With the acceleration of global energy transition, the field of power storage has become a core segment of the energy industry. Lithium-ion batteries, as one of the key components of electrochemical energy storage systems, have attracted widespread attention from both the market and researchers. In practical applications, the cycle life and safety characteristics of batteries after long-term use are critical technical indicators for evaluating energy storage cell performance. In this work, we systematically investigated the accelerated aging behavior of large-capacity lithium iron phosphate (LFP) energy storage cells under elevated temperature (45 °C) compared to room temperature (25 °C), and subsequently evaluated the safety characteristics of the aged cells via thermal runaway tests. Our findings provide a quantitative relationship between aging acceleration and thermal stability degradation, offering a theoretical basis for life prediction and safety monitoring of energy storage cells in real-world applications.
The aging mechanism of an energy storage cell is highly dependent on operating temperature. We designed a comparative cycling test using five different commercial energy storage cells, each with two samples, totaling ten cells. The cycling procedure followed the Chinese standard GB/T 36276–2023, section 6.6.2.1. For the 25 °C group, the cells were first kept at (25 ± 2) °C for 5 h, then charged at constant power Prc to the charge cut-off voltage, rested for 10 min, and then discharged at constant power Prd to the discharge cut-off voltage, also with a 10 min rest. This cycle was repeated 1000 times. The identical procedure was applied to the 45 °C group at (45 ± 2) °C.
For safety evaluation, we used 20 energy storage cells from 10 different manufacturers, and performed thermal runaway tests on both fresh cells and cells after 1000 cycles (aged cells). The test setup included a heating element and temperature sensors attached to the cell surface. The heating was initiated while the cell was charged at a constant current I = Prc/Unom. The thermal runaway trigger condition was defined as three consecutive temperature rise rates exceeding 3 °C/s, or the occurrence of fire or explosion. The temperature at the moment of thermal runaway was recorded as the thermal runaway temperature. The test was stopped once the condition was met, or when the temperature reached 300 °C, or after 4 h of testing.
2. Results and Discussion
2.1 Cycle Performance and Accelerated Aging Factor
The capacity retention curves for the five energy storage cells cycled at 25 °C are shown in the table below. All cells exhibited an initial capacity increase due to the activation effect, followed by a gradual decline. After 1000 cycles, the average capacity retention was 95.71%. In contrast, at 45 °C, the capacity decline was much faster, with an average retention of 90.81% after 1000 cycles, and no initial capacity rise was observed. The rapid degradation at elevated temperature is attributed to accelerated side reactions, SEI film thickening, and structural degradation of electrode materials.
| Cell ID | Capacity retention at 25 °C (%) | Capacity retention at 45 °C (%) |
|---|---|---|
| 1 | 95.82 | 90.18 |
| 2 | 95.81 | 90.26 |
| 3 | 96.24 | 90.24 |
| 4 | 95.63 | 91.88 |
| 5 | 95.06 | 91.47 |
| Average | 95.71 | 90.81 |
To quantify the acceleration effect of temperature on aging, we computed the ratio of capacity fade rate at 45 °C to that at 25 °C for each cycle interval. The capacity fade rate for a given cycle n is defined as
$$
r_{\text{fade}}(n) = \frac{C_0 – C_n}{n}
$$
where \(C_0\) is the initial capacity and \(C_n\) is the capacity after n cycles. The ratio \(R = r_{\text{fade,45°C}} / r_{\text{fade,25°C}}\) as a function of cycle number is shown in the following fitting formula. After the first 200 cycles, the ratio gradually approaches a constant value of approximately 2. The data from cycle 250 to cycle 1000 were fitted using an inverse function:
$$
R = a + \frac{b}{x}, \quad a = 0.4, \quad b = 22.7
$$
where \(x\) is the scaled cycle number (with cycle 250 mapped to 1, cycle 300 to 2, … , cycle 1000 to 16). The correlation coefficient \(R^2 = 0.9976\) indicates excellent agreement. This means that under the standard cycling conditions of GB/T 36276–2023, the aging acceleration factor for an energy storage cell at 45 °C relative to 25 °C stabilizes at around 2 after the initial period, providing a quantitative basis for accelerated life tests.
