In the context of global efforts to mitigate climate change, the transition from traditional fossil fuels to renewable energy sources has become imperative. Renewable energies such as solar, wind, and geothermal power are increasingly deployed to reduce carbon emissions. However, the inherent intermittency and spatial variability of these sources necessitate efficient energy storage solutions to balance supply and demand. Among various storage technologies, li-ion batteries have emerged as a dominant choice due to their high energy density, long cycle life, and lack of memory effect. Nonetheless, the safety of li-ion batteries is critically dependent on temperature management. Under abusive conditions, exothermic side reactions can be triggered, leading to thermal runaway—a rapid and uncontrollable increase in temperature that poses significant fire and explosion hazards. This study investigates the efficacy of a mist cooling system in suppressing thermal runaway in li-ion batteries, with a focus on understanding the cooling mechanisms under varying operational parameters.
Thermal runaway in li-ion batteries is characterized by a chain of exothermic reactions that escalate once critical temperature thresholds are crossed. Typically, the solid electrolyte interface (SEI) layer begins to decompose around 90°C, followed by more severe reactions as temperature rises, eventually culminating in thermal runaway with peak temperatures exceeding several hundred degrees Celsius within seconds. The heat generation during this process can reach rates as high as 1.5 kW for individual cells, making conventional thermal management systems inadequate for suppression. Traditional approaches, such as enhanced battery thermal management systems (BTMS) incorporating phase change materials or insulation, often add complexity and reduce energy density. Alternatively, fire suppression agents like water mist or liquid nitrogen have shown promise but may not be optimized for integrated cooling within compact spaces. Mist cooling, which involves introducing fine droplets into an air stream, offers a potential solution by combining convective cooling with evaporative heat absorption, thereby enhancing cooling capacity without substantial system overhead. In this work, I explore the fundamental aspects of mist cooling for thermal runaway suppression, aiming to delineate critical conditions and optimize cooling strategies for li-ion batteries.
To conduct this investigation, I designed and constructed an experimental setup tailored for studying mist cooling effects on li-ion battery thermal runaway. The setup comprises a mist mixing section and a thermal runaway testing section, with a total length of 1.2 meters and a square cross-section of 0.1 m × 0.1 m. The mixing section, fabricated from acrylic glass to minimize droplet adhesion, includes side ports for droplet introduction and flow-straightening perforated plates to ensure uniform mist distribution. Droplets are generated via ultrasonic atomization and transported through tubing into the airstream. The testing section, made of stainless steel with mica lining for safety, houses a cylindrical electric heater with an inner diameter of 4 cm to externally heat the battery and trigger thermal runaway. A pneumatic mechanism is employed to move the battery into the mist flow at predetermined temperature points. Temperature measurements are taken using K-type thermocouples (accuracy ±1.5°C) attached to the battery surface and at the air inlet, while airflow velocity is monitored with a precision anemometer (±0.2 m/s).
The li-ion batteries used in this study are commercial 18650-type ternary cells with a nominal capacity of 2.6 Ah and voltage of 3.7 V. Prior to experiments, each li-ion battery undergoes conditioning through at least three charge-discharge cycles at 0.5 C rate to stabilize capacity and screen for outliers, ensuring reproducibility. The experimental procedure involves externally heating the li-ion battery until safety vent opening, after which the heater is deactivated. The battery then self-heats due to internal exothermic reactions, leading to a near-uniform temperature rise. Mist cooling is initiated at specified trigger temperatures by activating the atomizer and fan beforehand to establish steady flow, followed by pneumatic insertion of the battery into the mist stream. Critical suppression temperatures are determined using a bisection method, iteratively testing midpoints between safety vent opening and thermal runaway temperatures until the range narrows to within 3°C, accounting for cell-to-cell variations.
The typical thermal runaway curve for a li-ion battery without cooling exhibits a temperature drop post-vent opening due to heat dissipation, followed by a resurgence of heating driven by internal reactions. As shown in prior studies, the maximum heat generation rate can approximate 1.5 kW, posing a challenge for cooling systems. In mist cooling experiments, I observe that successful suppression occurs only if cooling is applied before thermal runaway triggers; once triggered, the heat generation overwhelms cooling capabilities. For instance, with a mist flow rate of 0.2 mL/s and airflow velocity of 1.5 m/s, cooling initiated at 219.4°C leads to continuous temperature decline to 154.5°C at an average rate of 1.1°C/s, whereas cooling at 261.2°C results in temporary cooling followed by runaway with peak temperatures around 449.8°C. This underscores the importance of early intervention in managing li-ion battery hazards.
