As the global energy structure accelerates its transition toward low-carbon and renewable sources, technologies such as wind power, photovoltaics, hydroelectricity, and biomass have become central drivers of power system innovation. These clean energy sources provide significant environmental benefits and inherent renewability; however, their distributed and intermittent characteristics—such as the diurnal fluctuations of solar power and seasonal variability of wind energy—lead to unstable grid power quality. In some cases, without effective regulation, this variability has been referred to as “junk electricity.” The breakthrough development of energy storage technology has fundamentally changed this situation. By leveraging the energy time-shifting capability of “peak shaving and valley filling,” energy storage systems successfully achieve dynamic matching between renewable power generation and grid load, becoming the critical hub for building new power systems.
Among the diverse energy storage technology routes, electrochemical energy storage, particularly lithium-ion battery energy storage, has gradually become the mainstream choice for grid-scale storage power stations, owing to its modular design, fast response characteristics, and flexible deployment. However, lithium-ion batteries generate heat during the charging and discharging processes due to internal resistance and internal chemical reactions. If this heat is not removed promptly and effectively, the temperature of the energy storage battery will rise, which can first degrade battery performance and, in severe cases, trigger thermal runaway. Moreover, inconsistent temperatures among individual cells within a battery pack can lead to different aging rates. Over long-term operation, the accumulation of these cell-to-cell differences can result in reduced overall pack capacity, lower safety, and diminished charging/discharging efficiency.
Establishing the optimal operating temperature range for lithium-ion batteries is crucial for the safe and efficient operation of energy storage systems. Extensive research indicates that the best working temperature range is approximately 293 K to 313 K, while temperature uniformity should be maintained within 5 K. When the temperature difference inside a battery pack exceeds 5 K, the cycle life can be reduced by more than 30%. Furthermore, lithium embedded in graphite can react with electrolyte components and binders at temperatures above 393 K, which is widely recognized as the high-risk threshold for thermal runaway. These constraints highlight the essential role of battery thermal management systems in determining the economic viability and reliability of energy storage installations.
In practical grid operations, when renewable generation power suddenly increases, the energy storage battery must possess a high-rate charging capability to quickly absorb excess electricity and prevent overvoltage problems. Conversely, during peak demand periods, the battery must discharge at high rates to support grid stability. Without effective thermal management, such instantaneous high-rate charge and discharge processes inevitably lead to rapid temperature changes and large temperature differences within the battery pack. Therefore, to ensure that grid-scale energy storage lithium-ion batteries operate safely within the optimal temperature window, efficient cooling is indispensable. Common cooling techniques for lithium-ion batteries include air cooling, liquid cooling, and phase-change cooling. Among these, liquid cooling stands out because of its fast heat removal capability, strong stability, and significant advantages in reducing maximum cell temperature and improving temperature uniformity within modules.
Several researchers have investigated liquid cooling approaches. Some have built numerical models of lithium iron phosphate batteries using ANSYS Fluent and compared serpentine flow channel layouts in both transversal and longitudinal orientations, analyzing the effect of flow rate on cooling performance. Their findings indicated that transversal serpentine channels achieved lower pressure drop, whereas longitudinal serpentine channels provided superior cooling effectiveness. In addition, increasing coolant flow rate improves the overall cooling effect but leads to higher energy consumption. Other studies have demonstrated that serpentine and tree-like or fractal cooling plate configurations can effectively control temperature uniformity in battery thermal management systems, with optimized designs keeping battery temperature below 313.15 K and temperature differences under 5 K.
Further research focused on vertical liquid-cooling thermal management systems proposed a gradient flow velocity optimization strategy. The experimental results showed that increasing the flow velocity gradient reduces the maximum temperature difference before the steady-state phase, and the gradient flow velocity design helps balance the heat transfer efficiency of different sections of the battery module. Investigations on rectangular microchannel cooling systems for battery thermal management systematically examined the effects of microchannel number, channel dimensions, coolant flow rate, and coolant temperature. When using five microchannels with a width of 4 mm, a coolant mass flow rate of 0.003 kg/s, and an inlet temperature of 298.15 K, the battery temperature could be maintained in the ideal range of 298.15 K to 313.15 K.
Another study aimed at optimizing the cooling performance of a lithium-ion battery module used the maximum temperature and maximum temperature difference as objective functions. Through simulation, the effects of coolant velocity, coolant temperature, cooling channel number, and water inlet direction were analyzed under constant current conditions. The optimized parameters derived from the constant current study were then validated under actual storage peak-shaving and frequency-regulation conditions, leading to the introduction of a variable flow velocity strategy for liquid cooling structures.
