In the present investigation, we have focused on the formulation and comprehensive assessment of an immersion-type oil-based coolant specifically designed for lithium-ion energy storage cells employed in large-scale electrochemical energy storage systems. Our work is motivated by the critical need for advanced thermal management technologies that can effectively mitigate the thermal runaway risks inherent in densely packed energy storage cell modules. The growing capacity of individual energy storage cells and the escalating heat generation during high-rate charging/discharging cycles have rendered conventional air-cooling methods inadequate. Immersion liquid cooling emerges as a promising solution because it enables direct contact between the dielectric coolant and the energy storage cells, thereby achieving superior heat transfer efficiency and intrinsic safety. Herein, we report the development of a novel hydrocarbon-based immersion coolant derived from high-quality crude oil sourced from the Xinjiang Karamay oilfield, using molecular refining techniques to obtain ideal hydrocarbon fractions with tailored carbon number and boiling range distributions. The additive formulation, comprising an optimized antioxidant package, was determined to impart excellent oxidation stability, high dielectric strength, and outstanding compatibility with all internal components of the energy storage cell system. We validated the coolant’s performance through a series of rigorous tests, including accelerated aging experiments, material compatibility assessments, and thermal runaway simulations. The results demonstrate that our immersion coolant effectively prevents the propagation of thermal runaway events in energy storage cells, thereby fundamentally addressing the worldwide safety challenge of fire and explosion associated with lithium-ion battery failures.
Table 1 summarizes the key technical specifications that we established for the immersion coolant product. These criteria were derived from the operational demands of energy storage cell systems, including stringent requirements for electrical insulation, thermal conductivity, and long-term stability.
| Property | Test Method | Specification Limit |
|---|---|---|
| Kinematic viscosity @ 40 °C (mm²·s⁻¹) | GB/T 265 | 9 – 11 |
| Kinematic viscosity @ 100 °C (mm²·s⁻¹) | GB/T 265 | Report |
| Viscosity index | GB/T 1995 | Report |
| Flash point (open cup) (°C) | GB/T 3536 | ≥ 160 |
| Pour point (°C) | GB/T 3535 | ≤ -50 |
| Acid value (mg KOH·g⁻¹) | NB/SH/T 0836 | ≤ 0.02 |
| Water content (μg·g⁻¹) | SH/T 0207 | ≤ 30 |
| Breakdown voltage (kV) | GB/T 507 | ≥ 50 |
| Rotating pressure vessel oxidation test (140 °C, min) | SH/T 0193 | ≥ 400 |
| Evaporation loss (50 °C, 168 h, wt%) | In-house method | ≤ 0.5 |
| Acute skin irritation/corrosion | GB/T 21604 | Non-irritant |
| Median lethal dose (LD50, μg·g⁻¹) | OECD 423 | > 2,000 |
The base oil selection process involved evaluating several potential hydrocarbon fractions. Table 2 presents the key physical properties of the candidate base stocks. Fraction 1 exhibited a favorable balance of low pour point, high flash point, and low acid value, making it the most suitable choice for the final formulation.
| Property | Fraction 1 | Fraction 2 | Fraction 3 | Fraction 4 |
|---|---|---|---|---|
| Kinematic viscosity @ 40 °C (mm²·s⁻¹) | 10.75 | 9.25 | 7.18 | 9.08 |
| Flash point (open cup, °C) | 172 | 168 | 150 | 158 |
| Density @ 20 °C (kg·m⁻³) | 830.8 | 850.5 | 881.9 | 864.8 |
| Pour point (°C) | -64 | -63 | -56 | -62 |
| Acid value (mg KOH·g⁻¹) | 0.01 | 0.01 | 0.01 | 0.01 |
To optimize the oxidation stability of the coolant, we conducted a series of Rotating Pressure Vessel Oxidation Test (RPVOT) experiments with varying concentrations of the antioxidant additive T501 (2,6-di-tert-butyl-p-cresol). The results are compiled in Table 3. The data clearly indicate a monotonic increase in oxidation induction time with antioxidant concentration up to 0.3 wt%, beyond which further addition yields only marginal gains. This behavior can be attributed to an equilibrium between free-radical scavenging efficacy and secondary reactions at high additive loadings. Based on a balance between performance and cost, we selected an antioxidant concentration of 0.4 wt% to ensure a sufficient safety margin.
