With the increasing degree of electrification in passenger transportation, the demand for lithium-ion batteries (LIBs) has significantly increased. However, they still face technical challenges, including improving energy density, shortening charging time, reducing production costs, and safety issues. Among them, thermal runaway is a key issue in battery safety research, especially for positive electrode material batteries with high nickel content. Integrating sensors into batteries is a promising method for monitoring their internal behavior during operation and failure.
In recent years, researchers have been working to integrate various sensors into batteries of different sizes and shapes to monitor internal temperature, such as fiber Bragg grating (FBG) sensors, Fabry Perot (FP) and FBG hybrid sensor networks, resistance temperature detectors (RTD), NTC thermistors, thin-film K-type thermocouples (TFTC), and thermocouples. However, most existing methods are limited to studying the internal temperature behavior of small-sized soft pack, cylindrical, or button cells in conventional cyclic testing. This article uses two integrated thermocouples to study the temperature behavior of a large-sized prismatic automotive grade lithium-ion battery (energy density of 246 Wh/kg, NMC811/C electrode chemistry) during thermal runaway.
1 Test
Battery preparation and sensor integration
We conducted experiments using commercial prismatic lithium-ion batteries and introduced the specifications of the battery, including capacity, voltage, energy, energy density, size, electrode material, electrode and separator thickness, etc. The process of integrating thermocouples into the interior of batteries is described in detail, including battery pretreatment, six steps in the glove box (cutting the top, cutting to prevent short circuits, extracting the core and cover, drilling, inserting thermocouples, folding and adding electrolyte, and sealing) to integrate thermocouples, and setting thermocouples on the surface and surrounding environment of the battery. The battery with integrated sensors is called CellINT.
Cycle pre-test
Perform low cycle charging and discharging tests on CellINT, with a charging current of 24A (0.25C) and a discharging current of 14A (0.15C), at 100% DOD. Record data from internal and external temperature sensors during the test to evaluate the temperature gradient during normal battery operation. The test will be conducted using a thermal runaway test chamber, battery tester, and data acquisition system.
Thermal runaway test
Perform thermal runaway testing chamber suitable for testing, equipped with temperature control and smoke extraction filtration system. Connect the battery to a battery tester and record temperature data. The initial SOC of the battery is about 0%. Charge it continuously with a constant current of 26A until thermal runaway occurs. At the same time, test CellINT and CellREF without integrated sensors. If overcharging fails to cause thermal runaway, use an external heating machine as a backup trigger.
2 Test result
Cycle performance (with/without integrated thermocouple)
By comparing the discharge capacity and quasi open circuit voltage (qOCV) curves of CellINT at low C rates before and after sensor integration, it was found that the qOCV deviation was the largest at SOC 0%, at 0.16%. The discharge capacity was 1.22% higher than the initial value at 22 ° C and 0.15C, which may be due to excessive compensation of electrolyte added during the integration process. However, sensor insertion did not significantly damage the electrochemical behavior of the battery, and the long-term impact on battery aging cannot be determined.
In further charge discharge cycles, the internal and external temperatures of the battery were analyzed, and the average temperature gradient between the center and outer shell surfaces of the battery at 0.15C was calculated to be 0.1 ° C/mm. This gradient is mainly attributed to poor thermal conductivity perpendicular to the electrode surface and high heat dissipation efficiency on the battery surface.
Thermal runaway performance
Comparing the surface temperatures of CellREF and CellINT during thermal runaway, CellREF showed a surface temperature rise rate of 1 ° C/s at 4.315 seconds after reaching the cut-off charging voltage, with T2 at measurement point 1 being 107.5 ° C. Comparing the voltage data of the two, there is a significant similarity during overcharging, and further determination is needed in the future to determine the proportion of deviation that may be caused by battery modifications.
3 Conclusion
This study inserted two thermocouples into a commercial prismatic high-energy lithium-ion battery (NMC811/C, 95 Ah) to characterize thermal runaway behavior. Overcharging induced thermal runaway was performed on the battery with thermocouples and a reference battery without thermocouples. Based on the voltage, charging current, surface temperature, and internal temperature data of both batteries, the following conclusions were drawn:
The universality of sensor integration methods: The method of integrating temperature sensors into small-sized batteries in existing literature can be adjusted appropriately and used for large-sized prismatic batteries without significant impact on battery performance, but the amount of evaporated solvent needs to be accurately determined to prevent overcompensation.
Internal temperature differences and sensor value: In large-sized prismatic high-energy batteries, even at low charging currents, there are differences in internal and external temperatures, which confirms the important value of integrating temperature sensors into batteries for understanding internal processes.
Advantages of early detection of thermal runaway: By utilizing integrated thermocouples, irreversible points can be detected 21 seconds in advance compared to surface temperature measurement in the event of thermal runaway caused by overcharging. This is of great significance for early detection of potential hazardous conditions in battery chemical systems with low thermal stability.