It is evident that liquid cooling is an effective technical approach for dispersing heat from energy storage lithium-ion batteries during high-rate charge and discharge operations. Moreover, the configuration of the liquid cooling plate has a decisive impact on the heat dissipation performance of the battery module. However, most current liquid cooling research has focused on conventional charge and discharge rates for lithium-ion batteries. There is limited research on liquid cooling techniques for energy storage lithium-ion batteries under high-rate and ultra-high-rate charge/discharge conditions. This gap motivates the design of a novel bidirectional counter-flow heat exchange plate capable of resolving the issues of excessive temperature and excessive temperature differences during high-rate and ultra-high-rate operations.
In this article, we present a comprehensive numerical study comparing three liquid cooling configurations for an energy storage battery pack. The baseline configuration uses a conventional unidirectional flow heat exchange plate placed at the bottom of the battery pack. The second configuration uses a unidirectional flow plate placed on the large side surfaces of each cell. The third configuration, which is the primary contribution of this work, is a bidirectional counter-flow heat exchange plate also placed on the side surfaces. We evaluate the thermal management performance of these three configurations under conventional (1 C), high (3 C), and ultra-high (5 C) charge/discharge rates. The results demonstrate the superior capability of the bidirectional counter-flow design in maintaining both low maximum temperature and low temperature differences across a wide range of operating conditions, thereby providing a robust technical solution for grid-scale energy storage battery safety.
The remainder of this article is organized as follows: the numerical model describing the geometry, governing equations, boundary conditions, and grid independence verification is presented; then, the simulation results for each cooling scheme under different C-rates are discussed in detail; finally, a heat transfer analysis of the heat exchange plates and the conclusions are provided.
Numerical Model
Geometry Description
The energy storage battery pack model constructed in this study consists of ten blade-type lithium-ion battery cells connected in series. The thermophysical and geometric parameters of the cell are as follows: the cell material is lithium iron phosphate (LiFePO₄); the cell capacity is 135 Ah; the nominal voltage is 3.2 V; the thermal conductivity is 18.3 W/(m·K) in the in-plane directions and 1.1 W/(m·K) in the thickness direction; the specific heat capacity is 1100 J/(kg·K); the density is 1715 kg/m³; and the cell dimensions are 945 mm in length, 14 mm in thickness, and 90 mm in height.
Due to the internal layered structure of the lithium-ion battery, which consists of electrodes, electrolytes, and separators, the overall thermal conductivity of the battery pack is relatively low. Moreover, the thermal conductivity exhibits anisotropy, with the in-plane (tangential) thermal conductivity [18.3 W/(m·K)] being significantly higher than the through-plane (normal) thermal conductivity [1.1 W/(m·K)]. This anisotropic nature strongly influences the heat transfer paths within the cell and, consequently, the effectiveness of different cooling configurations.
To clarify the advantages of the proposed side-mounted bidirectional counter-flow heat exchange plate, three distinct cooling schemes are compared. Scheme 1 corresponds to a conventional unidirectional flow heat exchange plate placed at the bottom of the battery pack. The plate features an eight-channel unidirectional straight flow path, with coolant entering from Inlet1 and exiting from Outlet2. The liquid cooling plate is positioned under the battery pack, which is the classic arrangement for liquid-cooled battery modules.
Scheme 2 places the same unidirectional straight-flow heat exchange plate between the large side surfaces of adjacent cells. In this case, coolant also flows from Inlet1 to Outlet2 through the eight parallel channels. Placing the plate at the side surface significantly increases the effective heat transfer area between the plate and the battery cell compared to the bottom placement.
Scheme 3, the primary innovation of this work, employs a bidirectional counter-flow heat exchange plate mounted on the large side surfaces of the cells. Unlike the previous two schemes, this bidirectional plate has two separate flow paths. Two inlets are both located at the bottom position, while two outlets are located at the top position. Coolant enters from Inlet1 and Inlet2 simultaneously and leaves through Outlet1 and Outlet2, respectively. The two streams of fluid perform cross counter-current flow inside the cooling plate, enabling counter-flow heat exchange. Additionally, an air layer is integrated between the bidirectional counter-flow heat exchange plates, which provides thermal insulation between adjacent battery cells. The unique dual-inlet (bottom) and dual-outlet (top) layout, combined with the central air insulation layer, achieves efficient heat dissipation and thermal isolation. This configuration is designed to promote heat transfer from the top region of the cell downward, thereby mitigating the large vertical temperature gradient that is typical of bottom cooling approaches.
