Since their transition from theory to practice in the late 20th century, lithium-ion batteries have been widely adopted across various sectors, including power grid storage, electric vehicles, and portable electronics, due to their superior performance. Today, with the rapid advancement of the new energy industry, lithium-ion cells, modules, and systems play an increasingly vital role in human production and daily life. However, due to their intrinsic safety limitations and the diverse, complex environments in which they are used, thermal runaway incidents can occur at any stage of a lithium-ion battery’s life cycle—during production, transportation, usage, storage, and recycling. Thermal runaway is often accompanied by fire, explosion, release of toxic gases, and high voltage, characterized by high destructiveness, complexity, and concealment. These incidents pose severe threats to life and property and hinder the development of the new energy industry and the achievement of national “dual carbon” goals. Consequently, in-depth research into the thermal runaway characteristics of lithium-ion batteries holds significant theoretical and practical value for scientifically understanding this behavior and guiding emergency response.
This study focuses on a commercially prevalent 18650 cylindrical lithium-ion cell with a nominal voltage of 3.7V and a rated capacity of 2.8Ah. The cell utilizes a LiNi0.6Co0.2Mn0.2O2 (NCM622) cathode, a graphite anode, and an electrolyte composed of LiPF6 in a 1:1 mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC). We employed external stress to accelerate the aging process, obtaining samples at different states of health (SOH) and states of charge (SOC). A comprehensive analysis was conducted using charge-discharge cycling, electrochemical impedance spectroscopy (EIS), hybrid pulse power characterization (HPPC), and adiabatic thermal runaway tests via an Accelerating Rate Calorimeter (ARC). The aim is to elucidate the fundamental electrochemical performance degradation and the adiabatic thermal runaway behavior of lithium-ion batteries under varying conditions.
Electrochemical Performance Degradation with Cycling
The capacity fade of the lithium-ion battery with cycling is a critical indicator of its state of health. The remaining discharge capacity relative to the initial capacity after different cycle counts is presented in the table below.
| Cycle Count | Relative Remaining Capacity (%) | Remarks |
|---|---|---|
| 0 (Fresh) | 100 | Baseline reference. |
| 20 | 95 | Initial rapid fade. |
| 200 | 90 | Fade rate slows down. |
| 1000 | 80 | Fade rate stabilizes but continues slowly. |
The data clearly shows that the capacity fade rate is initially high but gradually decreases and stabilizes over prolonged cycling, though a slow, continuous decline persists. This suggests that after a certain amount of internal degradation, the lithium-ion battery system reaches a more stable, albeit degraded, state.
Complementing the capacity data, the internal resistance, a key parameter influencing heat generation and thermal stability, was measured using the HPPC method at various depth-of-discharge (DOD) points. The direct current internal resistance (DCIR) at 50% SOC, where it is typically minimal, is summarized for different cycle counts.
| Cycle Count | DCIR at 50% SOC (mΩ) |
|---|---|
| 0 (Fresh) | 32.30 |
| 20 | 32.74 |
| 200 | 33.14 |
| 900 | 34.05 |
| 1000 | 34.16 |
The internal resistance of the lithium-ion battery increases with cycle count, rising from 32.30 mΩ for a fresh cell to 34.16 mΩ after 1000 cycles. The increase is most pronounced in the early cycles and then tends to plateau, mirroring the trend observed in capacity fade. A higher internal resistance leads to greater Joule heating during operation and can accelerate heat release during a thermal runaway event, thereby reducing the inherent thermal stability of the lithium-ion battery.
To gain deeper insight into the internal changes, electrochemical impedance spectroscopy was performed. The Nyquist plots were fitted to an equivalent circuit model to extract key impedance parameters. The results for cells at different cycle stages are compiled in the following table.
| Impedance Component | Fresh Cell (Ω) | After 100 Cycles (Ω) | After 200 Cycles (Ω) | After 300 Cycles (Ω) |
|---|---|---|---|---|
| Ohmic Resistance (Rs) | 17.3 | 23.6 | 26.7 | 28.3 |
| SEI Film Resistance (RSEI) | 1.1 | 2.7 | 2.9 | 4.6 |
| Charge Transfer Resistance (Rct) | 7.7 | 47.2 | 93.8 | 125.0 |
The EIS analysis reveals a consistent increase in all major impedance components with cycling. The ohmic resistance (Rs), related to electrolyte conductivity and electrode bulk resistance, increases by a factor of approximately 1.6. The SEI film resistance (RSEI) increases by a factor of about 4.2, indicating growth and restructuring of the solid-electrolyte interphase layer on the anode. Most dramatically, the charge transfer resistance (Rct), which governs the kinetics of the electrochemical reactions at the electrode interfaces, increases by a factor of over 16. This substantial rise in Rct is the primary contributor to the increased cell polarization and DCIR, signaling significant degradation of the electrode materials’ activity, possible loss of active material, and increased difficulty in the charge transfer process.
