The integration of renewable energy sources into the global power grid has dramatically increased the importance of electrochemical energy storage for ensuring grid stability and reliability. Among various technologies, lithium-ion battery energy storage systems have become a cornerstone due to their high energy density, long cycle life, and efficiency. However, the safe and efficient operation of these battery energy storage system is intrinsically linked to effective thermal management. During charging and discharging, especially at high rates, lithium-ion batteries generate significant heat. If this heat is not adequately dissipated, it can lead to elevated temperatures, accelerated degradation, and in extreme cases, thermal runaway—a critical safety hazard. Therefore, developing advanced cooling solutions is paramount for the longevity and safety of large-scale battery energy storage system installations.
Air cooling, while simple and cost-effective, often proves insufficient for high-density, high-power battery packs. Liquid cooling, particularly indirect cooling using cold plates, offers superior heat transfer capabilities and precise temperature control, making it the preferred method for demanding applications. The core challenge lies in designing a liquid cooling system that provides excellent thermal uniformity across a large battery module while minimizing the pumping power required to circulate the coolant. This work focuses on optimizing the thermal performance of a liquid-cooled module for a large-capacity LiFePO4 battery energy storage system. We propose a novel parallel-serpentine flow channel design, investigate strategic cooling plate arrangements, and explore differentiated flow distribution to achieve an optimal balance between cooling effectiveness and system energy consumption.
Thermal Modeling and Design of the Cooling Plate
The thermal behavior of a battery module under operation is governed by heat generation and transfer mechanisms. The total heat generation rate ($$Q_{gen}$$) within a lithium-ion battery can be modeled as a sum of irreversible joule heating, reversible entropic heat, and side reaction heat. For engineering analysis, a simplified approach often uses a volumetric heat source derived from the battery’s internal resistance and current. The governing energy balance equation for a battery cell can be expressed as:
$$\rho C_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + q”’$$
where $$\rho$$ is density, $$C_p$$ is specific heat capacity, $$k$$ is the thermal conductivity tensor, $$T$$ is temperature, $$t$$ is time, and $$q”’$$ is the volumetric heat generation rate. For a battery energy storage system module, this equation is solved for all cells coupled through conduction and convection boundaries at the cooling interfaces.
We base our study on a module comprising 30 commercial 208 Ah LiFePO4 prismatic cells (170 mm × 200 mm × 50 mm), arranged in three columns of ten cells each. The primary design variable is the internal flow channel structure of the aluminum cold plates placed between the cell columns. Three configurations were analyzed computationally:
- Traditional Serpentine Channel: A single, continuous winding path.
- Traditional Parallel Channel: Multiple straight, parallel flow paths.
- Proposed Parallel-Serpentine Channel: A hybrid design featuring main parallel branches with integrated local serpentine sections to enhance flow distribution and heat exchange.
Numerical simulations were conducted to evaluate the thermal performance and pressure drop of each design under a 1C charging scenario. The battery’s thermal properties and operating conditions used in the simulation are summarized in Table 1. A 50% ethylene glycol-water solution was selected as the coolant.
| Parameter | Value |
|---|---|
| Cell Density (kg/m³) | 2405 |
| Cell Specific Heat (J/kg·K) | 1329 |
| Cell Thermal Conductivity (X, Z, Y) (W/m·K) | 3.72, 28, 26 |
| Volumetric Heat Gen. Rate @1C (W/m³) | 6362 |
| Coolant Density (kg/m³) | 1070 |
| Coolant Specific Heat (J/kg·K) | 3396 |
| Coolant Thermal Conductivity (W/m·K) | 0.399 |
| Coolant Dynamic Viscosity (kg/m·s) | 0.00339 |
| Ambient/Inlet Temperature (°C) | 27 / 25 |
The simulation results revealed a critical trade-off. The traditional serpentine channel achieved the best thermal performance, yielding the lowest maximum module temperature and temperature difference. However, it incurred the highest pressure drop ($$\Delta P \approx 1253 \text{ Pa}$$), implying high pumping power. The traditional parallel channel had the lowest pressure drop ($$\Delta P \approx 101 \text{ Pa}$$) but resulted in the poorest thermal uniformity and higher maximum temperature. The proposed parallel-serpentine design struck an effective balance. It maintained a thermal performance close to the serpentine design while reducing the pressure drop significantly ($$\Delta P \approx 749 \text{ Pa}$$), which is only 59.8% of the serpentine channel’s pressure drop. This makes it an energy-efficient choice for a large-scale battery energy storage system where operational costs matter.
System Integration and Experimental Platform
To validate the simulation findings, a full-scale experimental platform for the battery energy storage system module was constructed. The optimized parallel-serpentine cold plates were manufactured from aluminum via a milling and friction stir welding process. Four identical cold plates were installed vertically in the module, creating three battery columns. The entire assembly was housed in a stainless-steel enclosure to mimic a realistic operating environment.