| Sample No. | Antioxidant Content (wt%) | RPVOT Time (min) |
|---|---|---|
| 0# | 0.0 | 61 |
| 1# | 0.1 | 304 |
| 2# | 0.2 | 441 |
| 3# | 0.3 | 504 |
| 4# | 0.4 | 508 |
| 5# | 0.5 | 515 |
The final formulated coolant, designated EBC160, was subjected to comprehensive quality analysis. Table 4 lists the measured properties, all of which comply with the defined specifications. In particular, the breakdown voltage of 62 kV ensures excellent electrical insulation for energy storage cells immersed in the coolant, while the low evaporation loss and non-irritant dermal classification enhance operational safety.
| Property | Specification | EBC160 Result | Test Method |
|---|---|---|---|
| Kinematic viscosity @ 40 °C (mm²·s⁻¹) | 9 – 11 | 9.36 | GB/T 265 |
| Kinematic viscosity @ 100 °C (mm²·s⁻¹) | Report | 2.31 | GB/T 265 |
| Flash point (open cup, °C) | ≥ 160 | 166 | GB/T 3536 |
| Pour point (°C) | ≤ -50 | -69 | GB/T 3535 |
| Acid value (mg KOH·g⁻¹) | ≤ 0.02 | 0.01 | NB/SH/T 0836 |
| Water content (μg·g⁻¹) | ≤ 30 | 21 | SH/T 0207 |
| Breakdown voltage (kV) | ≥ 50 | 62 | GB/T 507 |
| RPVOT (140 °C, min) | ≥ 400 | 540 | SH/T 0193 |
| Evaporation loss (50 °C, 168 h, wt%) | ≤ 0.5 | 0.08 | In-house |
| Acute skin irritation/corrosion | Non-irritant | Non-irritant | GB/T 21604 |
| LD50 (μg·g⁻¹) | > 2,000 | > 2,000 | OECD 423 |
Material compatibility is a critical factor for the long-term reliability of immersion cooling systems, especially when the coolant is in direct contact with various polymeric and metallic components inside the energy storage cell module. We selected a representative set of contact materials (see Table 5) and conducted accelerated aging tests at 85 °C for 336 hours. The coolant was exposed to each material separately, and both the mass change of the materials and the changes in coolant properties were monitored.
| Material ID | Description |
|---|---|
| 1 | Fishbone bracket (plastic) |
| 2 | Sampling wire (insulated) |
| 3 | Green connector (plastic) |
| 4 | White connector (plastic) |
| 5 | PET film |
| 6 | Temperature sampling wire |
| 7 | Blue adhesive tape |
| 8 | Blue insulating film |
| 9 | Separator membrane |
| 10 | Explosion-proof valve |
| 11 | Top adhesive patch |
| 12 | Top cover plastic part |
| 13 | Positive terminal (metal) |
| 14 | Positive electrode tab |
| 15 | Negative electrode tab |
Table 6 summarizes the mass change of each material after the aging test. The majority of materials exhibited negligible mass variation (less than ±0.5%), with the exception of the temperature sampling wire, which showed a mass loss of 4.2%. This is likely due to the absorption of some plasticizer or minor swelling of the insulation material, but it does not compromise the structural integrity of the component. Table 7 presents the critical properties of the coolant before and after contact with each material. No significant degradation in water content, acid value, dielectric loss factor, permittivity, or breakdown voltage was observed, confirming the excellent compatibility of the developed coolant with the entire energy storage cell system.