In all schemes, a thermally conductive silicone gap filler is applied between the cooling plate and the battery cells to ensure good thermal contact. The geometry of the three schemes is illustrated schematically in the original article (the corresponding figures are omitted here for brevity). For clarity, we define the coordinate system such that the x-axis corresponds to the cell thickness direction, the y-axis to the cell height direction (top-bottom), and the z-axis to the cell length direction (coolant main flow direction).
Governing Equations and Heat Generation Model
For the heat generation rate inside the lithium-ion battery, the widely used Bernardi model is adopted. The volumetric heat generation rate is expressed as:
$$ q = \frac{I}{V_b} \left[ \left( E_0 – U \right) – T \frac{dE_0}{dT} \right] = \frac{1}{V_b} \left( I^2 R – I T \frac{dE_0}{dT} \right) $$
where \(V_b\) is the volume of the lithium-ion battery cell (m³), \(E_0\) is the open-circuit voltage (V), \(U\) is the terminal voltage (V), \(T\) is the instantaneous temperature of the battery (K), \(I\) is the charging/discharging current (A), and \(R\) is the internal resistance (Ω). The first term \(I^2R\) represents irreversible ohmic heat, while the second term \(-IT\,dE_0/dT\) represents reversible entropic heat.
The three-dimensional heat conduction within the battery cell is anisotropic and governed by:
$$ \rho C_p \frac{\partial T}{\partial t} = \lambda_x \frac{\partial^2 T}{\partial x^2} + \lambda_y \frac{\partial^2 T}{\partial y^2} + \lambda_z \frac{\partial^2 T}{\partial z^2} + Q $$
where \(\rho\) is the cell density (kg/m³), \(C_p\) is the cell specific heat capacity (J/(kg·K)), \(\lambda_x\), \(\lambda_y\), and \(\lambda_z\) are the thermal conductivities in the three orthogonal directions (W/(m·K)), and \(Q\) is the volumetric heat generation rate (W/m³).
Inside the cooling plate, the coolant flow is governed by the continuity, momentum, and energy equations for incompressible Newtonian fluids. The continuity equation is:
$$ \frac{\partial \rho_w}{\partial t} + \nabla \cdot \left( \rho_w \mathbf{v} \right) = 0 $$
The momentum equation is:
$$ \frac{\partial \mathbf{v}}{\partial t} + \left( \mathbf{v} \cdot \nabla \right) \mathbf{v} = -\frac{\nabla p}{\rho_w} + \frac{\mu}{\rho_w} \nabla^2 \mathbf{v} + \mathbf{g} $$
where \(\mathbf{v}\) is the velocity vector (m/s), \(p\) is the pressure (Pa), \(\rho_w\) is the coolant density (kg/m³), \(\mu\) is the dynamic viscosity (Pa·s), and \(\mathbf{g}\) is the gravitational acceleration (m/s²).
The energy equation for the coolant is:
$$ \frac{\partial}{\partial t} \left( \rho_w C_{p,w} T_w \right) + \nabla \cdot \left( -k_w \nabla T_w \right) + \rho_w C_{p,w} \mathbf{v} \cdot \nabla T_w = 0 $$
where \(C_{p,w}\) is the coolant specific heat capacity (J/(kg·K)), \(T_w\) is the coolant temperature (K), and \(k_w\) is the coolant thermal conductivity (W/(m·K)).
For the solid cooling plate body, the steady energy conservation equation is:
$$ \frac{\partial}{\partial t} \left( \rho_n C_{p,n} T_n \right) + \nabla \cdot \left( -k_n \nabla T_n \right) = 0 $$
where \(\rho_n\), \(C_{p,n}\), \(T_n\), and \(k_n\) are the density, specific heat capacity, temperature, and thermal conductivity of the cooling plate material, respectively.
Boundary Conditions and Simulation Setup
The initial temperature of the battery cells and the ambient temperature are both set to 298 K. A velocity inlet boundary condition is imposed at the coolant inlet, while a pressure outlet boundary condition is applied at the coolant outlet. The heat generation power of the battery pack is calculated according to the Bernardi model for three different C-rates: conventional rate (1 C), high rate (3 C), and ultra-high rate (5 C). The total coolant flow rate at the liquid cooling plate inlet is set to 12 L/min, and the inlet temperature is 293.15 K. The coolant is a 50% by mass aqueous ethylene glycol solution, which is commonly used in battery thermal management systems to provide freeze protection and corrosion resistance.