In summary, the aging of a lithium-ion battery manifests externally as a reduction in deliverable capacity. Internally, this correlates with increased parasitic side reactions, progressive electrode polarization, and reduced ionic conduction efficiency. These factors collectively lead to a rise in the internal resistance of the lithium-ion battery. The reaction kinetics slow down considerably, primarily due to the drastic increase in charge transfer resistance. This comprehensive electrochemical analysis confirms that the stability of a lithium-ion battery decreases with use, providing a fundamental basis for understanding its evolving safety characteristics.
Thermal Runaway Characteristics at Different States of Charge
The state of charge is a paramount factor influencing the safety and thermal runaway behavior of a lithium-ion battery. We investigated the adiabatic thermal runaway process for cells at 0% SOC (fully discharged, ~2.96V), 50% SOC (~3.76V), and 100% SOC (fully charged, ~4.18V) using an ARC. Key temperature points and events were identified during the tests:
- T0: Self-heating onset temperature (detected by ARC).
- T1: Jet/venting temperature (characterized by a sudden temperature drop point indicating separator collapse and internal pressure release).
- T2: Thermal runaway onset temperature (defined as the point where the self-heating rate exceeds 2 °C/min).
- T3: Maximum temperature during burning.
The recorded temperature data for the three SOC conditions are summarized in the table below.
| State of Charge (SOC) | T0 (°C) | T1 (°C) | T2 (°C) | T3 (°C) |
|---|---|---|---|---|
| 0% | 115 | 128.5 | 183.4 | 290.9 |
| 50% | 95 | 135.0 | 152.9 | 288.3 |
| 100% | 85 | 110.2 | 128.9 | 186.5 |
A clear trend is evident: as the SOC increases, all characteristic temperatures (T0, T1, T2, T3) decrease. The self-heating onset temperature T0 drops by 30°C from the 0% SOC to the 100% SOC condition. More critically, the thermal runaway onset temperature T2 decreases by approximately 54°C. This indicates that a fully charged lithium-ion battery is significantly more prone to initiating exothermic reactions and entering thermal runaway at a much lower temperature, aligning with the fact that it contains more active lithium and chemical energy. The lower T3 for the 100% SOC cell may be attributed to extensive material ejection during violent venting, leaving less combustible material inside the cell for subsequent sustained combustion. The 50% SOC cell exhibited the most violent jetting event, completely rupturing the safety vent and ejecting electrode material, likely due to a buildup of gaseous electrolysis products during a longer self-heating phase before venting.
The progression of the thermal runaway event can be analyzed by examining the time intervals between these key stages. We define the following time segments:
- t0: Time from test start to self-heating onset (T0).
- t1: Duration from self-heating onset to jetting (T0 → T1).
- t2: Duration from jetting to thermal runaway onset (T1 → T2).
- t3: Duration from thermal runaway onset to peak burning (T2 → T3).
The measured time intervals are presented in the following table.
| Time Segment | 0% SOC (min) | 50% SOC (min) | 100% SOC (min) |
|---|---|---|---|
| t0 | 1407 | 963 | 780 |
| t1 | 649 | 1073 | 817 |
| t2 | 433 | 101 | 82 |
| t3 | 31 | 39 | 35 |
The fully charged lithium-ion battery (100% SOC) reached the self-heating stage (t0) in the shortest time. Following venting (T1), the high-SOC cells transitioned to full thermal runaway (T2) very rapidly (t2 = 82-101 min), as the open cell structure allowed intense interaction of ejected materials with oxygen. In contrast, the 0% SOC cell, which remained mostly sealed, took much longer (433 min) to reach the same stage after its initial internal reactions. Once thermal runaway commenced (T2), the time to reach the peak burning temperature (t3) was relatively consistent across all SOC levels (31-39 min), indicating a similarly fierce reaction rate once the critical threshold is crossed.
The cell voltage provides another critical signature of internal failure. For all SOC levels, the voltage remained stable until the onset of significant self-heating. Subsequently, a sharp voltage drop to below 2V was observed, corresponding to the onset of internal short circuits caused by separator shrinkage/melting. This was followed by a rapid collapse to 0V, coinciding with the massive internal short circuit and the jetting event (T1).
