and potential combustion of the electrolyte can occur. These side reactions can subsequently trigger further reactions, and cause the cell to release tremendous amounts of thermal energy, sometimes accompanied by fire or cell rupture [1, 2]. The heat released during this thermal runaway event may propagate to adjacent cells, causing them to release additional heat until the entire module goes into thermal runaway. The potential for this module-level thermal runaway scenario must be avoided, especially in applications where safety is paramount such as in the aerospace and automotive industries [3, 4]. This work focuses on the experimental testing of various methods to characterize the thermal behavior of different module configurations and determine safe practices in lithium-ion battery module design. This testing consists of a modified oven test where a single cell in modules of various configurations is artificially stimulated to trigger thermal runaway via a heating element. The thermal behavior of the surrounding cells is analyzed to determine the performance of the propagation mitigation methods. Results have shown that increasing the inter-cell spacing beyond that of typical module layouts significantly decreases the probability of thermal propagation to adjacent cells of the cylindrical variety, as shown in Figure 1. The data gathered from these modified oven tests are utilized in coupled electrochemical-thermal simulations that model the various high-temperature side reactions to predict the thermal response for a given configuration [5-7]. It is expected that this model will permit potential pathways toward determining safe battery module configurations.