In this study, I explore the application of immersion liquid cooling technology for thermal management in energy storage batteries, leveraging numerical simulations and theoretical heat transfer analysis. With the growing adoption of electrochemical energy storage systems, particularly in data centers to reduce electricity costs and enhance renewable energy utilization, efficient thermal control has become paramount. Energy storage batteries, especially lithium-ion variants, require precise temperature regulation to maintain performance, safety, and longevity. Traditional air cooling methods face limitations as energy density and scale increase, prompting a shift toward liquid cooling solutions that offer higher heat dissipation efficiency. Immersion liquid cooling, where batteries are directly submerged in dielectric coolants, represents a promising approach due to its ability to utilize the entire battery surface for heat exchange and minimize thermal resistance. I focus on single-phase immersion cooling, comparing three common coolants—fluorinated liquid, silicone oil, and mineral oil—to evaluate their cooling effectiveness for energy storage batteries under typical operating conditions.
The immersion liquid cooling process involves submerging energy storage batteries in a dielectric fluid, allowing direct heat transfer from the battery surfaces to the coolant. During charge-discharge cycles, energy storage batteries generate heat, which is absorbed by the fluid. A circulation pump moves the heated coolant to an external heat exchanger for cooling before returning it to the battery enclosure, maintaining optimal temperatures. This method eliminates the need for thermal interface materials, reduces contact resistance, and ensures uniform temperature distribution across battery modules. In my analysis, I consider steady-state conditions where the coolant undergoes no phase change, relying solely on convective heat transfer. The direct immersion enables 100% utilization of the battery exterior for散热, facilitating multidirectional heat flow and enhancing thermal homogeneity. For energy storage batteries, which are often arranged in packs, this technology can mitigate hot spots and prevent thermal runaway, critical for large-scale deployments.
To assess the cooling performance, I developed a numerical model using Ansys Icepak software, simulating a battery pack comprising 13 prismatic lithium iron phosphate (LFP) energy storage batteries. The pack dimensions are 204 mm in width, 1038 mm in depth, and 235 mm in height, with batteries spaced 6 mm apart and 15 mm from the enclosure walls. The coolant flows in a bottom-inlet and top-outlet configuration, with inlet and outlet ports on the same side for practical tubing arrangements. I simplified the model by neglecting internal battery structures to improve computational efficiency, assuming uniform material properties and heat generation within each energy storage battery. The batteries are arranged horizontally within a sealed enclosure, and the coolant circulates through 8 mm diameter ports. The overall geometry is designed to replicate typical energy storage battery modules used in stationary applications.
The heat generation in energy storage batteries is modeled using the Bernardi heat generation rate equation, which accounts for both joule heating and reversible reaction heat. The model assumes uniform internal heat generation, consistent material properties, and negligible internal convection and radiation. The equation is expressed as:
$$q = \frac{I}{V_b} \left[ (E_0 – E) – T \frac{dE_0}{dT} \right] + I^2 R$$
where \(q\) is the volumetric heat generation rate (W/m³), \(I\) is the current (A, positive for discharge), \(V_b\) is the battery volume (m³), \(E_0\) is the open-circuit voltage (V), \(E\) is the operating voltage (V), \(T\) is the battery temperature (K), \(\frac{dE_0}{dT}\) is the temperature coefficient of open-circuit voltage (V/K), and \(R\) is the internal resistance (Ω). For the LFP energy storage batteries in this study, key parameters are summarized in Table 1. The batteries operate at a 1C discharge rate, considered a high rate for energy storage applications, with an assumed working temperature of 30°C. Using the Bernardi model, the heat generation per battery is calculated as 24.2 W, resulting in a total heat load of 314.6 W for the pack. This heat load is critical for evaluating coolant performance in maintaining temperature limits for energy storage batteries.