The thermal runaway process in a lithium-ion battery is governed by complex chain reactions. The overall heat release rate can be considered a sum of exothermic reactions from different components (SEI decomposition, anode-electrolyte reaction, cathode decomposition, electrolyte decomposition, etc.). The reaction kinetics for each step often follow an Arrhenius-type relationship:
$$k_i = A_i \exp\left(-\frac{E_{a,i}}{RT}\right)$$
where \(k_i\) is the rate constant for reaction \(i\), \(A_i\) is the pre-exponential factor, \(E_{a,i}\) is the activation energy, \(R\) is the universal gas constant, and \(T\) is the absolute temperature. The self-heating rate (dT/dt) of the lithium-ion battery under adiabatic conditions is proportional to the sum of the heat generation rates from all ongoing reactions:
$$\frac{dT}{dt} = \frac{1}{m C_p} \sum_i Q_i m_i A_i \exp\left(-\frac{E_{a,i}}{RT}\right)$$
Here, \(m\) and \(C_p\) are the mass and specific heat capacity of the cell, \(Q_i\) is the reaction enthalpy, and \(m_i\) is the relevant mass of the reacting component. At high SOC, the anode is lithiated (LiC6), which is highly reactive. The exothermic reaction between the lithiated anode and the electrolyte has a lower activation energy and/or a higher heat of reaction \(Q_i\) compared to a discharged anode. This results in a significantly higher heat generation rate at a given temperature, explaining the lower T0 and T2 observed for the fully charged lithium-ion battery. Furthermore, the charged cathode (e.g., Li1-xNi0.6Co0.2Mn0.2O2) is thermally less stable and releases oxygen at a lower temperature, which can fuel combustion reactions with the electrolyte, creating a feedback loop that dramatically accelerates the thermal runaway.
In summary, the hazard severity of thermal runaway in a lithium-ion battery escalates with increasing SOC. Higher SOC leads to earlier and more violent onset of exothermic reactions. While the fully discharged lithium-ion battery (0% SOC) showed the most benign response in our test, it is crucial to note that under different abuse conditions (e.g., severe crushing), even a low-SOC cell can enter thermal runaway.
Conclusions and Practical Recommendations
This investigation into the thermal runaway performance of cylindrical NCM/graphite lithium-ion cells yielded the following key conclusions:
- Correlation between Capacity Fade and Internal Impedance: The aging of a lithium-ion battery, manifested as capacity fade, is intrinsically linked to the increase in its internal impedance. Cycling leads to increased electrode polarization, growth of the SEI layer, and a dramatic rise in charge transfer resistance. This degradation reduces ionic conduction efficiency and slows down reaction kinetics, making the lithium-ion battery less stable and more susceptible to failure.
- SOC-Dependent Thermal Runaway Hazard: The thermal runaway hazard of a lithium-ion battery is strongly dependent on its state of charge. Higher SOC results in lower characteristic temperatures (self-heating onset, thermal runaway onset) and a faster progression to catastrophic failure. A fully charged lithium-ion battery presents the highest thermal hazard under adiabatic conditions.
- Violence of Failure: The violence of cell failure, particularly the jetting event, can be severe at intermediate SOC (e.g., 50%) due to potential gas accumulation before venting.
Based on these findings, we propose the following practical recommendations for the handling, transportation, and use of lithium-ion batteries to mitigate risks:
- Transportation: Lithium-ion batteries, whether single cells or systems, should ideally be transported at a reduced state of charge, typically below 30% SOC. This significantly lowers their inherent chemical energy and reduces the probability and severity of a thermal runaway event if accidentally abused during transit.
- Usage Practices: Avoid frequent overcharging and high-current charging, as these practices accelerate the degradation mechanisms (electrode stress, SEI growth) that increase the internal resistance and reduce the stability of the lithium-ion battery.
- Operating Environment: Operate lithium-ion batteries within their specified temperature and pressure ranges. Extreme conditions can accelerate aging and trigger internal failures.
- End-of-Life Management: For applications like electric vehicles, a lithium-ion battery pack should be considered for replacement or repurposing when its usable capacity degrades below a certain threshold (e.g., 70-80% of initial capacity). For secondary use (e.g., in energy storage systems), continuous monitoring of the state of health is critical, especially in demanding environments. Batteries showing significant performance decay should be promptly retired and processed to prevent potential safety incidents.
This study provides fundamental insights into the thermal runaway behavior of lithium-ion batteries, linking electrochemical degradation to safety performance. The findings underscore the importance of managing both the state of health and the state of charge throughout the lifecycle of a lithium-ion battery to ensure safe and reliable operation.