Our research is driven by the critical need to enhance the thermal safety of power li ion battery systems, which are the cornerstone of new energy vehicles. Despite their advantages, the inherent risk of thermal runaway (TR), a complex cascade of exothermic reactions leading to potentially catastrophic failure, remains a significant safety concern. While considerable research has focused on the temperature rise characteristics of failing cells and modules, a comprehensive understanding of the total heat generation from large-capacity li ion battery cells and the hazardous thermal characteristics of their high-velocity, flammable gas efflux remains incomplete. This study aims to fill this gap by providing a detailed experimental investigation into the thermal failure process.
We designed an experiment to study a commercial 50 Ah prismatic li ion battery with a Li(Ni0.6Co0.2Mn0.2)O2 cathode. The cell was preconditioned to 90% state of charge (SOC). The thermal failure was triggered using a non-contact, radiant heater with a constant power of 500 W, simulating an external thermal abuse scenario. The experimental setup was instrumented to capture multi-faceted data: thermocouples were attached to the battery surface (side, negative terminal, front, back, and vent) and positioned vertically in the efflux plume at heights of 0 cm, +30 cm, -30 cm, +60 cm, and -60 cm relative to the vent. A load cell recorded mass loss in real-time. Gas analysis (O2, CO2, CO) in the exhaust duct, combined with measured flow rates, allowed for the calculation of heat release rate (HRR) and total heat release (THR) based on the oxygen consumption principle. Visual and infrared recordings documented the event progression.
The specifications of the li ion battery under investigation are summarized in the table below.
| Parameter | Value |
|---|---|
| Dimensions | 148 mm × 96 mm × 27.4 mm |
| Initial Mass | 909.84 g |
| Rated Capacity | 50 Ah |
| Nominal Voltage | 3.65 V |
| Voltage Range | 2.75 V – 4.25 V |
| Cathode Material | Li(Ni0.6Co0.2Mn0.2)O2 |
| Anode Material | Graphite |
Experimental Methodology and Analysis Principles
The core of our analysis hinges on quantifying heat generation and release. The heat generated by the li ion battery body itself, $Q_{body}$, and its power, $P_{body}$, were estimated based on its temperature rise and mass:
$$Q_{body} = c_p m \Delta T$$
$$P_{body} = \frac{dQ_{body}}{dt}$$
where $c_p$ is the specific heat capacity, $m$ is the instantaneous mass, and $\Delta T$ is the temperature increase.
The heat released by the combustion of ejected gases in the environment was calculated using oxygen consumption calorimetry. The critical formulas are:
$$ \dot{Q}_{efflux} = E \cdot \dot{m}_a \cdot \left( \frac{\chi_{O_2}^0 – \chi_{O_2}}{1 + \alpha (\chi_{O_2}^0 – 1)} \right) – E_{CO} \cdot \dot{m}_a \cdot \left( \frac{\chi_{CO}}{1 + \alpha (\chi_{O_2}^0 – 1)} \right) \cdot \left( \frac{M_{O_2}}{2 M_{CO}} \right)$$
where $\dot{Q}_{efflux}$ is the Heat Release Rate (HRR), $E$ and $E_{CO}$ are the heat release per kg of oxygen consumed for general combustion and CO combustion, respectively, $\dot{m}_a$ is the mass flow rate of exhaust gases, $\chi_{O_2}^0$ and $\chi_{O_2}$ are the initial and instantaneous oxygen mole fractions, $\chi_{CO}$ is the instantaneous carbon monoxide mole fraction, $\alpha$ is the expansion factor (taken as 1.105), and $M$ denotes molar mass.
The Total Heat Release (THR) is the integral of HRR over time:
$$THR = \int_{0}^{t} \dot{Q}_{efflux} \, d\tau$$
These principles allow us to dissect the thermal contributions from the li ion battery‘s internal reactions and its external fire plume separately.
Results and Discussion: The Thermal Runaway Cascade
Characteristic Temperature Evolution
The thermal failure of the li ion battery exhibited a distinct two-stage eruption process. The surface temperature and voltage evolution are plotted below, defining four key phases.
Phase I – Heating and Initial Reactions (0–1200 s): After an initial stabilization period, radiant heating commenced. The average surface temperature increased nearly linearly. Around 90–120 °C, the initial decomposition of the Solid Electrolyte Interphase (SEI) layer on the anode began, followed by reactions between the lithiated graphite and the electrolyte. These reactions produce gases like ethylene ($C_2H_4$):
$$(CH_2OCO_2Li)_2 \rightarrow Li_2CO_3 + C_2H_4 + CO_2 + \frac{1}{2}O_2$$
$$2Li_{(s)} + C_3H_4O_3(EC) \rightarrow Li_2CO_3 + C_2H_4$$
These processes generated mild internal pressure but no significant temperature excursion.
Phase II – First Venting and Eruption (1200–2050 s): A critical juncture was reached at approximately 1908 s when the cell voltage plummeted to zero, indicating severe internal short circuiting likely due to separator collapse. This triggered accelerated heat generation. At 1994 s, internal gas pressure exceeded the vent’s mechanical integrity, leading to the first violent ejection of hot gases and aerosols. This first eruption caused a sharp peak in the average surface temperature rise rate ($\alpha_{avg}$) to 8.5 K s-1, with a peak average surface temperature ($\theta_{avg}$) of 189.4 °C. The highest localized temperature was 280 °C at the negative terminal tab.
