1. Introduction
Lithium-ion battery has become a crucial part of modern energy storage systems due to their high energy density, long cycle life, and environmental friendliness. However, the safety issue, especially the thermal runaway problem, has been a major concern. Thermal runaway can lead to serious consequences such as fire and explosion, which not only endanger human life and property but also limit the further development and application of lithium-ion battery. In this context, a detailed study on the thermal runaway behavior of different types of lithium-ion battery is of great significance. This article focuses on the comparison between lithium-iron battery (LiFePO4) and ternary lithium-ion battery.
1.1 The Importance of Studying Thermal Runaway Behavior
The thermal runaway process involves complex chemical reactions inside the battery, resulting in rapid heat release and the emission of a large amount of toxic and harmful gases. Understanding the characteristics and mechanisms of thermal runaway can help in the design and development of safer battery, as well as in formulating effective prevention and early warning strategies.
1.2 Objectives of the Study
The main objectives of this study are to investigate the thermal runaway characteristics of LiFePO4 and ternary lithium-ion battery under external heating conditions, including the thermal runaway phenomena, temperature changes, heat release, gas composition, and voltage variations. By comparing these two types of battery, we aim to provide a comprehensive understanding of their safety performance and offer valuable references for battery safety management.
2. Experimental Setup and Methods
2.1 Sample Preparation
- LiFePO4 Battery: A 3.2V 100Ah LiFePO4 lithium-ion battery (produced in Shandong, with dimensions of 200mm×170mm×30mm) was used. Before the thermal runaway experiment, the battery state was adjusted to 100% state of charge (SOC) using a charge-discharge tester (produced in the US). The adjustment method was as follows: charged at 1/3V, rested for 30min, then charged at 1/3C to 3.6V with a constant current, then charged at C to 3.6V with a constant voltage, and finally discharged at C to 2.0V, and then charged at 1/20C to 3.6V with a constant voltage and rested for 30min.
- Ternary Lithium-Ion Battery: A 3.6V 90Ah ternary LiNi0.5Co0.2Mn0.3O2 (NCM523) ion battery (produced in Changzhou, with dimensions of 150mm×100mm×50mm) was selected. The battery state was adjusted to 100% SOC in a similar way. The adjustment process included charging at 1/3C to 4.2V, discharging at C to 3.0V, charging at 1/3C to 4.2V with a constant voltage, resting for 30min, then discharging at 1/3C to 3.0V, resting for 30min, and finally charging at 1/3C to 4.2V with a constant voltage and resting for 30min.
2.2 Thermal Runaway Experiment
- The experiment was carried out in a room-temperature and open environment. A heating film was used to trigger the battery thermal runaway. During the experiment, the temperature changes at various parts of the battery were monitored and collected by 1mm diameter K-type thermocouples (produced in the US). The thermocouple measurement points were located at the geometric center of the large surface A (TC1), the geometric center of the side surface on the positive electrode side (TC2), the positive electrode tab (TC3), the safety valve port (TC4), the geometric center of the large surface B (TC5), and the geometric center of the side surface on the negative electrode side (TC6). After the thermocouples and heating films were arranged, asbestos was covered on both sides of the sample and clamped tightly with fixtures.
- The heating power was set at 500W and continued until the battery underwent thermal runaway. During the entire process, the temperature and voltage data were continuously recorded. After the thermal runaway occurred, the gas composition was analyzed using an MKS – MG6000 Fourier transform infrared spectroscopy online detector. The hydrogen gas was analyzed using an H240000 hydrogen electrochemical sensor (produced in Shanghai), and the smoke release rate and heat release rate were calculated using a Motis – 3MW heat release rate tester (produced in Jiangsu). In addition, video data was collected using a camera.
3. Results and Discussion
3.1 Thermal Runaway Phenomena
- LiFePO4 Battery: As shown in Figure 2, during the heating process, the LiFePO4 lithium-ion battery experienced a series of processes including temperature rise, safety valve opening, and intense smoke release. After 6 minutes of the safety valve opening, thermal runaway occurred (when the temperature rise rate of at least 3 measurement points was not less than 3°C/s or when a fire or explosion occurred). During the thermal runaway process, only a large amount of white smoke was released, and there was no combustion.