The experimental setup comprised several key subsystems:
1. Battery Module: 30 LiFePO4 cells connected in a configured series-parallel arrangement.
2. Liquid Cooling Circuit: Included a reservoir, a variable-speed pump, a chiller set to 25°C, and the network of parallel-serpentine cold plates.
3. Battery Cycler: To charge and discharge the module under controlled conditions (1C rate).
4. Data Acquisition System: Twenty T-type thermocouples were attached to the surface of selected cells in the outer and center columns to record temperature evolution.
The experimental procedure involved preconditioning the batteries, setting the coolant flow rate and temperature, and then performing a constant-current charge to the upper voltage limit while logging temperatures. This platform allowed for testing different thermal management strategies.
Optimization Strategies and Experimental Verification
Beyond the base design, two optimization strategies were proposed and tested to further enhance the performance of the battery energy storage system thermal management.
Strategy 1: Staggered vs. Aligned Cooling Plate Arrangement.
In a standard “aligned” setup, the coolant inlets for all four cold plates are located on the same side of the module. This can lead to a temperature gradient along the flow path, as the coolant warms up. We proposed a “staggered” arrangement, where adjacent cold plates have their inlets on opposite sides of the module. This creates a counter-flow-like effect, improving thermal uniformity.
Experimental Result: At a uniform inlet flow velocity of 0.1 m/s, the staggered arrangement reduced the maximum temperature of the module by 0.3°C (from 34.6°C to 34.3°C) compared to the aligned arrangement. More importantly, it significantly improved temperature uniformity within the most critical column, reducing the inter-cell temperature spread by 25% (from 1.2°C to 0.9°C). This was achieved without any increase in pumping power, demonstrating a simple yet effective hardware modification for the battery energy storage system.
| Battery Position | Max Temp. – Aligned (°C) | Max Temp. – Staggered (°C) |
|---|---|---|
| Cell 1 | 32.6 | 32.7 |
| Cell 2 | 33.2 | 33.1 |
| Cell 3 | 33.5 | 33.0 |
| Cell 4 | 33.7 | 33.6 |
| Cell 5 | 33.4 | 33.3 |
| Cell 6 | 33.7 | 33.5 |
| Cell 7 | 33.2 | 33.6 |
| Cell 8 | 33.6 | 33.3 |
| Cell 9 | 33.8 | 33.2 |
| Cell 10 | 33.8 | 32.9 |
| Column ΔT | 1.2 | 0.9 |
Strategy 2: Differentiated Flow Velocity Distribution.
In a multi-plate battery energy storage system, heat generation is not uniform; cells in the center of the module typically experience worse thermal conditions than those on the edges due to constrained heat dissipation paths. Applying a uniform high flow rate to all plates is effective but energy-inefficient. We proposed a differentiated flow strategy: assigning a higher flow rate to the cold plates cooling the inner, hotter battery columns and a lower flow rate to those cooling the outer columns.
Experimental Result: Two scenarios were compared with the cold plates in a staggered arrangement: 1) Uniform flow at 0.2 m/s for all plates. 2) Differentiated flow with 0.3 m/s for the two inner plates and 0.1 m/s for the two outer plates. The total coolant volumetric flow rate was kept approximately constant between the two scenarios. The differentiated strategy successfully lowered the module’s maximum temperature by an additional 0.2°C (from 32.9°C to 32.7°C) compared to the uniform high-flow case. This confirms that intelligently distributing cooling resources based on thermal demand can yield better performance for the same overall pumping power in a battery energy storage system.
| Thermal Management Condition | Module Maximum Temperature (°C) |
|---|---|
| Natural Convection (No Cooling) | 40.3 |
| Liquid Cooling – Base (Aligned, 0.1 m/s uniform) | 34.6 |
| Liquid Cooling – Staggered (0.1 m/s uniform) | 34.3 |
| Liquid Cooling – Staggered (0.2 m/s uniform) | 32.9 |
| Liquid Cooling – Staggered (Diff. Flow: 0.1/0.3 m/s) | 32.7 |
Conclusion
This work comprehensively addressed the thermal management challenge for a large-capacity LiFePO4 battery energy storage system module. The core innovation is a parallel-serpentine cold plate design that effectively balances high cooling performance with low hydraulic resistance, leading to reduced operational energy consumption. Through coupled simulation and experimental validation, two practical system-level optimizations were demonstrated:
- Staggered Cold Plate Arrangement: This simple mechanical design change improves thermal uniformity and slightly lowers peak temperature without any energy penalty.
- Differentiated Flow Velocity Control: This active control strategy allocates coolant flow preferentially to hotter regions, enhancing cooling efficiency for a given total pump power.
The final optimized system—featuring parallel-serpentine cold plates in a staggered arrangement with differentiated flow control—successfully maintained the module’s maximum temperature at 32.7°C during a 1C charge in a 27°C ambient environment. This represents a significant 7.6°C reduction compared to natural convection cooling, effectively keeping the battery energy storage system within a safe and optimal operating range. The strategies presented provide valuable guidelines for designing energy-efficient and high-performance thermal management systems for next-generation grid-scale energy storage applications.