| Parameter | Value |
|---|---|
| Rated Capacity (Ah) | 280 |
| Rated Voltage (V) | 3.2 |
| Internal Resistance (mΩ) | 0.2 |
| Dimensions (mm) | 205 × 174 × 72 |
| Thermal Conductivity (W/m·K) | 24 in width/height, 0.6 in depth |
| Open-Circuit Voltage Temperature Coefficient (mV/K) | -0.1 |
I selected three coolants for comparison: fluorinated liquid, silicone oil, and mineral oil, commonly used in immersion cooling for data centers and now adapted for energy storage batteries. Their thermophysical properties, listed in Table 2, influence heat transfer efficiency. Fluorinated liquid exhibits low viscosity and high density, silicone oil offers balanced properties, and mineral oil has higher viscosity but favorable specific heat. These coolants are evaluated under identical boundary conditions to ensure a fair comparison for cooling energy storage batteries.
| Property | Fluorinated Liquid | Silicone Oil | Mineral Oil |
|---|---|---|---|
| Thermal Conductivity (W/m·K) | 0.0623 | 0.134 | 0.136 |
| Specific Heat Capacity (J/kg·K) | 1014 | 1800 | 2150 |
| Density (kg/m³) | 1830 | 940 | 805 |
| Kinematic Viscosity (mm²/s) | 1.3 | 10 | 20 |
| Volumetric Expansion Coefficient (1/K) | 0.0014 | 0.0011 | 0.0007 |
The simulation employs steady-state assumptions with fully developed fluid flow, constant heat source power, and adiabatic enclosure walls to isolate coolant effects. Boundary conditions include inlet fans set as intake with specified flow rates and temperatures, outlet openings, and fixed heat generation for each energy storage battery. The mesh consists of approximately 5.47 million elements to ensure accuracy. Coolant flow rates are adjusted to maintain a 2°C temperature difference between inlet and outlet, with inlet temperature fixed at 25°C for all cases. Table 3 details the工艺 parameters, where volume flow rates are derived from thermal balance considerations to achieve the target温差 for energy storage battery cooling.
| Coolant | Inlet Temperature (°C) | Outlet Temperature (°C) | Volume Flow Rate (L/min) | Mass Flow Rate (kg/s) |
|---|---|---|---|---|
| Fluorinated Liquid | 25 | 27 | 5.09 | 0.155 |
| Silicone Oil | 25 | 27 | 5.59 | 0.0874 |
| Mineral Oil | 25 | 27 | 5.45 | 0.0732 |
Simulation results reveal significant differences in cooling performance among the coolants for energy storage batteries. As shown in Table 4, under 1C discharge and a 2°C inlet-outlet温差, fluorinated liquid achieves the lowest maximum surface temperature and smallest temperature spread across the battery pack, followed by silicone oil and mineral oil. The average surface temperature rise relative to inlet coolant is restrained to 3–4°C, with pack temperature differences below 5°C, meeting typical thermal management requirements for energy storage batteries. This demonstrates the superior temperature control capability of immersion liquid cooling, ensuring energy storage batteries operate within the optimal 15–35°C range and minimizing risks of thermal runaway.
