Batteries are widely used in energy storage systems (ESS), and thermal runaway in different types of batteries presents varying safety risks. Therefore, comparative research on the thermal runaway behaviors of various batteries is essential. This study investigates the thermal runaway characteristics of sodium-ion batteries (NIBs), lithium iron phosphate batteries (LFP), and lithium-ion batteries with NCM523 and NCM622 cathodes. The experiments were conducted in a nitrogen-filled constant-volume sealed chamber. The results show that the critical surface temperatures at the time of thermal runaway are as follows: LFP (346 °C) > NIBs (292 °C) > NCM523 (290 °C) > NCM622 (281 °C), with LFP batteries exhibiting the highest thermal runaway critical temperature. NIBs have the lowest thermal runaway triggering energy (158 kJ), while LFP has the highest (592.8 kJ). During the thermal runaway of all four battery types, the primary gases produced include carbon dioxide, hydrogen, carbon monoxide, methane, ethylene, propylene, and ethane. For NCM622 and NCM523, carbon monoxide is the dominant combustible gas, with volume fractions of 35% and 29%, respectively. In contrast, hydrogen is the main flammable gas for LFP and NIBs, with volume fractions of 44% and 30%, respectively. Among these, NIBs have the lowest lower flammability limit (LFL), indicating the highest explosion risk. The thermal runaway characteristics of 50 Ah batteries provide valuable insights for battery selection and design in energy storage applications.

Lithium batteries are widely used in fields such as engineering micro-machines, robotics, and transportation. However, safety issues caused by battery thermal runaway limit their further promotion. This study used a sealed heating pressure chamber (SHPC) to perform “heat-wait-seek (HWS)” stepwise heating on a 50 Ah lithium iron phosphate (LiFePO₄) battery to trigger thermal runaway. It was found that the state of charge (SOC) has a significant impact on the safety of the battery. There was no significant correlation between the valve opening temperature (T₁) and the temperature at which the battery’s thermal runaway rapidly self-heats (T₂) and SOC. However, as SOC increased, the maximum temperature (T₃) of the battery’s thermal runaway increased, reaching up to 357.4 °C. The mass loss rate due to thermal runaway increased with SOC. The critical point of the battery’s safety valve was essentially independent of SOC and was mainly influenced by temperature. After thermal runaway, the mixed gas was passed through a gas chromatograph (GC) to detect its composition. When the SOC was below 50%, the total gas production from thermal runaway increased slowly (0.68–0.90 mol). Above 50% SOC, the total gas production from the battery increased sharply (at 75% SOC, 1.17504 mol; at 100% SOC, 2.33047 mol). Among these gases, the amount of H₂ increased sharply with SOC (from 0.01 mol at 0% SOC to 0.93 mol at 100% SOC), while the amount of CO₂ remained almost constant. Considering the inerting effect of CO₂ in the gas produced during thermal runaway of LiFePO₄ batteries, the lower flammability limit of the mixed gas increased as SOC decreased (from 6.91% at 100% SOC to 55.43% at 0% SOC). The risk of explosion during thermal runaway of high SOC batteries significantly increased. Notably, within the SOC range of 25% to 100%, the flammable range remained stable at 34–43%, but at 0% SOC, it sharply dropped to 0.5%. Therefore, batteries that are deeply discharged have higher safety.