Steady-state simulations are performed using the ANSYS Fluent software. The heat generation parameters for the battery cell are taken from the literature [16]. To resolve the temperature and flow fields accurately, local mesh refinement is applied to the thermally conductive silicone layer, the liquid cooling plate, the insulation layer, and the fluid domain. Six different grid densities are tested to verify grid independence. The number of mesh elements and the corresponding computed pressure drop between inlet and outlet are summarized in Table 1.
| Mesh count (×10⁴) | 133 | 236 | 289 | 316 | 423 | 513 |
|---|---|---|---|---|---|---|
| Inlet-outlet pressure drop (kPa) | 13.60 | 13.90 | 14.20 | 14.80 | 14.90 | 14.93 |
As the number of grid elements increases from 133×10⁴ to 513×10⁴, the pressure drop changes less and less, indicating that the solution is essentially grid-independent. To balance computational accuracy and cost, the mesh with 423×10⁴ elements is adopted for all subsequent simulations.
Results and Discussion
Thermal Performance at Conventional Rate (1 C)
The simulation results for the three schemes under a conventional 1 C charging condition are presented in Figure 4, Figure 5, and Figure 6 in the original article. Here we describe the key findings quantitatively.
For Scheme 1 (bottom liquid cooling), the cell temperature distribution ranges from 294.5 K to 301 K. Although the maximum temperature is within the acceptable range, the maximum temperature difference across the entire battery pack reaches 6.5 K. This value violates the 5 K uniformity requirement. The temperature gradient along the coolant flow direction is relatively small; the dominant temperature variation occurs between the top and bottom of the cells. The bottom region, which is in direct contact with the cooling plate, is maintained at a lower temperature, while the top region becomes hotter. Because of the low thermal conductivity of the battery pack in the vertical direction, the heat generated at the top cannot be conducted efficiently downward to the cooling plate, resulting in a large top-to-bottom temperature difference.
For Scheme 2 (side-mounted unidirectional flow cooling), the liquid cooling plate is placed parallel to the large surface of the cell, which enlarges the contact area and shortens the internal heat conduction path. The highest battery temperature is 294.7 K, and the maximum temperature difference is about 1.5 K, which satisfies the 5 K uniformity requirement. The temperature distribution is much more uniform than that of Scheme 1. The maximum temperature occurs in the middle of the cell, while the side regions adjacent to the cooling plate are cooler. This improved uniformity is attributed to the enhanced heat transfer in the vertical direction, which is achieved by directly cooling the large side surface where the in-plane thermal conductivity is higher.
For Scheme 3 (side-mounted bidirectional counter-flow cooling), the cell temperature field is similar in trend to Scheme 2, but with even better uniformity. The highest battery temperature is 294 K, and the maximum temperature difference is approximately 1 K, which is well below the 5 K limit. The bidirectional counter-flow design ensures that the coolant temperature remains more uniform along the flow path, which further reduces the cell-to-cell and within-cell temperature variations.
Thus, under the conventional 1 C condition, all three schemes can control the maximum temperature below the recommended 313 K threshold. However, Scheme 1 fails to meet the temperature uniformity criterion due to the inadequate vertical heat transfer path. Schemes 2 and 3 both provide excellent temperature uniformity with maximum differences below 2 K, demonstrating clear advantages over bottom cooling.
Thermal Performance at High Rate (3 C)
When the charge/discharge rate is increased to 3 C, the heat generation rate scales significantly, and the differences among the three schemes become more pronounced. The simulation results are shown in Figure 7 of the original article. We summarize the key findings as follows.
In Scheme 1, the maximum cell temperature reaches 357 K, which is dangerously close to the 393 K thermal runaway threshold and far above the normal operating temperature for lithium-ion batteries. The maximum temperature difference across the pack is as high as 52 K. Under such conditions, the risk of thermal runaway is extremely high, and long-term operation is completely infeasible. The bottom cooling plate simply cannot remove heat fast enough from the upper regions of the cells, leading to severe heat accumulation at the top.
In Scheme 2, the highest cell temperature is 307 K. This value is significantly lower than that of Scheme 1, but it remains above the ideal operating window. More importantly, the temperature distribution shows a strong gradient along the coolant flow direction. The coolant temperature rises continuously from the inlet to the outlet, causing the cell region near the outlet to become much hotter. The maximum temperature difference is approximately 13 K, which fails the 5 K uniformity criterion. Although the risk of thermal runaway is lower, the large temperature difference can still accelerate cell aging, reduce cycle life, and degrade charging/discharging performance.