| Coolant | Minimum Surface Temperature (°C) | Maximum Surface Temperature (°C) | Surface Temperature Difference (°C) | Average Surface Temperature (°C) |
|---|---|---|---|---|
| Fluorinated Liquid | 26.8 | 29.1 | 2.3 | 28.1 |
| Silicone Oil | 26.2 | 30.5 | 4.3 | 28.2 |
| Mineral Oil | 26.3 | 31.3 | 5.0 | 28.6 |
To explain these outcomes, I delve into heat transfer theory, analyzing the interplay between forced and natural convection in cooling energy storage batteries. The flow regime is assessed using the Grashof number (\(Gr\)) for natural convection and Reynolds number (\(Re\)) for forced convection. The ratio \(Gr/Re^2\) indicates the dominance of buoyancy versus inertia forces. For the vertical battery surfaces, where natural convection is significant, the Grashof number is defined as:
$$Gr = \frac{g \beta \Delta T L^3}{\nu^2}$$
where \(g\) is gravitational acceleration (9.81 m/s²), \(\beta\) is the volumetric expansion coefficient (1/K), \(\Delta T\) is the temperature difference (K), \(L\) is the characteristic length (m), and \(\nu\) is the kinematic viscosity (m²/s). The Reynolds number is given by:
$$Re = \frac{UL}{\nu}$$
with \(U\) as the flow velocity (m/s). The ratio \(Gr/Re^2\) simplifies to:
$$\frac{Gr}{Re^2} = \frac{g \beta \Delta T L}{U^2}$$
In all cases, \(Gr/Re^2\) exceeds 10, indicating that natural convection effects are negligible compared to forced convection, validating the primary role of pump-driven flow in cooling energy storage batteries. However, for vertical surfaces, natural convection still contributes, and the heat transfer coefficient \(h\) can be derived from empirical correlations for natural convection over vertical plates. For \(10^4 \leq Gr \cdot Pr \leq 10^9\), where \(Pr\) is the Prandtl number, the Nusselt number (\(Nu\)) correlation is:
$$Nu = 0.59 (Gr \cdot Pr)^{1/4}$$
The Prandtl number is \(Pr = \frac{\nu}{\alpha}\), with \(\alpha = \frac{k}{\rho c_p}\) as thermal diffusivity, where \(k\) is thermal conductivity, \(\rho\) is density, and \(c_p\) is specific heat. The heat transfer coefficient is then:
$$h = \frac{Nu \cdot k}{L}$$
Combining these, \(h\) can be expressed as:
$$h = 0.59 \left( \frac{g \beta \rho^2 c_p k^2}{\mu L} \right)^{1/4} \Delta T^{1/4}$$
where \(\mu\) is dynamic viscosity. This shows that \(h\) is proportional to \(\beta\), \(c_p\), \(\rho\), and \(k\), and inversely proportional to \(\nu\). Using property data from Table 2, I calculate \(h\) values: approximately 55.6 W/m²·K for fluorinated liquid, 53.3 W/m²·K for silicone oil, and 43.1 W/m²·K for mineral oil. Fluorinated liquid’s high \(h\) stems from its low viscosity and high \(\beta\), enhancing heat removal from energy storage batteries. Silicone oil’s moderate properties yield a comparable \(h\), while mineral oil’s higher viscosity reduces \(h\), aligning with the simulation results.
Newton’s law of cooling, \(Q = h S \Delta T\), where \(Q\) is heat transfer rate, \(S\) is surface area, and \(\Delta T\) is temperature difference, further elucidates the findings. For a fixed \(Q\) and \(S\), higher \(h\) results in lower \(\Delta T\), explaining why fluorinated liquid minimizes temperature rise in energy storage batteries. The theoretical analysis corroborates that coolant selection profoundly impacts thermal performance, with fluorinated liquid offering the best cooling, silicone oil as a viable alternative, and mineral oil less effective due to its thermophysical limitations. These insights are crucial for designing immersion cooling systems for energy storage batteries, where maintaining narrow temperature spreads is essential for longevity and safety.
Beyond the base case, I extend the analysis to explore parametric influences on cooling energy storage batteries. Variations in discharge rate, inlet temperature, and flow arrangement can alter thermal behavior. For instance, at higher discharge rates (e.g., 2C), heat generation increases, necessitating higher flow rates or enhanced coolant properties to prevent excessive温升. Similarly, lower inlet temperatures may reduce average temperatures but require more energy for coolant chilling. To generalize, I derive dimensionless groups that govern heat transfer in immersion-cooled energy storage batteries. The overall energy balance for the battery pack is:
$$Q_{total} = \dot{m} c_p (T_{out} – T_{in})$$
where \(\dot{m}\) is mass flow rate, and \(T_{in}\) and \(T_{out}\) are inlet and outlet temperatures. Combining with the heat transfer equation, the required flow rate for a target \(\Delta T_{pack}\) (battery pack temperature rise) can be estimated as:
$$\dot{m} = \frac{Q_{total}}{c_p (T_{out} – T_{in})}$$
This highlights the importance of coolant \(c_p\) in determining flow requirements for energy storage batteries. Fluorinated liquid’s lower \(c_p\) necessitates higher mass flow rates compared to silicone oil, as seen in Table 3, to achieve the same outlet temperature.
