In the context of the global transition towards renewable energy sources, battery energy storage systems have become increasingly vital for power regulation and supply-demand balancing. The rapid growth of lithium-ion battery installations, particularly in large-scale stationary energy storage containers, demands efficient thermal management to ensure safety, performance, and longevity. As an engineer deeply involved in this field, I have focused on enhancing the air-cooling efficiency of container-level battery energy storage systems through systematic CFD simulations. The challenge lies in the inherent low heat transfer coefficient of air, which often leads to large temperature gradients and local hotspots within the battery racks, potentially triggering thermal runaway. This work investigates the effects of cold aisle configuration (open vs. closed), air supply arrangement (underfloor vs. inter-column air conditioning), and exhaust outlet location (back vs. side) on the thermal field of a 20-foot equivalent unit (TEU) containerized battery energy storage system. Through transient three-dimensional simulations validated against experimental data, we identify a set of design guidelines that significantly improve temperature uniformity and reduce peak temperatures.
1. Model and Methodology
We constructed a half-symmetry CFD model of a standard container (4000 mm × 2440 mm × 2590 mm) containing 12 battery cabinets arranged in two rows. Each cabinet houses multiple battery cells (LiFePO₄, 50 Ah, 3.2 V). The cold aisle (2650 mm × 1180 mm × 1960 mm) sits between the two rows. In the baseline configuration, air enters through five floor grilles (each 530 mm × 530 mm) located below the cold aisle, flows upward through the aisle, and is drawn into each cabinet by fans at the front. Heated air exits through side outlets (1180 mm × 400 mm). All battery cells are modeled as anisotropic solid blocks with volumetric heat generation measured experimentally at various discharge rates. The governing equations for mass, momentum, and energy conservation are solved using the k-ε turbulence model, with gravity acting along the vertical y-direction. The air properties are listed in Table 1.
| Property | Value |
|---|---|
| Density ρ | 1.2 kg/m³ |
| Specific heat Cp | 1006.4 J/(kg·K) |
| Thermal conductivity λ | 2.4×10⁻² W/(m·K) |
| Dynamic viscosity μ | 1.8×10⁻⁵ kg/(m·s) |
| Thermal expansion coefficient β | 3.4×10⁻³ K⁻¹ |
The volumetric heat generation rate for the battery cells was obtained from controlled discharge experiments. Under a 3C discharge rate (150 A), the temperature rise of a single cell over 20 minutes was 12.47 °C, yielding a volumetric heat source q = 37,959.2 W/m³ according to Eq. (1):
$$
q = \frac{m C_p \Delta T}{t \cdot V} \tag{1}
$$
where m is the cell mass (0.785 kg), Cp is the specific heat of the battery core (1399 J/(kg·K)), ΔT is the temperature difference, t is the discharge duration, and V is the cell volume (0.185×0.135×0.030 m³). The anisotropic thermal conductivities of the cell are given in Table 2.
| Parameter | Value |
|---|---|
| λx (through-plane) | 1.0 W/(m·K) |
| λy (in-plane width) | 10.0 W/(m·K) |
| λz (in-plane height) | 21.0 W/(m·K) |
| Specific heat (core) | 1399 J/(kg·K) |
| Density | 1934 kg/m³ |
The transient heat conduction within the battery follows Eq. (2):
$$
\rho_b C_p \frac{\partial T}{\partial t} = \frac{\partial}{\partial x} \left( \lambda_x \frac{\partial T}{\partial x} \right) + \frac{\partial}{\partial y} \left( \lambda_y \frac{\partial T}{\partial y} \right) + \frac{\partial}{\partial z} \left( \lambda_z \frac{\partial T}{\partial z} \right) + q \tag{2}
$$
The fluid flow and heat transfer are governed by the incompressible Navier-Stokes equations:
$$
\frac{\partial \rho}{\partial t} + \nabla \cdot (\rho \mathbf{u}) = 0 \tag{3}
$$
$$
\frac{\partial (\rho \mathbf{v})}{\partial t} + \nabla \cdot (\rho \mathbf{u} \mathbf{u}) = -\nabla P + \rho \mathbf{g} \tag{4}
$$
$$
\frac{\partial T}{\partial t} + \nabla \cdot (\mathbf{u} T) = \frac{\lambda}{\rho C} \nabla^2 T \tag{5}
$$
The model was validated against the benchmark experimental data of a data center cooling scenario (Abdelmaksoud et al.), where the temperature difference between the rack inlet and the floor tile (T – Ttile) was compared at various heights. Our simulation results matched the measurements within 10% deviation, confirming the reliability of the CFD approach for predicting airflow and temperature distributions in large-scale battery energy storage systems.
