Our investigation into the fire propagation characteristics of double-layer energy storage batteries has revealed critical insights into the energy transfer mechanisms that drive thermal runaway escalation in vertically stacked configurations. The widespread deployment of energy storage systems using lithium iron phosphate batteries, while beneficial for grid stability, has introduced significant safety challenges due to thermal runaway incidents. Understanding how fire propagates between vertically arranged modules is essential for designing safer energy storage systems. In this study, we focused on 100 Ah lithium iron phosphate batteries, conducting three sets of double-layer battery module experiments with configurations of 1-1, 2-2, and 3-3 batteries per layer.
Our experimental methodology involved heating the bottom batteries until they vented, then actively igniting the released thermal runaway gas. We systematically recorded experimental phenomena and temperature variations, analyzing temperature rise rates and staging to quantitatively assess the cumulative energy transferred from bottom batteries to top batteries during fire propagation. Crucially, we decoupled heat transfer contributions from different pathways to understand the underlying mechanisms.
Experimental Setup and Methodology
The energy storage battery used in our experiments was a prismatic lithium iron phosphate cell with a nominal capacity of 100 Ah. The key specifications are summarized in the following table:
| Parameter | Value |
|---|---|
| Nominal Capacity | 100 Ah |
| Mass | (2245 ± 5) g |
| Dimensions (L × W × H) | 130 mm × 36 mm × 211.8 mm |
| Cathode Material | Lithium Iron Phosphate |
| Anode Material | Graphite |
| Nominal Voltage | 3.2 V |
| Upper Cutoff Voltage | 3.65 V |
| Lower Cutoff Voltage | 2.5 V |
| State of Charge (SOC) | 100% |
| Thermal Conductivity (Large Face) | 21.24 W/(m·℃) |
All experiments were conducted in a combustion chamber built according to standard GB/T 25207—2010. The battery modules were placed on a stainless steel platform, and an exhaust fan was operated throughout the experiment to ensure safety. We used K-type thermocouples and voltage leads connected to a HIOKI LR8450 data acquisition system, recording data at 10 Hz. High-definition cameras recorded the experimental phenomena. The heating process initiated with an 800 W heating plate attached to the first bottom battery. After the battery vented, a pulse igniter was used to ignite the combustible gas. Thermal runaway was identified when the backside temperature exceeded 5 ℃/s for three consecutive readings and the voltage dropped below 1 V.
Our three experimental groups were designed to represent different stages of fire propagation:
| Group | Configuration (Bottom × Top) | Fire Propagation Status |
|---|---|---|
| Experiment 1 (Non-Propagated) | 1 × 1 | No |
| Experiment 2 (Non-Propagated) | 2 × 2 | No |
| Experiment 3 (Fire-Propagated) | 3 × 3 | Yes |
Temperature monitoring points were strategically placed on each battery: front (Tif), side (Tis), back (Tib), and vent (Tiv). Additionally, thermocouples were placed on the bottom surface of the top battery (Tibo) to monitor flame temperature. Mica plates were attached to both ends of each battery group to minimize heat dissipation, and aluminum alloy fixtures clamped the modules with a pre-tightening force of 2 N·m. The vertical spacing between layers was set at 5 cm, representing a severe but realistic scenario.
Experimental Observations
Non-Propagated Groups
In Experiment 1 (1-1 configuration), the bottom battery vented at 623 seconds, with a front temperature of 287 ℃. After ignition, the flame burned stably in a baking state. Thermal runaway occurred at 930 seconds, with the front temperature reaching 648.2 ℃ at 999 seconds. The backside peak temperature was 442.3 ℃ at 1234 seconds, marking the end of thermal runaway after 304 seconds. The average temperature of the top battery peaked at 120.9 ℃ at 1405 seconds, but it never vented or experienced thermal runaway.
In Experiment 2 (2-2 configuration), the two bottom batteries sequentially experienced venting and thermal runaway between 788 and 1348 seconds. The top batteries reached an average peak temperature of 162.6 ℃ at 1609 seconds. Notably, one top battery vented at 1456 seconds, and the other at 1657 seconds, but neither achieved thermal runaway. The voltage of one top battery dropped to 2.97 V at 3447 seconds before recovering, indicating a localized internal short circuit that self-extinguished due to heat dissipation.