In Scheme 3, the highest cell temperature is 299 K, which is within the recommended operating range. The maximum temperature region appears in the middle of the cell, while the side surfaces adjacent to the cooling plate remain cooler. The maximum temperature difference is about 4.8 K, which is below the 5 K limit. Thus, Scheme 3 simultaneously satisfies both the maximum temperature and temperature uniformity requirements under the 3 C high-rate condition. The excellent performance is attributed to the counter-flow arrangement: the two coolant streams flowing in opposite directions exchange heat among themselves, preventing the coolant temperature from rising monotonically along a single path. This reverse thermal compensation mechanism minimizes the spatial temperature gradient across the cooling plate and within the cell.
Thermal Performance at Ultra-High Rate (5 C)
Grid-scale energy storage systems may occasionally be subjected to extreme conditions, such as sudden faults or severe grid disturbances, requiring ultra-high-rate charging or discharging. To evaluate the thermal management capability under such extreme conditions, we further simulate a 5 C charge/discharge rate. The results are presented in Figure 8 of the original article.
For Scheme 1, the cell and cooling plate temperatures range from 363 K to 470 K. The maximum temperature far exceeds the 393 K high-risk threshold. In this situation, the lithium-ion battery is at extreme risk of thermal runaway, possibly leading to fire or explosion, posing a severe safety hazard. The bottom cooling scheme is entirely inadequate for ultra-high-rate operations.
For Scheme 2, the maximum cell temperature reaches 332 K. Although this value is below 393 K, it is still significantly higher than the normal working temperature range. Local overheating can lead to accelerated electrode material degradation and electrolyte decomposition. The temperature difference across the pack is approximately 39 K, which causes highly nonuniform aging, shortens cycle life, and could induce irreversible performance degradation or safety concerns. The unidirectional side cooling improves heat transfer compared to bottom cooling, but the coolant heating along the flow path creates a large temperature nonuniformity at high heat generation rates.
For Scheme 3, the cell and liquid cooling plate temperatures are distributed in the range from 294 K to 308 K. This range is still within the acceptable operating window for energy storage batteries. The maximum temperature is 308 K, well below the high-risk threshold. The overall temperature difference is 14 K. While this exceeds the 5 K limit, it is important to note that ultra-high-rate charge/discharge scenarios are uncommon and typically last for a short duration. Under such short-term extreme conditions, the impact on battery life is limited, and the safety of the system is preserved since the maximum temperature remains controlled. In contrast, the other two schemes would experience either dangerous thermal runaway or severe degradation.
To provide a comprehensive comparison, Table 2 and Table 3 summarize the maximum temperature and maximum temperature difference for the three schemes under the three C-rates, respectively.
| Scheme | 1 C | 3 C | 5 C |
|---|---|---|---|
| Scheme 1 (bottom unidirectional) | 301.0 | 357.0 | 470.0 |
| Scheme 2 (side unidirectional) | 294.7 | 307.0 | 332.0 |
| Scheme 3 (side bidirectional counter-flow) | 294.0 | 299.0 | 308.0 |
| Scheme | 1 C | 3 C | 5 C |
|---|---|---|---|
| Scheme 1 (bottom unidirectional) | 6.5 | 52.0 | 107.0 |
| Scheme 2 (side unidirectional) | 1.5 | 13.0 | 39.0 |
| Scheme 3 (side bidirectional counter-flow) | 1.0 | 4.8 | 14.0 |
These results clearly demonstrate the superiority of the bidirectional counter-flow heat exchange plate for the thermal management of energy storage batteries under both high-rate and ultra-high-rate conditions. Whereas Scheme 1 is only acceptable at 1 C and Scheme 2 fails the temperature uniformity requirement at 3 C and 5 C, Scheme 3 maintains the maximum temperature below 310 K and provides reasonable temperature uniformity across all evaluated conditions.
Heat Transfer Analysis of the Cooling Plates
To gain deeper insight into the mechanisms behind the observed performance differences, we analyze the temperature distribution of the coolant inside the heat exchange plates for the three schemes under the 3 C high-rate condition. The results are shown in Figure 10 of the original article.
For Scheme 1, the coolant flows from left to right through the eight straight channels of the bottom plate. The coolant temperature rises substantially along the flow direction, with a temperature increase of about 15 K. The plate temperature distribution is characterized by a central region that is hotter than the edges. This is due to the nonuniform flow distribution and the limited vertical heat transfer from the center of the battery pack toward the edges. As the coolant warms up, its cooling capacity decreases, which further exacerbates the temperature nonuniformity inside the battery pack.
