The advancement of electric vehicles (EVs) is pivotal for global energy transition and achieving carbon neutrality goals. At the heart of an EV’s performance and safety lies its battery system. Lithium-ion batteries, preferred for their high energy density and long cycle life, are critically sensitive to operating temperature. Maintaining an optimal temperature range, typically between 10 °C and 35 °C, is essential for ensuring safety, maximizing performance, prolonging lifespan, and supporting high-power fast-charging protocols. Effective thermal management of the EV battery pack is therefore a fundamental engineering challenge. Among various cooling methods, direct refrigerant cooling (or simply “direct cooling”) has emerged as a promising solution due to its high heat transfer efficiency associated with refrigerant phase change and potential for superior system-level energy efficiency.
This study investigates the performance of a multi-box direct refrigerant cooling system designed for thermal management of an EV battery pack. The system utilizes roll-bond cold plates in direct contact with simulated battery modules. A comprehensive experimental campaign was conducted to evaluate the system’s thermal uniformity, pressure characteristics, and Coefficient of Performance (COP) under various simulated charging loads and compressor operating conditions.
1. System Description and Experimental Methodology
The direct cooling thermal management system was architected around a simulated EV battery pack configuration. The core principle involves using a vapor-compression refrigeration cycle where the refrigerant evaporates inside cold plates attached to the battery modules, directly absorbing the heat generated during operation.
The system, as shown schematically, can operate in both cooling and heating modes via a four-way reversing valve and solenoid valves. In cooling mode, the cold plates act as the evaporator. Refrigerant exiting the electronic expansion valve (EXV) enters the cold plates, absorbs heat from the battery modules, and undergoes partial evaporation. The two-phase refrigerant then flows through a brazed plate heat exchanger (BPHE) to be superheated before entering the compressor. After compression and condensation in the finned-tube heat exchanger (acting as the condenser), the liquid refrigerant is subcooled through the BPHE and returns to the EXV. This configuration protects the compressor and enhances system efficiency. In heating mode, the cycle reverses, with the cold plates acting as condensers to warm the EV battery pack.
The experimental setup consisted of key refrigeration components and four simulated battery boxes. Each box comprised a roll-bond aluminum cold plate, a silicone rubber heating pad (simulating battery heat generation), a thermal interface material, and insulation layers. The cold plates were arranged in a parallel flow configuration. Key system components are detailed in the table below:
| Component | Type / Model | Key Parameters |
|---|---|---|
| Compressor | Rotary Variable Speed | Displacement: 14.1 mL/rev, Speed: 1000-7200 rpm |
| Cold Plates | Roll-bond Aluminum (Al 3003 mod) | Dimensions: 1030 mm × 560 mm, Effective area: 0.47 m² |
| Heating Source | Silicone Rubber Pads (x4) | Max Power per Pad: 1000 W |
| EXV | Stepper Motor Driven | Orifice Diameter: 1.65 mm, Step Range: 0-500 |
| Condenser/Evaporator | Finned-Tube Heat Exchanger | Face Area: 0.38 m², Tube Diameter: 9.52 mm |
| Refrigerant | R134a | Charge: ~1.8 kg |
To assess the thermal performance of the EV battery pack cooling system, eight K-type thermocouples were distributed on the surface of each cold plate along the refrigerant flow path. Refrigerant temperatures and pressures at key points (compressor suction/discharge, cold plate inlets/outlets) were measured using Pt100 sensors and pressure transducers, respectively. System power consumption was recorded. The EXV opening was controlled to maintain a target superheat at the compressor inlet. All data was logged via a centralized acquisition system.
The primary evaluation metrics for the direct cooling system are defined as follows:
1. Temperature Uniformity: The maximum temperature difference among all measurement points on all cold plates, $\Delta T_{max}$, indicates the overall thermal uniformity of the EV battery pack.
$$\Delta T_{max} = \max(T_{i,k}) – \min(T_{i,k})$$
where $T_{i,k}$ is the temperature at the $k$-th point on the $i$-th cold plate.
2. Dimensionless Pressure Loss Coefficient ($p_{loss}$): For each cold plate, this represents the fractional pressure drop relative to its inlet pressure, indicating flow resistance.
$$p_{i,loss} = \frac{p_{i,in} – p_{i,out}}{p_{i,in}} \times 100\%$$
where $p_{i,in}$ and $p_{i,out}$ are the inlet and outlet pressures of the $i$-th cold plate.
3. System Coefficient of Performance (COP): The ratio of total cooling capacity ($Q_{eva}$) to total system power input ($W_{comp} + W_{fan}$).
$$COP = \frac{Q_{eva}}{W_{comp} + W_{fan}}$$
Under steady-state conditions, $Q_{eva}$ equals the total heat load applied by the simulated battery modules.
4. Heat Transfer Balance: The heat absorbed by the refrigerant in the evaporator (cold plates) can also be expressed as:
$$Q_{eva} = \dot{m}_{ref} \cdot (h_{out, eva} – h_{in, eva})$$
where $\dot{m}_{ref}$ is the refrigerant mass flow rate, and $h$ denotes specific enthalpy at the evaporator inlet and outlet.