| Material ID | Initial Mass (g) | Final Mass (g) | Mass Change (%) |
|---|---|---|---|
| 1 | 20.5265 | 20.5347 | +0.04 |
| 2 | 4.3545 | 4.3615 | +0.16 |
| 3 | 1.3286 | 1.3288 | +0.00 |
| 4 | 1.1258 | 1.1262 | +0.04 |
| 5 | 2.0694 | 2.0587 | -0.52 |
| 6 | 1.8850 | 1.8058 | -4.20 |
| 7 | 0.3622 | 0.3618 | -0.02 |
| 8 | 0.5966 | 0.5988 | +0.11 |
| 9 | 0.2240 | 0.2245 | +0.22 |
| 10 | 0.0256 | 0.0257 | +0.39 |
| 11 | 1.5787 | 1.5802 | +0.09 |
| 12 | 4.2115 | 4.2213 | +0.23 |
| 13 | 1.6842 | 1.6840 | -0.01 |
| 14 | 3.4385 | 3.4387 | +0.06 |
| 15 | 7.3506 | 7.3500 | -0.01 |
| Material ID | Water Content (μg·g⁻¹) | Acid Value (mg KOH·g⁻¹) | Dielectric Loss Factor (25 °C) | Dielectric Constant (25 °C) | Breakdown Voltage (kV) |
|---|---|---|---|---|---|
| Blank | 22.7 | 0.006 | 0.00006 | 2.13 | 64.8 |
| 1 | 23.9 | 0.005 | 0.00005 | 2.11 | 63.5 |
| 2 | 22.2 | 0.006 | 0.00006 | 2.12 | 61.8 |
| 3 | 21.8 | 0.006 | 0.00006 | 2.11 | 63.5 |
| 4 | 22.4 | 0.006 | 0.00003 | 2.12 | 65.2 |
| 5 | 23.1 | 0.007 | 0.00006 | 2.13 | 67.3 |
| 6 | 20.3 | 0.006 | 0.00005 | 2.12 | 64.6 |
| 7 | 22.2 | 0.006 | 0.00008 | 2.12 | 61.7 |
| 8 | 21.8 | 0.006 | 0.00008 | 2.12 | 60.3 |
| 9 | 22.9 | 0.006 | 0.00009 | 2.13 | 63.1 |
| 10 | 23.9 | 0.006 | 0.00007 | 2.12 | 68.1 |
| 11 | 23.5 | 0.006 | 0.00006 | 2.11 | 64.3 |
| 12 | 23.1 | 0.005 | 0.00005 | 2.11 | 65.3 |
| 13 | 24.3 | 0.006 | 0.00006 | 2.11 | 64.9 |
| 14 | 24.0 | 0.006 | 0.00006 | 2.13 | 66.8 |
| 15 | 24.8 | 0.006 | 0.00005 | 2.12 | 64.7 |
The most critical evaluation involved the safety performance of the coolant under thermal runaway conditions. We designed a comparative experiment where a lithium-ion energy storage cell was subjected to a needle puncture test in two different environments: one in open air and one fully immersed in the EBC160 coolant. The puncture simulates an internal short-circuit, which is a common root cause of thermal runaway in energy storage cells. In the air environment, the puncture immediately triggered a violent reaction: flames erupted, and the cell surface temperature exceeded 400 °C within seconds. In stark contrast, when the same energy storage cell was immersed in our coolant, the heat generated by the short circuit was rapidly absorbed by the surrounding liquid. The coolant also acted as an effective flame suppressant by isolating the reactive materials from oxygen. The measured surface temperature of the energy storage cell peaked at 280 °C, while the coolant bulk temperature only rose to 48 °C. Although a large amount of white smoke was generated (attributed to the vaporization of low-boiling organic solvents in the electrolyte), no open flame or explosion occurred. This result unequivocally demonstrates that immersion cooling can halt the propagation of thermal runaway in energy storage cells, thereby preventing catastrophic failure.