For Scheme 2, the coolant also exhibits a temperature increase along the flow direction, but the total temperature rise is reduced to about 10 K. Because the plate is mounted on the side surface of the cell, the direction perpendicular to the flow corresponds to the cell height direction, which has a relatively high in-plane thermal conductivity. This enables more effective heat spreading, resulting in a milder temperature gradient across the width of the plate. Nonetheless, the cumulative heating of the coolant along its path still leads to a significant temperature difference between the inlet side and the outlet side of the cell.
For Scheme 3, the bidirectional counter-flow design dramatically reduces the coolant temperature variation across the plate. The maximum temperature difference within the cooling plate is only about 1 K. This remarkable uniformity is achieved because the two streams of coolant flow in opposite directions, so that the local hot spots created by one stream are partially compensated by the cooler flow of the other stream at the same axial position. In other words, the plate performs a continuous internal heat exchange between the two streams, equalizing the temperature along the entire length. Moreover, the air insulation layer between adjacent plates prevents unwanted heat exchange between neighboring cells, preserving the local temperature uniformity.
The unique flow design of the bidirectional counter-flow plate can be further explained by considering the heat transfer enhancement factor. For a conventional unidirectional flow plate, the local temperature difference between the cell and the coolant decreases along the flow direction as the coolant warms up, leading to reduced heat transfer effectiveness downstream. In the bidirectional plate, the two counter-flow streams create a more uniform driving temperature difference, as expressed by the log mean temperature difference (LMTD) method commonly used for heat exchangers. For a counter-flow arrangement, the LMTD is defined as:
$$ \Delta T_{lm} = \frac{\Delta T_1 – \Delta T_2}{\ln \left( \Delta T_1 / \Delta T_2 \right) } $$
where \(\Delta T_1\) and \(\Delta T_2\) are the temperature differences between the hot and cold streams at the two ends of the heat exchanger. In a counter-flow configuration, \(\Delta T_1\) and \(\Delta T_2\) are less disparate than in a parallel-flow configuration, resulting in a higher LMTD for the same inlet and outlet temperatures. This higher effective temperature difference enhances the total heat transfer rate and simultaneously reduces the spatial temperature variation within the plate.
Additionally, the placement of the inlets at the bottom and outlets at the top exploits the natural tendency of heated coolant to rise, potentially reducing the pumping power required; however, in the present study, forced convection is dominant and the effect of natural convection is negligible.
Comparative Analysis and Engineering Implications
The results presented above clearly highlight the inadequacy of conventional bottom cooling for energy storage battery packs, especially at elevated C-rates. The fundamental bottleneck is the poor through-plane thermal conductivity of the cell, which severely limits the vertical heat conduction. As a result, heat generated near the top of the cell cannot be effectively removed by a cooling plate attached at the bottom. This leads to an unacceptably high top-to-bottom temperature gradient, which is only acceptable at very low C-rates.
Side-mounted cooling, as in Scheme 2, offers a substantial improvement because it exploits the higher in-plane thermal conductivity of the cell. However, the unidirectional flow pattern causes the coolant to heat up progressively along its path, which creates a temperature gradient perpendicular to the flow direction. At high C-rates, this gradient reach values of 13 K or more, which is detrimental to battery life and safety. Some researchers have attempted to mitigate this problem by using variable flow rates or multiple inlets, but such approaches increase the complexity of the cooling system.
The bidirectional counter-flow plate proposed in this work addresses the root cause by creating a symmetric flow arrangement that actively balances the coolant temperature distribution. The two opposing streams act as distributed heat sources and sinks for each other, effectively providing an internal thermal compensation mechanism. This simple yet powerful design reduces the maximum temperature difference in the plate to about 1 K at high C-rates and enables the battery pack to operate at ultra-high rates with a maximum temperature of only 308 K. The concept is analogous to the counter-flow heat exchanger principle, which is well established in thermal engineering but has not been widely exploited in battery thermal management plates.
From an engineering perspective, the bidirectional counter-flow heat exchange plate can be manufactured with conventional processes such as aluminum extrusion or brazed plate heat exchangers. The dual-inlet and dual-outlet configuration requires additional piping, but the overall system complexity is modest. The air insulation layer can be incorporated as a hollow chamber or filled with a low-thermal-conductivity material to minimize heat transfer between neighboring plates. The design is scalable to larger battery packs and can be adapted to different cell form factors, including prismatic and pouch cells.
To further quantify the advantages of the proposed scheme, we define a performance indicator, the cooling efficiency ratio, as the ratio of the heat removed to the pumping power. Although pumping power is not directly reported here, the pressure drop and flow rate can be used to estimate the pumping power. For a given total flow rate of 12 L/min, the pressure drop for the bidirectional counter-flow plate is comparable to that of the unidirectional side plate, since the total flow area is divided into two parallel paths. The additional pressure drop due to the counter-flow arrangement is minimal. Therefore, the enhanced thermal performance does not come at the expense of significantly higher pumping power.