Fire-Propagated Group
Experiment 3 (3-3 configuration) revealed the full fire propagation process. The bottom batteries (1#, 2#, 3#) vented and experienced thermal runaway sequentially from 537 to 1493 seconds. Critically, at 1598 seconds, all three top batteries (4#, 5#, 6#) experienced simultaneous thermal runaway. The flame jet from the three top batteries combined, producing an intense roar and intermittent reignition due to the high jet velocity.
Temperature analysis showed that the peak temperature of the top batteries reached 640.2 ℃, which was 115.9 ℃ (22.1%) higher than the bottom batteries’ peak of 524.3 ℃. This indicates that during fire propagation, the top energy storage battery experienced significantly higher temperatures, making it more hazardous. The maximum temperature rise rate for the top batteries was 14.0 ℃/s, compared to 7.5 ℃/s for the bottom batteries, a difference of 6.5 ℃/s (86.7%).
Temperature Rise Characteristics
Analysis of the temperature rise stages revealed a three-step ladder pattern for the top batteries before thermal runaway. The temperature profile of a representative top battery showed distinct stages:
| Stage | Duration (s) | Temperature Rise (℃) | Average Rate (℃/min) |
|---|---|---|---|
| 1st Flame Baking | 221 | 24.9 | 6.8 |
| 1st Flame Jet | 110 | 25.5 | 13.9 |
| 2nd Flame Baking | 295 | 31.5 | 6.4 |
| 2nd Flame Jet | 123 | 31.1 | 15.2 |
| 3rd Flame Baking | 195 | 24.7 | 7.6 |
| 3rd Flame Jet | 117 | 34.4 | 17.6 |
The temperature rise during flame jet stages was approximately double that during flame baking stages. While the absolute temperature rise in each stage was similar (24.7-34.4 ℃), the rate of temperature increase in flame jet stages (13.9-17.6 ℃/min) was significantly higher than in flame baking stages (6.4-7.6 ℃/min). This demonstrates that the intense thermal runaway phase of bottom batteries provides more efficient heating to the top energy storage battery due to the higher velocity flame jet that imparts greater convective heat transfer to the bottom surface.
Energy Transfer Analysis
To quantify the cumulative energy transferred during fire propagation, we utilized the characteristic temperatures from adiabatic calorimetry data:
| Characteristic Temperature | Definition | Value (℃) |
|---|---|---|
| T1 | Self-heating onset | 116.6 |
| T2 | Thermal runaway onset | 179.7 |
| Tv | Vent temperature | 158.4 |
The energy required to trigger specific events was calculated using:
$$Q = cm\Delta T$$
where c = 985.3 J/(kg·℃), m = 2.24 kg, and ΔT represents the temperature increase from ambient. The energy required for venting was calculated as 294.3 kJ, and for thermal runaway as 341.8 kJ.
The cumulative energy transferred to the top energy storage battery in each experiment is summarized below:
| Experiment | Bottom Batteries | Cumulative Energy (kJ) | Percentage of TR Energy |
|---|---|---|---|
| Experiment 1 | 1 | 249.1 | 84.6% (of venting energy) |
| Experiment 2 | 2 | 334.3 | 97.6% (of TR energy) |
| Experiment 3 | 3 | 379.7 | 111.1% (of TR energy) |
The energy accumulated from one bottom battery (249.1 kJ) was below the venting threshold, consistent with the observation that no top battery vented. Two bottom batteries provided 334.3 kJ, which fell between the venting and thermal runaway thresholds, explaining the observed venting but no thermal runaway. Three bottom batteries delivered 379.7 kJ, exceeding the thermal runaway threshold by 37.9 kJ (11.1%), which is consistent with the successful fire propagation observed.
Heat Transfer Pathway Decoupling
The total energy triggering fire propagation in the top battery can be decomposed into four components:
$$Q_{trig} = Q_{bo} + Q_{side} + Q_{self} + Q_{diss}$$
where Qtrig is the total triggering energy, Qbo is heat transfer through the bottom surface, Qside is heat transfer through side surfaces, Qself is self-heating (negligible during the short heating period), and Qdiss is heat dissipation to the environment (negligible due to flame envelopment).
The average bottom surface temperature of the top batteries was calculated as:
$$T_{bo} = \frac{T_{4bo} + T_{5bo} + T_{6bo}}{3}$$
Heat transfer through the bottom surface was determined using Fourier’s law:
$$Q_{bo} = \frac{\lambda_1 A_{bo}}{\sigma_1} \int_{t_0}^{t_1} (T_{bo} – T_{ave}) dt$$
where λ1 = 21.24 W/(m·℃), Abo = 0.0047 m², and σ1 = 0.106 m.
