Are Sodium-Ion Batteries Safer Than Lithium-Ion Batteries? Insights from Thermal Runaway Experiments and Modeling

Thermal runaway (TR) represents a critical safety concern in battery technologies, often resulting in the rapid release of heat and toxic gases. In severe cases, TR propagates from cell to cell leading to fire or explosion of the entire battery pack. While TR phenomena have been extensively studied for lithium-ion batteries (LIBs), significantly less is known about the thermal safety of sodium-ion batteries (SIBs), particularly at the full-cell level. Therefore, further research is needed to assess the potential hazards of SIBs and to examine whether their often-claimed safety advantages over LIBs hold true. In this presentation, we evaluate the thermal safety behavior of a commercially available 18650-format SIB utilizing a NFM cathode and hard carbon anode and compare its safety performance with LFP and NMC based LIBs of identical format. A series of experimental techniques is employed to characterize TR onset, progression, and severity to provide robust data for the development and validation of a predictive TR model. First, the specific heat capacity and through-plane thermal conductivity of the jelly roll is determined based on the variant convective cooling method using a customized wind tunnel test bench. These thermophysical properties define the thermal transport behavior within the cell and serve as inputs to the heat transfer model. Accelerating rate calorimetry tests under quasi-adiabatic conditions are used to identify the onset temperatures of exothermic reactions and extract reaction kinetics parameters, i.e., activation energy and pre-exponential factor, for use in the Arrhenius-based TR model. The total heat release is assessed using fractional thermal runaway calorimetry, which enables the decoupling of heat retained in the cell body from heat expelled via the ejection of hot gases and particles. In addition to the heat release fractions, the measured mass loss is also used for model parameterization. Furthermore, in a separate overheating test setup, the venting dynamics of SIBs and LIBs are recorded using a high-speed camera with a resolution of 10,000 frames per second, allowing for the determination of gas velocity and precise venting duration. The experimental data are integrated into a three-dimensional multiphysics model, incorporating the described reaction kinetics and heat transfer parameters. Additionally, the phase transformation from solid to gas and mass transfer to the vent opening are considered. With this comprehensive modeling approach, different internal short-circuit scenarios are simulated to evaluate the associated TR risk for SIBs and LIBs. Altogether, this work contributes to the growing experimental dataset on SIB safety and to enhanced TR modeling, which is essential for risk mitigation and the development of safer next-generation SIBs.