- Ternary Lithium-Ion Battery: As depicted in Figure 3, the ternary lithium-ion battery underwent processes such as temperature rise, safety valve opening, smoke release, and combustion. Immediately after the safety valve opened within 1 second, thermal runaway occurred, accompanied by smoke release and intense combustion, and a continuous jet flame was formed until the combustible materials were burned out.
3.2 Temperature Analysis
- Temperature Variation Curves: The temperature changes of the measurement points of the LiFePO4 and ternary lithium-ion battery with time are shown in Figure 4. For the LiFePO4 battery, in the first stage, the self-heating rate was low, and the temperature rise of the measurement points was mainly due to the heat transfer from the heating plate. After 37 minutes of heating, the safety valve opened, and a small amount of viscous substance flowed out along with the release of thin white gas near the safety valve port, corresponding to the leakage and deflation of the electrolyte at the valve port. The opening of the safety valve transferred some heat to the external environment, causing a slight decrease in the temperature of each measurement point. In the second stage, after a slight decrease, the temperature of each measurement point rose rapidly due to the superposition of the heat transfer from the heating plate and the large amount of self-generated heat inside the battery. After 43 minutes of heating, LiFePO4 battery underwent thermal runaway and released a large amount of white smoke without combustion. For the ternary lithium-ion battery, the temperature rise characteristics in the first stage were similar to those of the LiFePO4 battery. In the second stage, after 33 minutes of heating, the safety valve opened, and immediately there was intense combustion, accompanied by a large amount of heat release and flame generation. The temperature at the valve port of the ternary lithium-ion battery was the highest during the entire combustion process, which developed rapidly.
- Peak Temperatures and Arrival Times: The peak temperatures () and arrival times () of the measurement points during the heating process are listed in Table 1. The temperatures of the measurement points differed due to different heat conduction paths inside LiFePO4 battery and the low thermal conductivity of the safety valve surface material. For the LiFePO4 battery without combustion, the peak temperature at the safety valve port, which was farther from the heating plate, was the lowest. For the ternary lithium-ion battery with combustion, significant differences were observed among the measurement points. After thermal runaway combustion, the valve port temperature was the highest, and the average peak temperature of all measurement points was 352°C higher than that of the LiFePO4 battery, and the average arrival time was shortened by 845s. This indicates that the sampled LiFePO4 lithium-ion battery has higher safety, while the sampled ternary lithium-ion battery would cause greater harm if it undergoes thermal runaway.
- Comparison of Temperature Changes: As shown in Figure 5, during the thermal runaway process of the ternary lithium-ion battery, the temperature and temperature rise rate of each measurement point at the same time point were higher than those of the LiFePO4 lithium-ion battery. In an open environment with heat convection, the determination conditions for the self-heating start temperature, thermal runaway start temperature, and thermal runaway maximum temperature in an adiabatic experimental environment cannot be used to evaluate the characteristic behavior of thermal runaway. The difference between the time when the temperature changes suddenly and the time when the peak temperature is reached is listed in Table 2. The average temperature rise rate of the LiFePO4 lithium-ion battery during temperature mutation was 0.79°C/s, while that of the ternary lithium-ion battery was 10.52°C/s, which is 13.32 times that of the LiFePO4 lithium-ion battery. This shows that the energy release during the thermal runaway process of the sampled ternary lithium-ion battery is more intense.
3.3 Heat Analysis
- Heat Release Mechanisms: During the thermal runaway process, complex chemical reactions occur inside LiFePO4 battery, and most of them are exothermic reactions. As the temperature rises and time elapses, reactions such as the decomposition of the solid electrolyte interphase (SEI) film, the decomposition of the electrolyte, the reaction between the electrolyte and the negative electrode, the oxygen release reaction of the positive electrode, and the oxidation reaction of the electrolyte and oxygen, and the reaction between the negative electrode and the electrolyte at high temperatures occur in sequence. The oxygen release reaction of the positive electrode and the oxidation reaction of the electrolyte are the main sources of heat generation.