Phase III – Second and Major Eruption (2050–2250 s): Following the first venting, temperatures continued to climb, reaching the regime where the cathode active material decomposes exothermically, releasing oxygen:
$$Li(Ni_{0.6}Co_{0.2}Mn_{0.2})O_2 \rightarrow \frac{1}{2}Li_2O + \frac{3}{10}NiO + \frac{1}{10}CoO + \frac{1}{5}MnO_2 + \frac{1}{2}O_2$$
The released oxygen violently reacts with the remaining electrolyte and carbonaceous materials, leading to an extremely intense secondary eruption at 2154 s. This event was markedly more severe. The peak average surface temperature skyrocketed to 489.2 °C, with a maximum temperature of 650 °C recorded at the terminal. The peak average temperature rise rate reached 27.7 K s-1, over three times that of the first eruption. This dramatic intensification is governed by the Arrhenius law, where reaction rates increase exponentially with temperature:
$$k = A \exp\left(-\frac{E_a}{RT}\right)$$
The table below summarizes the key temperature metrics, highlighting the stark contrast between the two eruptions.
| Thermal Metric | First Eruption | Second Eruption | Relative Increase |
|---|---|---|---|
| Peak Avg. Surface Temp. ($\theta_{avg}^{peak}$) | 189.4 °C | 489.2 °C | +158% |
| Peak Avg. Temp. Rise Rate ($\alpha_{avg}^{peak}$) | 8.5 K s-1 | 27.7 K s-1 | +226% |
| Max Localized Temp. | 280 °C (Tab) | 650 °C (Tab) | +132% |
Phase IV – Post-TR Cooling (2250 s onward): After the second eruption, the exothermic reactions subsided, and the cell began to cool down.
Mass Loss Dynamics
The mass loss profile directly corresponds to the ejection of material during venting. Two distinct mass drop events were recorded, synchronized with the two eruptions. The first eruption resulted in a mass loss of 74 g, with a peak mass loss rate ($\beta$) of 13.3 g s-1. The second, more violent eruption led to a much larger mass loss of 199 g, with a peak rate of 32.7 g s-1. This indicates that the second event involved the ejection of a significantly larger quantity of active material, electrolytes, and reaction products at a higher velocity, corresponding to the more intense chemical reactions occurring at higher internal temperatures within the li ion battery.
Battery Body Heat Generation Analysis
Calculating the heat generated by the li ion battery body itself (excluding the combustion of ejected gases) reveals the intensity of internal reactions. The heat generation power ($P_{body}$) showed two clear peaks. The first, minor peak of 7 kW corresponded to the initial venting. The second, dominant peak reached 32 kW during the major eruption. Integrating the heat generation over time yielded a total body heat output of approximately 1.05 MJ. This substantial energy release within the cell casing is the primary driver for cell-to-cell propagation in a module.
Efflux Plume Temperature Field
The temperature distribution in the space above the failing li ion battery is critical for assessing fire spread risk. During the first eruption, the highest ambient temperature (209.8 °C) was recorded at the +30 cm height, indicating the flame plume reached this level. The second eruption was far more powerful. The peak ambient temperature of 705.3 °C was recorded at the +60 cm height, which was 44% higher than the maximum battery surface temperature itself. This demonstrates that the high-temperature, high-speed combustible jet poses a more severe immediate thermal threat to surrounding components than the surface of the failing li ion battery cell. The vertical temperature profile data is consolidated below.
| Thermocouple Position | Peak Temp. (First Eruption) | Peak Temp. (Second Eruption) |
|---|---|---|
| Vent (0 cm) | 166.5 °C | 467.0 °C |
| +30 cm | 209.8 °C | 228.8 °C |
| +60 cm | 73.3 °C | 705.3 °C |
| -30 cm | 161.9 °C | 120.7 °C |
| -60 cm | 38.2 °C | 45.0 °C |
Efflux Gas Heat Release Characteristics
The combustion of gases vented from the li ion battery represents a major fire hazard. The Heat Release Rate (HRR) curve, derived from oxygen consumption, showed two primary peaks aligning with the eruptions. The first HRR peak was 122.8 kW m-2, and the second was 89.2 kW m-2. The Total Heat Release (THR) accumulated in a step-like manner, increasing by approximately 1.18 MJ m-2 during the first eruption and 1.52 MJ m-2 during the second. The final cumulative THR from the efflux combustion reached 6.56 MJ m-2. This immense energy release into the environment underscores the potent fire risk associated with a single large-format li ion battery undergoing thermal runaway.
Conclusion
Our experimental investigation into the thermal failure of a large-format Li(Ni0.6Co0.2Mn0.2)O2 li ion battery triggered by radiant heating provides critical insights into its thermal hazards. The process is characterized by a two-stage eruption, where the second stage is drastically more severe due to high-temperature cathode decomposition and accelerated reaction kinetics. Key quantified findings include a maximum battery surface temperature of 489.2 °C, a peak internal heat generation power of 32 kW, and a total body heat release of 1.05 MJ. Most significantly, the combustion of ejected gases created an efflux plume with a peak ambient temperature (705.3 °C) exceeding the cell surface temperature, and released a total of 6.56 MJ m-2 of energy into the environment.
These results highlight a crucial safety consideration: the primary immediate threat during the failure of a li ion battery often comes not from the cell casing, but from the high-velocity jet of burning gases it projects. This understanding is fundamental for designing effective early warning systems, developing targeted thermal runaway mitigation strategies—such as flame-arresting venting channels or intra-module fire suppression—and for modeling fire propagation risks in battery packs. Ultimately, managing the efflux hazard is as important as managing the cell-internal thermal runaway process for ensuring the overall safety of li ion battery systems.