The decoupling analysis revealed:
| Pathway | Energy (kJ) | Percentage |
|---|---|---|
| Bottom Surface | 180.5 | 47.5% |
| Side Surfaces | 199.2 | 52.5% |
| Self-heating & Dissipation | 0 | 0% |
| Total | 379.7 | 100% |
Both pathways contributed almost equally to the fire propagation triggering energy, with side surfaces accounting for a slightly larger share (52.5%) compared to the bottom surface (47.5%). This balanced distribution highlights the importance of protecting both the bottom and side surfaces of energy storage batteries in vertical configurations.
Energy Transfer Mechanism
The energy transfer mechanism during fire propagation in double-layer energy storage batteries can be summarized as follows:
After the first bottom battery vents and experiences thermal runaway, the top batteries undergo their first temperature rise stage, accumulating 249.1 kJ of energy. This is insufficient to cause venting or thermal runaway. Following the second bottom battery’s venting and thermal runaway, the top batteries experience the second temperature rise stage, accumulating 334.3 kJ. This energy is sufficient to cause venting but still 45.4 kJ short of triggering thermal runaway. After the third bottom battery completes its cycle, the top batteries reach the third temperature rise stage, accumulating 379.7 kJ, which exceeds the thermal runaway threshold. At this point, all three top batteries experience simultaneous thermal runaway.
The entire process from first venting to fire propagation took approximately 1061 seconds, providing a critical window for intervention. This “golden time” is sufficient for fire suppression systems to activate and potentially prevent catastrophic propagation.
Implications for Energy Storage System Design
Our findings provide several important guidelines for improving the safety of energy storage battery systems:
First, since fire propagation to the upper layer depends on horizontal thermal runaway propagation in the lower layer, installing suitable insulating materials between batteries in the same layer is crucial to prevent or delay horizontal propagation. This can significantly reduce the cumulative energy reaching the upper batteries.
Second, the existence of a substantial time window (over 1000 seconds) between first venting and fire propagation presents an opportunity for active thermal management and fire suppression. Early detection systems combined with targeted cooling or inerting agents could prevent the energy accumulation from reaching the critical threshold.
Third, the nearly equal contribution of bottom and side surface heat transfer (47.5% from bottom, 52.5% from sides) indicates that both pathways must be addressed in thermal protection design. Insulating materials should be applied to both the bottom surfaces and side surfaces of energy storage batteries to reduce heat transfer from flames below.
Finally, the significantly higher peak temperatures (22.1% higher) and faster temperature rise rates (86.7% higher) observed in the top batteries during fire propagation underscore the greater hazard potential of vertical propagation compared to horizontal propagation. This must be considered in risk assessment and mitigation strategies for energy storage power stations.
Conclusions
Our systematic investigation of fire propagation in double-layer energy storage battery modules has yielded several key conclusions:
1. The fire propagation threshold for vertically arranged energy storage batteries requires energy accumulation from multiple bottom batteries. One bottom battery provided 249.1 kJ (below venting), two provided 334.3 kJ (enabling venting but not thermal runaway), and three provided 379.7 kJ (triggering full fire propagation).
2. During fire propagation, top energy storage batteries reach peak temperatures 115.9 ℃ (22.1%) higher and maximum temperature rise rates 6.5 ℃/s (86.7%) higher than bottom batteries, indicating significantly greater hazard potential.
3. The temperature rise of top batteries during fire propagation follows a three-stage ladder pattern, with flame jet stages providing approximately double the heating rate of flame baking stages (13.9-17.6 ℃/min vs. 6.4-7.6 ℃/min).
4. Heat transfer to the top energy storage battery during fire propagation is nearly equally distributed between bottom surface (47.5%) and side surfaces (52.5%), emphasizing the need for comprehensive thermal protection.
5. The approximately 1061-second window between first venting and fire propagation provides a critical opportunity for intervention, which should be exploited in active safety systems.
These findings provide essential scientific guidance for the design of safer energy storage battery systems and the development of effective fire propagation suppression strategies. Future work should focus on validating these results in larger-scale module configurations and developing practical thermal management solutions based on the energy transfer mechanisms identified in this study.