To quantify cooling performance, I analyze the heat balance during mist cooling. The governing equation for battery thermal dynamics during cooling is:
$$ c m (T_{\text{start}} – T_{\text{finish}}) + H_g = \int h_W A (T – T_0) dt $$
where \( c \) is the specific heat capacity (0.85 kJ/(kg·°C) for the li-ion battery used), \( m \) is the mass (44.5 g), \( T_{\text{start}} \) and \( T_{\text{finish}} \) are initial and final battery temperatures, \( H_g \) is the heat generated during cooling, \( h_W \) is the mist cooling coefficient, \( A \) is the surface area, \( T \) is the instantaneous temperature, and \( T_0 \) is the ambient temperature. The mist cooling coefficient \( h_W \) incorporates both convective and evaporative effects. By comparing with pure air cooling (measured using a copper rod of identical dimensions to eliminate \( H_g \)), I derive the enhancement due to droplets. Table 1 summarizes the cooling parameters under different airflow velocities, with a constant mist flow rate of 0.2 mL/s.
| Cooling Parameter | Airflow Velocity (m/s) | 0.2 | 0.6 | 1.0 | 1.4 | 1.8 | 2.2 | 3.0 |
|---|---|---|---|---|---|---|---|---|
| Critical Suppression Temperature \( T_c \) (°C) | 241.2 | 245.1 | 246.4 | 247.9 | 247.9 | 247.3 | 247.3 | |
| Mist Cooling Coefficient \( h_W \) (W/(m²·°C)) | 50.9 | 53.8 | 56.6 | 61.1 | 66.5 | 73.7 | 82.5 | |
| Air Cooling Coefficient \( h_0 \) (W/(m²·°C)) | 31.6 | 37.4 | 43.5 | 48.2 | 56.2 | 62.9 | 72.0 | |
| Cooling Enhancement \( h_W – h_0 \) (W/(m²·°C)) | 19.3 | 16.4 | 13.1 | 12.9 | 10.3 | 10.8 | 10.5 | |
| Heat Generated During Cooling \( H_g \) (kJ) | 3.2 | 4.4 | 3.7 | 3.0 | 2.4 | 3.2 | 4.1 | |
| Total Temperature Rise \( T_{\text{total}} \) (°C) | 326.0 | 361.7 | 344.4 | 327.4 | 311.5 | 332.1 | 343.5 |
As evident from Table 1, the critical suppression temperature \( T_c \) for the li-ion battery increases with airflow velocity, plateauing around 247.9°C for velocities above 1.4 m/s. The cooling enhancement \( h_W – h_0 \), representing the contribution of droplet evaporation, decreases at higher velocities, stabilizing near 10.5 W/(m²·°C). This trend is attributed to the evaporation dynamics: at lower velocities, droplets have more time to evaporate in the high-temperature region near the li-ion battery surface, enhancing local cooling. At higher velocities, droplets may evaporate fully within the airstream before reaching the battery, diminishing additional evaporative cooling. The maximum evaporative capacity \( E_u \) can be estimated as:
$$ E_u = (1 – Rh) \times A_h $$
where \( Rh \) is relative humidity and \( A_h \) is absolute humidity. Under controlled conditions, \( E_u \approx 10.9 \, \text{g/m}^3 \), implying that for velocities exceeding 1.8 m/s, most droplets evaporate in the flow, reducing localized effects. The heat generated during cooling \( H_g \) varies with velocity, reflecting changes in internal reaction rates due to temperature gradients. The total temperature rise \( T_{\text{total}} \), calculated from \( H_g \) and cooling data, represents the hypothetical temperature increase if no cooling occurred, often exceeding 300°C even in suppressed cases. This indicates that while surface temperatures are controlled, internal reactions may proceed partially, emphasizing the need for timely cooling in li-ion battery safety protocols.
Further experiments examine the influence of mist flow rate on cooling behavior. Table 2 presents data for three mist flow rates at a constant airflow velocity of 1.5 m/s and cooling trigger temperature of 219.4°C.
| Cooling Parameter | Mist Flow Rate (mL/s) | 0.1 | 0.2 | 0.4 |
|---|---|---|---|---|
| Minimum Temperature During Cooling (°C) | 176.1 | 174.2 | 162.1 | |
| Post-Cooling Rebound Peak Temperature (°C) | 186.0 | 183.3 | 183.4 | |
| Total Cooling Heat Removal (kJ) | 2.0 | 2.1 | 2.1 |
Higher mist flow rates lead to more pronounced surface cooling, as seen by lower minimum temperatures. However, the total heat removal remains similar across flow rates, suggesting that increased cooling is offset by reduced efficiency due to steeper temperature gradients. This phenomenon arises because rapid surface cooling can insulate the li-ion battery core, limiting heat extraction. Thus, while higher mist flow rates improve cooling rates, they may not enhance overall energy removal, a critical consideration for designing mist cooling systems for li-ion battery packs.