- Heat Release Parameters: The smoke release rate (SRR), total smoke release amount (TSR), heat release rate (HRR), and total heat release amount (THR) are important parameters to describe the heat release characteristics of LiFePO4 battery. The formulas for calculating these parameters are given. The SRR is related to the density of the smoke, and the larger the SRR, the lower the visibility and the lower the evacuation efficiency of personnel. The TSR is the integral of the SRR with respect to time. The HRR is calculated using the oxygen consumption method, and the larger the HRR, the faster the material pyrolysis speed and the more volatile combustible substances are generated, accelerating the spread of the flame. The THR is the integral of the HRR with respect to time. The typical heat release parameters of the LiFePO4 and ternary lithium-ion batteries with time are shown in Figure 6, and the characteristic heat values are listed in Table 3. The total smoke release amount of the LiFePO4 battery was 395.0m², and the highest smoke release rate was 7.5377m²/s. The total smoke release amount of the ternary lithium-ion battery was 38.8m², and the highest smoke release rate was 2.2802m²/s. The total heat release amount of the LiFePO4 battery was 0.162MJ, and the highest heat release rate was 1.81kW. The total heat release amount of the ternary lithium-ion battery was 3.147MJ, and the highest heat release rate was 134.85kW. The size and speed of heat release are mainly related to the chemical composition of the materials undergoing pyrolysis or combustion and the combustion conditions. The heat release amount of the sampled ternary lithium-ion battery during thermal runaway is greater than that of the LiFePO4 battery, indicating that the thermal runaway of the ternary lithium-ion battery is more harmful.
3.4 Gas Analysis
- Gas Release Phenomena: During the thermal runaway experiment, the LiFePO4 lithium-ion battery emitted a large amount of white smoke without combustion, while the ternary lithium-ion battery emitted white smoke and underwent intense combustion.
- Gas Composition: The gas compositions released by the LiFePO4 and ternary lithium-ion batteries during thermal runaway are shown in Figure 7. The main components of the gases released by both batteries are H₂, CO₂, CO, and hydrocarbons. The content of CO₂ in the ternary lithium-ion battery is higher than that in the LiFePO4 battery. The combustion of combustible gases during the thermal runaway process is the main reason for the difference. H₂ is mostly generated from the reaction of the binder polyvinylidene fluoride (PVDF) and carboxymethyl cellulose (CMC) with metallic lithium. CO₂ is mainly derived from the decomposition reaction of carbonates in various electrolytes including the SEI film and the combustion of various combustible gases. The hydrocarbon compounds are mainly generated from the complex redox reactions of carbonates in the electrolyte. The reactions equations for the generation of these gases are given. Since the air almost does not contain H₂ and CO, H₂ and CO gases can be used as criteria for the early detection of thermal runaway. The experimental results show that the three gases H₂, CO₂, and CO account for more than 75% of the volume fraction of the released gas, and the influence of the positive electrode system on the gas composition of thermal runaway is not significant. Gas detection can be used as one of the main signals for the early warning and judgment of the thermal runaway of lithium-ion batteries.
3.5 Voltage Analysis
- Voltage Variation Phenomena: The changes in temperature and voltage during the thermal runaway process of the LiFePO4 and ternary lithium-ion batteries are shown in Figure 8. During the entire heating process, the voltage of the LiFePO4 lithium-ion battery decreased twice. The first decrease was from 3.388V (at 2237s) to 1.998V (at 2245s), and then it fluctuated and rose back to 2.888V (at 2645s). The second decrease was from 2.888V (at 2576s) to 0V. For the ternary lithium-ion battery, the voltage also decreased twice. The first decrease was from 4.154V (at 1949s) to 3.629V (at 1961s), and then it remained constant for 10s. The second decrease was from 3.629V (at 1960s) to 0V.
- Voltage Variation Mechanisms: The first voltage change was mainly due to the dissolution of the positive and negative electrode materials as the temperature increased. The second voltage change was due to the contraction and melting of the diaphragm, resulting in a large-area short circuit inside LiFePO4 battery.
- Voltage and Thermal Runaway Time Relationship: The times of safety valve opening, thermal runaway, and voltage change for different batteries are listed in Table 4. For the LiFePO4 battery, the voltage started to change 373s (more than 5min) earlier than the temperature corresponding to the thermal runaway. For the ternary lithium-ion battery, the voltage started to change 46s earlier than the temperature corresponding to the thermal runaway. The time differences between the temperature and voltage inflection points provide ideas for selecting early warning signals. Adding voltage signal monitoring and using both temperature and voltage signals as the basis for early warning devices can improve the accuracy and timeliness of early warning.