Table 4 summarizes the qualitative ranking of the three schemes for various evaluation criteria under high-rate operation.
| Criterion | Scheme 1 (bottom unidirectional) | Scheme 2 (side unidirectional) | Scheme 3 (side bidirectional counter-flow) |
|---|---|---|---|
| Max temperature control | Poor | Moderate | Excellent |
| Temperature uniformity | Poor | Moderate | Excellent |
| Scalability | Good | Good | Excellent |
| Manufacturing complexity | Low | Low | Moderate |
| Pumping power | Low | Low | Moderate |
| Safety at ultra-high rates | Unacceptable | Risky | Safe |
The enhanced thermal management performance of the bidirectional counter-flow plate has important implications for the operation of grid-scale energy storage systems. When renewable generation fluctuates rapidly, the energy storage battery may be required to absorb or release large amounts of power within a short time window. With the proposed cooling scheme, the battery pack can safely operate at up to 5 C for short durations while maintaining the maximum temperature below 308 K. This capability improves the flexibility and reliability of the grid, enables more aggressive use of storage assets, and extends the calendar and cycle life of the batteries.
From a safety perspective, the prevention of thermal runaway is the ultimate goal of any battery thermal management system. The simulation results indicate that Scheme 1 would lead to cell temperatures exceeding 393 K at 5 C, which is a direct thermal runaway trigger. Scheme 2 remains below the threshold but still exposes the cells to temperatures near 332 K, which accelerates aging and could eventually lead to internal short circuits or electrolyte decomposition. Scheme 3 maintains the cells below 308 K, providing a large safety margin and significantly reducing the risk of thermal runaway.
It is also worth discussing the transient behavior of the cooling system under rapidly varying load conditions. Although the current study adopts a steady-state analysis, the thermal mass of the battery and the cooling plate affects the temperature response during transients. The bidirectional counter-flow design, with its superior spatial uniformity, ensures that even under a rapidly changing heat generation rate, the temperature gradients within the cell remain moderate. The air insulation layer further reduces the heat exchange between adjacent cells, preventing any single cell from being excessively heated by its neighbors. Future transient simulations and experimental validations will provide more insights into dynamic thermal behavior.
Conclusions and Recommendations
This study investigates the thermal management performance of three liquid cooling configurations for an energy storage battery pack under various charge/discharge rates. Based on the numerical simulation results, the following main conclusions can be drawn:
- The conventional bottom-mounted liquid cooling plate (Scheme 1) fails to meet the temperature uniformity requirement (ΔT≤5 K) at all tested C-rates. At 1 C, the maximum temperature difference reaches 6.5 K; at 3 C and 5 C, the maximum temperature differences are 52 K and over 100 K, respectively, with maximum temperatures reaching 357 K and 470 K. This configuration poses a severe thermal runaway risk at high and ultra-high rates.
- The side-mounted unidirectional cooling plate (Scheme 2) significantly improves the maximum temperature control, achieving 294.7 K at 1 C, 307 K at 3 C, and 332 K at 5 C. However, the temperature uniformity degrades sharply as the C-rate increases, with maximum temperature differences of 1.5 K, 13 K, and 39 K at 1 C, 3 C, and 5 C, respectively. The unidirectional flow path leads to monotonic coolant heating and unacceptable gradients at high rates.
- The side-mounted bidirectional counter-flow heat exchange plate (Scheme 3) outperforms the other two schemes across all operating conditions. It achieves maximum temperatures of 294 K at 1 C, 299 K at 3 C, and 308 K at 5 C, while keeping the maximum temperature difference at 1.0 K, 4.8 K, and 14 K, respectively. At the high rate of 3 C, it simultaneously satisfies the maximum temperature limit below 313 K and the temperature uniformity limit below 5 K. Even at the ultra-high rate of 5 C, the maximum temperature remains safely below the thermal runaway threshold, making the battery pack suitable for short-term extreme operation.
- The excellent thermal performance of the bidirectional counter-flow plate is attributed to its symmetrical flow channel design, which creates a reverse thermal compensation mechanism. The two counter-current coolant streams equalize the temperature distribution along the plate, while the integrated air insulation layer minimizes heat interference between adjacent cells. The effective log mean temperature difference in the counter-flow configuration is higher than that in a unidirectional flow configuration, enhancing heat transfer efficiency and uniformity simultaneously.