2. Results and Discussion
2.1 Effect of Cold Aisle Configuration
Two cold aisle configurations were examined: an open aisle where the top and two side ends are open to the surrounding space, and a closed aisle where these surfaces are sealed, forcing all cooling air to pass through the battery cabinets. Table 3 summarizes the key flow and temperature metrics in the container.
| Parameter | Open Cold Aisle | Closed Cold Aisle |
|---|---|---|
| Maximum air velocity in cold aisle (y=1000 mm) | 6.8 m/s | 10.0 m/s |
| Maximum air temperature near battery inlet (x=1820 mm) | 39.8 °C | 39.5 °C |
| Temperature uniformity index (ΔTmax at y=1000 mm) | 8.4 °C | 4.6 °C |
The closed cold aisle significantly improved airflow organization by preventing mixing of cold supply air with warm return air. The maximum temperature on a single battery cell surface was reduced from 44.6 °C to 43.7 °C, and the maximum surface temperature difference across a cell was lowered by 15% (from 6.1 °C to 5.2 °C). The average cell temperature decreased by 0.8 °C (Table 4).
| Metric | Open Cold Aisle | Closed Cold Aisle |
|---|---|---|
| Maximum surface temperature (°C) | 44.6 | 43.7 |
| Maximum cell temperature difference (°C) | 6.1 | 5.2 |
| Average surface temperature (°C) | 29.4 | 28.6 |
These results clearly indicate that isolating the cold aisle is a fundamental strategy for enhancing the cooling efficiency of battery energy storage systems. The closed configuration also reduces the required fan power by providing a higher pressure differential across the cabinets.
2.2 Effect of Air Supply Mode
Next, we compared two common air supply arrangements for the closed cold aisle: underfloor air supply (baseline) and inter-column air conditioning (AC) units placed between the battery cabinets. In the inter-column AC scenario, the floor grilles were sealed and the AC units blew cold air horizontally into the cold aisle. Table 5 shows the results at a horizontal plane (y=1000 mm).
| Parameter | Underfloor Supply | Inter-column AC Supply |
|---|---|---|
| Maximum air velocity (m/s) | 10.0 | 10.6 |
| Maximum air temperature (°C) | 29.6 | 30.6 |
| Maximum temperature difference in plane (°C) | 4.6 | 5.6 |
The inter-column AC supply created vortices near the inlets of cabinets adjacent to the AC units, as shown in the velocity vector field. These vortices reduced the effective airflow into certain cabinets, leading to uneven cooling. Consequently, the maximum battery surface temperature increased by 0.4 °C, and the cell temperature difference rose to 5.6 °C (Table 6).
| Metric | Underfloor Supply | Inter-column AC Supply |
|---|---|---|
| Maximum surface temperature (°C) | 43.7 | 44.1 |
| Maximum cell temperature difference (°C) | 5.2 | 5.6 |
| Average surface temperature (°C) | 28.6 | 28.7 |
The underfloor supply mode provided a more uniform velocity distribution across the cabinets because the air enters over a large area and is naturally guided upward by buoyancy and the cold aisle enclosure. Therefore, for large-scale battery energy storage systems, underfloor air supply is preferred over inter-column AC units to maintain low temperatures and small thermal gradients.
2.3 Effect of Exhaust Air Outlet Location
Three exhaust outlet positions were evaluated: two side outlets (left and right) near the top of the container, and one rear outlet located at the back wall centered on the battery cabinets. All outlets had the same area (1180 mm × 400 mm). Table 7 summarizes the flow conditions at the vertical plane x=1820 mm (mid-plane of the container).
| Outlet Location | Max Velocity (m/s) | Max Temperature (°C) | ΔTmax in plane (°C) |
|---|---|---|---|
| Side (left) | 6.4 | 39.5 | 14.5 |
| Side (right) | 6.5 | 39.6 | 14.6 |
| Rear (back) | 23.8 | 39.1 | 14.6 |
When the exhaust was located on the rear side, the flow path became shorter and the air resistance decreased, as evidenced by the high local velocity (23.8 m/s) near the rear outlet. This enhanced convection and reduced the peak air temperature slightly. More importantly, the maximum surface temperature difference across a single battery cell decreased to 4.9 °C, which is 4.1% lower than the left-side outlet and 3.2% lower than the right-side outlet (Table 8). The average surface temperature was also slightly lower (28.3 °C vs. 28.6 °C for side outlets).
| Outlet Location | Max Surface Temp (°C) | Cell ΔTmax (°C) | Average Temp (°C) |
|---|---|---|---|
| Side (left) | 43.7 | 5.2 | 28.6 |
| Side (right) | 44.0 | 5.1 | 28.2 |
| Rear (back) | 42.9 | 4.9 | 28.3 |
The rear exhaust design promotes a more direct flow path from the cold aisle through the cabinets to the back, reducing the formation of recirculation zones observed with side outlets. However, it should be noted that the rear outlet leads to higher temperatures near the bottom cabinets due to the shorter path; this can be mitigated by adjusting the supply airflow distribution (e.g., increasing flow to bottom cabinets). Overall, the rear exhaust configuration is recommended for battery energy storage systems to achieve the best thermal uniformity.
3. Conclusion
Through systematic CFD simulations of a container-level battery energy storage system, we have identified key design parameters that significantly improve air-cooling performance. The closed cold aisle configuration reduces the maximum cell temperature difference by 15% compared to an open aisle. Underfloor air supply provides more uniform airflow and lower maximum temperatures than inter-column AC units. Relocating the air exhaust to the rear of the container reduces the maximum cell temperature difference to 4.9 °C, outperforming side outlets by up to 4.1%. These combined strategies offer a practical and cost-effective solution for enhancing the safety and reliability of large-scale battery energy storage systems. Future work will explore the interaction of these parameters under transient operational cycles and partial loading conditions to further optimize the system design.
By implementing these findings, engineers can design battery energy storage systems with superior thermal management, thereby extending battery life and reducing the risk of catastrophic failures in grid-scale applications.