Furthermore, I examine temperature uniformity within energy storage battery packs, a critical factor for preventing cell degradation. The standard deviation of surface temperatures, \(\sigma_T\), can be related to coolant properties and flow distribution. For a well-designed flow path, \(\sigma_T\) is minimized when the coolant has high thermal diffusivity \(\alpha = k/(\rho c_p)\). Calculating \(\alpha\) from Table 2: fluorinated liquid (~3.4 × 10⁻⁸ m²/s), silicone oil (~7.9 × 10⁻⁸ m²/s), and mineral oil (~7.9 × 10⁻⁸ m²/s). Despite similar \(\alpha\) for silicone and mineral oils, fluorinated liquid’s lower \(\alpha\) is offset by its superior convection, leading to better uniformity in practice. This underscores the need for holistic design when implementing immersion cooling for energy storage batteries.
To validate the simulation approach, I compare results with analytical solutions for simplified geometries. For a single energy storage battery submerged in coolant, the steady-state temperature distribution can be approximated using conduction-convection models. Assuming the battery as a rectangular solid with internal heat generation \(q\), the surface temperature \(T_s\) relates to coolant bulk temperature \(T_{\infty}\) via:
$$T_s = T_{\infty} + \frac{q V_b}{h S}$$
For the LFP battery with \(V_b \approx 2.57 \times 10^{-3} \, \text{m}^3\) and \(S \approx 0.124 \, \text{m}^2\), using \(h\) from above, \(T_s – T_{\infty}\) estimates are ~3.5°C for fluorinated liquid, close to simulation averages. This consistency reinforces the reliability of numerical models for energy storage battery thermal analysis.
In addition to steady-state analysis, transient effects during charge-discharge cycles are relevant for real-world energy storage batteries. The thermal time constant \(\tau\) of the system, given by \(\tau = \frac{\rho_b c_{p,b} V_b}{h S}\), where \(\rho_b\) and \(c_{p,b}\) are battery density and specific heat, influences temperature fluctuations. For typical LFP batteries, \(\tau\) ranges from minutes to hours, implying that immersion cooling can quickly respond to thermal loads, enhancing stability for energy storage batteries in dynamic applications like frequency regulation.
Economic and environmental considerations also play a role in coolant selection for energy storage batteries. Fluorinated liquids, while effective, may have higher costs and environmental impacts compared to silicone or mineral oils. Lifecycle assessments should factor in coolant longevity, maintenance needs, and disposal regulations. For large-scale energy storage deployments, silicone oil could offer a balance of performance and sustainability, supporting the adoption of immersion cooling for energy storage batteries.
Future work could optimize flow channel designs within battery packs to further improve temperature homogeneity for energy storage batteries. Computational fluid dynamics (CFD) studies can explore baffles, jets, or distributed inlets to enhance coolant distribution. Additionally, experimental validation with prototype energy storage battery modules would strengthen the findings, addressing real-world factors like aging and coolant degradation.
In conclusion, my simulation and theoretical research demonstrate that immersion liquid cooling is highly effective for thermal management of energy storage batteries. Among the coolants evaluated, fluorinated liquid provides the best cooling performance, followed by silicone oil and mineral oil, based on their thermophysical properties and derived heat transfer coefficients. Under 1C discharge with a 2°C inlet-outlet温差, the technology maintains energy storage battery surface temperature rises within 3–4°C and pack temperature differences below 5°C, meeting stringent operational requirements. The analysis underscores the importance of coolant properties like viscosity, thermal conductivity, and specific heat in designing efficient cooling systems for energy storage batteries. As energy storage systems evolve toward higher densities and scales, immersion liquid cooling offers a robust solution to ensure safety, efficiency, and longevity, paving the way for advanced thermal management strategies in renewable energy integration and grid stabilization applications.