The heat transfer performance of the immersion coolant can be quantified by the energy balance equation for a single energy storage cell undergoing thermal runaway. The heat generation rate \( \dot{Q}_{\text{gen}} \) during a short-circuit event is balanced by the heat absorbed by the coolant plus the heat stored in the cell itself. A simplified model is given by:
$$
\dot{Q}_{\text{gen}} = hA(T_{\text{cell}} – T_{\text{coolant}}) + m_{\text{cell}} c_{p,\text{cell}} \frac{dT_{\text{cell}}}{dt}
$$
where \( h \) is the convective heat transfer coefficient between the cell surface and the coolant, \( A \) is the surface area of the energy storage cell, \( T_{\text{cell}} \) and \( T_{\text{coolant}} \) are the cell surface and coolant temperatures, respectively, \( m_{\text{cell}} \) is the cell mass, and \( c_{p,\text{cell}} \) is the specific heat capacity of the cell. In our immersion setup, the coolant provides a very high \( h \) value because of the direct contact and the high thermal conductivity of the hydrocarbon liquid. This effectively suppresses the temperature rise of the energy storage cell and prevents the exothermic chain reactions from reaching the auto-ignition point of the electrolyte. Furthermore, the dielectric nature of the coolant ensures that no electrical short circuits occur through the liquid itself, preserving the safety of the entire system.
We also evaluated the long-term thermal stability of the coolant under repeated thermal cycling conditions representative of normal energy storage cell operation. The coolant was subjected to 1,000 cycles between -40 °C and 85 °C, and no significant changes in viscosity, breakdown voltage, or oxidation stability were detected. This confirms that the coolant retains its protective properties over the entire lifespan of the energy storage cell system.
In addition to the experimental work, we developed a computational fluid dynamics (CFD) model to simulate the temperature distribution within a module containing multiple energy storage cells immersed in the coolant. The governing equation for the coolant flow under laminar conditions is the Navier-Stokes equation coupled with the energy equation. For an incompressible Newtonian fluid, the momentum balance is:
$$
\rho \frac{{D\mathbf{u}}}{{Dt}} = -\nabla p + \mu \nabla^2 \mathbf{u} + \rho \mathbf{g}
$$
where \( \rho \) is the density of the coolant, \( \mathbf{u} \) is the velocity vector, \( p \) is the pressure, \( \mu \) is the dynamic viscosity, and \( \mathbf{g} \) is the gravitational acceleration. The energy equation for the coolant is:
$$
\rho c_p \frac{{DT}}{{Dt}} = k \nabla^2 T + \Phi
$$
where \( c_p \) is the specific heat, \( k \) is the thermal conductivity, and \( \Phi \) is the viscous dissipation term. The boundary conditions at the surface of each energy storage cell include a heat flux that varies with the charge/discharge state. Our simulations showed that the maximum temperature difference among the energy storage cells in the module was less than 3 °C under the most demanding operating conditions, indicating excellent temperature uniformity—a critical requirement for extending the cycle life of lithium-ion batteries.
The economic implications of adopting immersion cooling for large-scale energy storage systems are also noteworthy. Although the initial cost of the coolant and the containment infrastructure is higher than that of air cooling, the benefits include longer battery life, reduced fire suppression system costs, and higher system reliability. For instance, we performed a life-cycle cost analysis assuming a 10-year operation period for a 100 MWh energy storage facility. The total cost of ownership for the immersion cooling system was found to be approximately 15% lower than that of an equivalent air-cooled system when accounting for the avoided thermal runaway incidents and the reduced need for battery replacements. This economic advantage is largely due to the superior thermal management provided by the immersion coolant, which keeps each energy storage cell within its optimal temperature window.
Furthermore, the environmental footprint of our developed coolant is minimal. The base oil is produced from a renewable crude source using energy-efficient refining processes, and the additive package is non-toxic and biodegradable. We conducted a standard OECD 301B ready biodegradability test and found that the coolant achieved 68% degradation within 28 days, meeting the criteria for “inherently biodegradable” classification. This ensures that any accidental spill during handling or disposal does not pose a persistent environmental hazard.
In summary, we have successfully developed a novel immersion oil-based coolant specifically tailored for lithium-ion energy storage cells. The product exhibits outstanding oxidation stability, high electrical insulation, excellent material compatibility, and unparalleled safety performance in preventing thermal runaway propagation. The comprehensive experimental and modeling results demonstrate that immersion liquid cooling using this coolant can become the mainstream thermal management technology for next-generation energy storage systems. Our work provides a robust technical foundation for the widespread adoption of safe, efficient, and reliable energy storage solutions.