The effect of initial cooling trigger temperature is also pivotal. Figure 1 (not shown) depicts temperature curves for different trigger temperatures, showing that lower trigger temperatures yield more extensive cooling. The cooling process typically exhibits three phases: an initial rapid drop due to conductive cooling of the metal casing, a plateau where internal heat generation balances cooling, and a final decline if cooling dominates. By analyzing \( H_g \) and \( T_{\text{total}} \) across trigger temperatures, I identify a safety threshold. For instance, when cooling starts below 225°C, \( T_{\text{total}} \) remains below the thermal runaway trigger temperature (296.3°C), virtually eliminating runaway risk. Between 225°C and 247.9°C, \( T_{\text{total}} \) exceeds the trigger temperature, indicating residual propagation hazard. Above 247.9°C, suppression fails. This leads to a tiered hazard prevention framework for li-ion batteries:
- Fundamental Suppression: Cooling initiated at temperatures where \( T_{\text{total}} < T_{\text{trigger}} \) (e.g., below 225°C), ensuring complete prevention of thermal runaway.
- Conditional Suppression: Cooling at intermediate temperatures (225–247.9°C) where \( T_{\text{trigger}} \leq T_{\text{total}} < T_{\text{runaway}} \), mitigating but not eliminating propagation risk.
- Critical Threshold: Cooling above 247.9°C, where suppression is ineffective, and thermal runaway proceeds unabated.
To delve deeper into the cooling mechanics, I model the heat transfer processes. The overall cooling capacity of mist flow combines convective heat transfer and droplet evaporation. The convective component can be expressed using Nusselt number correlations for flow over a cylinder, while evaporative cooling depends on droplet size, distribution, and local temperature. For a li-ion battery, the surface heat flux during mist cooling is:
$$ q” = h_W (T_s – T_\infty) + \dot{m}_d L_v $$
where \( q” \) is heat flux, \( T_s \) is surface temperature, \( T_\infty \) is bulk air temperature, \( \dot{m}_d \) is droplet evaporation rate per unit area, and \( L_v \) is latent heat of vaporization. The droplet evaporation rate is influenced by airflow velocity and mist concentration, aligning with experimental observations. At high temperatures, droplet evaporation near the li-ion battery surface dominates, as droplets instantly vaporize, absorbing significant heat. This local effect is maximized at moderate velocities where droplets are conveyed efficiently without premature evaporation. Additionally, the cooling enhancement factor \( \Delta h = h_W – h_0 \) correlates with mist concentration \( C_m \) and velocity \( U \) approximately as:
$$ \Delta h \propto \frac{C_m}{U} \quad \text{for} \quad U < U_{\text{crit}} $$
where \( U_{\text{crit}} \approx 1.8 \, \text{m/s} \) in this setup. Beyond \( U_{\text{crit}} \), \( \Delta h \) saturates as evaporation completes in the airstream. This relationship guides optimization of mist cooling parameters for li-ion battery applications.
In practical terms, integrating mist cooling into li-ion battery thermal management systems requires balancing cooling performance with energy and space constraints. The mist system can be activated upon detecting temperature anomalies via battery management systems (BMS), providing rapid response before thermal runaway escalates. For battery packs, mist distribution must be uniform to prevent hot spots, necessitating careful design of nozzles or atomizers. Moreover, the use of water-based mist raises concerns about electrical short circuits; however, in controlled environments or with insulated designs, this risk can be mitigated. Alternative fluids with higher latent heat or lower conductivity could be explored for enhanced safety in li-ion battery systems.
The economic and environmental implications of mist cooling are also noteworthy. Compared to traditional fire suppression systems that use large quantities of water or chemicals, mist cooling requires minimal fluid volume, reducing resource usage and potential damage. For stationary energy storage systems using li-ion batteries, mist cooling could be integrated into existing air-cooling infrastructure with modest modifications, offering a cost-effective upgrade for hazard prevention. Additionally, the ability to suppress thermal runaway early may extend battery pack lifespan by preventing catastrophic failures, aligning with sustainability goals for li-ion battery reuse and recycling.
Future research directions include scaling mist cooling for larger li-ion battery modules, studying multi-cell interactions during cooling, and exploring synergies with other cooling methods like phase change materials. Computational fluid dynamics (CFD) simulations could refine mist distribution patterns, while advanced sensors could enable real-time adaptive cooling. Furthermore, investigating mist cooling under diverse abuse conditions (e.g., electrical overcharge, mechanical damage) would broaden its applicability. The insights from this study contribute to the evolving paradigm of proactive safety in li-ion battery technologies, where cooling systems are not just for operational temperature control but also for emergency response.
In conclusion, this investigation demonstrates that mist cooling is a potent technique for suppressing thermal runaway in li-ion batteries, with effectiveness hinging on airflow velocity, mist concentration, and cooling initiation temperature. The critical suppression temperature for the tested li-ion battery is approximately 247.9°C, but safer operation is achieved below 225°C. Mist cooling enhances heat removal primarily through localized droplet evaporation, with diminishing returns at high velocities. A tiered hazard prevention framework based on total temperature rise offers practical guidance for system design. As li-ion batteries continue to power the renewable energy transition, integrating robust cooling strategies like mist systems will be essential for ensuring safety and reliability in diverse applications.