2.2 Thermal Runaway Behavior of Aged Cells
After 1000 cycles at 45 °C, the aged energy storage cells were subjected to thermal runaway tests alongside fresh cells. The thermal runaway temperatures are summarized in the table below. The fresh cells exhibited trigger temperatures ranging from 95 °C to 125 °C, while the aged cells showed higher values, between 115 °C and 135 °C. The increase is attributed to the growth of a thicker SEI layer during cycling, which decomposes exothermically and generates gases (CO₂, C₂H₄, O₂), accelerating the heat release and pressurizing the cell, thus shifting the thermal runaway onset to higher temperatures. The aged cells also display a narrower distribution of trigger temperatures, indicating a more homogeneous degradation state.
| Cell ID | Fresh cell thermal runaway temperature (°C) | Aged cell thermal runaway temperature (°C) |
|---|---|---|
| 1 | 98.1 | 115 |
| 2 | 100.1 | 117 |
| 3 | 98.3 | 116 |
| 4 | 100.4 | 123 |
| 5 | 103.7 | 125 |
| 6 | 103.9 | 125 |
| 7 | 121 | 128 |
| 8 | 125 | 131 |
| 9 | 121 | 131 |
| 10 | 124 | 133 |
We further investigated the relationship between the state of health (SOH) of the aged cells and their thermal runaway temperature. The SOH after 1000 cycles is defined as the ratio of the measured capacity to the rated capacity. For the ten aged cells, the SOH values ranged from 87.74% to 92.58%, with an average of 90.07%. The thermal runaway temperatures versus SOH are plotted, and a remarkable consistency is observed when we compute the ratio:
$$
\frac{T_{\text{TR}}}{\text{SOH}} \approx 130\text{–}145
$$
where \(T_{\text{TR}}\) is the thermal runaway temperature in °C and SOH is expressed as a percentage (e.g., 90% corresponds to 90). The ratio for each cell is listed below, confirming that this relationship holds for all tested energy storage cells from different manufacturers. This suggests that by measuring the SOH of an aged energy storage cell, one can estimate its thermal stability and thus predict the approximate thermal runaway temperature.
| Cell ID | SOH (%) | Thermal runaway temperature (°C) | Ratio |
|---|---|---|---|
| 1 | 87.74 | 115 | 1.31 |
| 2 | 88.96 | 117 | 1.32 |
| 3 | 88.61 | 116 | 1.31 |
| 4 | 90.18 | 123 | 1.36 |
| 5 | 90.24 | 125 | 1.39 |
| 6 | 90.25 | 125 | 1.39 |
| 7 | 90.26 | 128 | 1.42 |
| 8 | 91.26 | 131 | 1.44 |
| 9 | 90.58 | 131 | 1.45 |
| 10 | 92.58 | 133 | 1.44 |
We also examined the correlation between SOH and the time to trigger thermal runaway, but no clear trend emerged. The trigger time varied from 652 s to 1215 s, likely affected by cell-specific differences in internal resistance, electrode morphology, and SEI composition. Therefore, while SOH is a reliable indicator of the temperature threshold for thermal runaway, it does not directly predict the timing of the event.
3. Conclusion
In this study, we comprehensively investigated the accelerated aging behavior and post-aging safety characteristics of energy storage cells. The key conclusions are as follows:
- For large-capacity LFP energy storage cells cycled according to the standard GB/T 36276–2023, the capacity fade rate at 45 °C is approximately twice that at 25 °C after the initial 200 cycles, with the ratio stabilizing at about 2. This provides a quantitative acceleration factor for life prediction.
- After 1000 cycles at 45 °C, the thermal runaway temperature of aged energy storage cells increases compared to fresh cells, ranging from 115 °C to 135 °C. This is attributed to the buildup of a thicker SEI layer that releases more exothermic energy during decomposition.
- A strong linear-like relationship exists between the SOH of the aged energy storage cell and its thermal runaway temperature. The ratio of thermal runaway temperature (in °C) to SOH (in percentage) lies consistently between 130 and 145 across different manufacturers and aging states. This enables a simple safety assessment: by measuring SOH, one can estimate the thermal stability threshold and implement early warning strategies for energy storage cells in real-world applications.
These findings provide a scientific foundation for lifetime modeling, accelerated aging test design, and safety monitoring of energy storage cells in large-scale energy storage systems.