Based on these findings, the proposed bidirectional counter-flow heat exchange plate offers a reliable and robust technical solution for the thermal management of grid-scale energy storage batteries, particularly under high-rate and ultra-high-rate operating conditions. Future work should focus on experimental validation, optimization of channel dimensions and flow rates, and investigation of transient behaviors under real grid frequency regulation profiles.
List of Symbols and Abbreviations
| Symbol | Description | Unit |
|---|---|---|
| C | Capacity | Ah |
| \(C_p\) | Specific heat capacity | J/(kg·K) |
| \(E_0\) | Open-circuit voltage | V |
| \(I\) | Current | A |
| \(k\) | Thermal conductivity | W/(m·K) |
| \(p\) | Pressure | Pa |
| \(q\) | Volumetric heat generation rate | W/m³ |
| \(Q\) | Total heat generation rate | W/m³ |
| \(R\) | Internal resistance | Ω |
| \(T\) | Temperature | K |
| \(t\) | Time | s |
| \(U\) | Terminal voltage | V |
| \(\mathbf{v}\) | Velocity vector | m/s |
| \(V_b\) | Battery cell volume | m³ |
| \(\lambda_x,\lambda_y,\lambda_z\) | Thermal conductivities in orthogonal directions | W/(m·K) |
| \(\mu\) | Dynamic viscosity | Pa·s |
| \(\rho\) | Density | kg/m³ |
| \(\Delta T_{lm}\) | Log mean temperature difference | K |
References
[1] Zhang Z, Zhang N, Li J, et al. MXene-based flexible composite phase change material for battery thermal management application. Applied Thermal Engineering, 2025, 259: 124898.
[2] Spotnitz R, Franklin J. Abuse behavior of high-power, lithium-ion cells. Journal of Power Sources, 2003, 113(1): 81-100.
[3] Kvasha A, Gutiérrez C, Osa U, et al. A comparative study of thermal runaway of commercial lithium ion cells. Energy, 2018, 159: 547-557.
[4] Jaguemont J, Boulon L, Dubé Y. A comprehensive review of lithium-ion batteries used in hybrid and electric vehicles at cold temperatures. Applied Energy, 2016, 164: 99-114.
[5] Zeyu C, Rui X, Fengchun S. Research status and analysis for battery safety accidents in electric vehicles. Journal of Mechanical Engineering, 2019, 55(24): 93.
[6] Liu Y T, Gu Q F, Sun X, et al. Cost analysis of large-scale multi-type electrochemical energy storage. Energy Storage Science and Technology, 2023, 12(1): 312-318.
[7] Cai T A, Shen X Y, He C M, et al. Thermal behavior research and air-cooling optimization design of power lithium-ion batteries. Power Supply Technology, 2023, 47(2): 187-192.
[8] Chen K, Song M, Wei W, et al. Structure optimization of parallel air-cooled battery thermal management system with U-type flow for cooling efficiency improvement. Energy, 2018, 145: 603-613.
[9] Liu Z, Liu X, Meng H, et al. Numerical analysis of the thermal performance of a liquid cooling battery module based on the gradient ratio flow velocity and gradient increment tube diameter. International Journal of Heat and Mass Transfer, 2021, 175: 121338.
[10] Wang J, Huang Q, Li X, et al. Experimental and numerical simulation investigation on the battery thermal management performance using silicone coupled with phase change material. Journal of Energy Storage, 2021, 40(490): 102810.
[11] Li Y F, Xu W P, Wei Y T, et al. Thermal design and simulation analysis of immersion liquid cooling system for energy storage lithium battery pack. Energy Storage Science and Technology, 2024, 13(10): 3534-3544.
[12] Jing Y F, Yan N J, Li L. Design and simulation analysis of liquid cooling structure for lithium-ion energy storage battery. Science & Technology and Innovation, 2025, 113(2): 107-109.
[13] Li M, Wang J, Guo Q, et al. Numerical analysis of cooling plates with different structures for electric vehicle battery thermal management systems. Journal of Energy Engineering, 2020, 146(4): 04020037.
[14] Monika K, Chakraborty C, Roy S, et al. An improved mini-channel based liquid cooling strategy of prismatic LiFePO4 batteries for electric or hybrid vehicles. Journal of Energy Storage, 2021, 35: 102301.
[15] Wang L Y, Wu C, Yin W Q, et al. Optimal design and analysis of liquid cooling system for energy storage batteries applicable to multiple working conditions. Power Supply Technology, 2024, 48(12): 2434-2443.
[16] Sun Y. Design and numerical simulation of thermal management cold plates for blade lithium-ion batteries. Master’s thesis, Jiangsu University of Science and Technology, 2023.