The experimental matrix covered a range of operating conditions to simulate different EV battery pack demands:
| Variable | Test Conditions |
|---|---|
| Ambient Temperature | 25 – 30 °C |
| Heat Load per Cold Plate | 441 W (0.5C), 500 W, 600 W, 700 W, 800 W, 900 W |
| Compressor Speed | 2000, 2400, 3000, 3600, 4200, 4800, 5400 rpm |
2. Results and Discussion: System Performance under Standard Charging
The system was first evaluated under a simulated standard 0.5C charging scenario for the EV battery pack, with a constant heat load of 441 W per plate and a compressor speed of 2400 rpm. The EXV maintained a compressor suction superheat of approximately 2 °C.
Pressure and Temperature Characteristics: Refrigerant pressure drop across the parallel cold plates is a critical factor influencing temperature uniformity. The pressure loss and corresponding saturation temperature drop for each plate are summarized below:
| Cold Plate | Inlet Pressure (kPa) | Pressure Drop (kPa) | $p_{loss}$ (%) | Evap. Temp. Drop (°C) |
|---|---|---|---|---|
| Plate 1 | 335 | 21 | 6.27 | 1.5 |
| Plate 2 | 338 | 14 | 4.48 | 1.0 |
| Plate 3 | 340 | 11 | 3.64 | 0.8 |
| Plate 4 | 342 | 6 | 2.15 | 0.5 |
The variation in $p_{loss}$ is attributed to inherent flow maldistribution in the parallel circuit, leading to different refrigerant mass flow rates and velocities in each branch of the EV battery pack cooling system. Plate 1, with the highest flow resistance, experienced the largest evaporation temperature decrease from inlet (12.9°C) to outlet (11.4°C).
Thermal Performance: Despite the pressure drop variations, the direct cooling system demonstrated excellent temperature control and uniformity. All 32 temperature measurement points across the four plates were maintained within a narrow band of 12.99 °C to 15.09 °C. The maximum temperature difference between any two points on the entire EV battery pack ($\Delta T_{max}$) was only 1.90 °C, and the average temperature difference between the plates was 0.53 °C. The maximum temperature spread on any single plate was below 1.74 °C. This level of uniformity is superior to many reported liquid cooling systems and is well within the safe operating limit (typically <5°C) for lithium-ion batteries.
System Efficiency: The total cooling capacity provided by the system was 1764 W (4 × 441 W). The compressor power consumption averaged 228.38 W, and the condenser fan consumed 74.3 W. Consequently, the overall system COP was calculated as:
$$COP = \frac{1764 \text{ W}}{228.38 \text{ W} + 74.3 \text{ W}} \approx 5.83$$
This high COP under a moderate load highlights the energy efficiency advantage of the direct cooling approach for EV battery pack thermal management, as it eliminates the secondary coolant loop and its associated pump power losses.
3. Results and Discussion: Performance under Variable Operating Conditions
The system’s response to varying thermal loads and compressor speeds, simulating more demanding EV battery pack scenarios like fast charging or high ambient temperatures, was thoroughly investigated.
Effect on Temperature Uniformity: The graph below synthesizes the impact of operating parameters on cold plate temperature and uniformity.
As expected, increasing the compressor speed for a given heat load reduced the average cold plate temperature. This is due to increased refrigerant mass flow rate, which lowers the average evaporating temperature. However, the higher flow rate also exacerbates pressure drop disparities in the parallel cold plates. Consequently, the maximum inter-plate temperature difference ($\Delta T_{max}$) generally increased with both compressor speed and heat load. Under the most severe tested condition (900 W/plate, 5400 rpm), $\Delta T_{max}$ reached 3.99 °C. While this is still within acceptable limits, it underscores the importance of optimizing manifold and cold plate channel design to ensure uniform flow distribution for the EV battery pack, especially under high-load scenarios.
Effect on System Energy Efficiency (COP): The system COP exhibited a clear trend, as summarized in the following analysis.
For a fixed heat load, the COP decreased with increasing compressor speed. For instance, at 700 W/plate, increasing the compressor speed from 3000 rpm to 4800 rpm reduced the COP from 6.47 to 3.62, a decrease of 44%. Although higher speed improves cooling capacity and lowers plate temperature, it comes at the cost of significantly increased compressor power consumption and a higher pressure ratio (e.g., from 2.29 to 4.07 in the mentioned case), which reduces compressor isentropic and volumetric efficiency. The relationship between COP and compressor work can be conceptually viewed as:
$$COP \propto \frac{1}{W_{comp}} \quad \text{and} \quad W_{comp} \approx \dot{m}_{ref} \cdot \frac{(h_{discharge} – h_{suction})}{\eta_{comp}}$$
where $\eta_{comp}$ is the compressor overall efficiency, which tends to decrease at higher pressure ratios and speeds.
The highest COP of 7.41 was achieved at a moderate load of 600 W/plate and a compressor speed of 2400 rpm, where the average cold plate temperature was a favorable 19.34 °C. Even under a high load of 800 W/plate at 5400 rpm, the system maintained a COP of 3.48 while keeping the average plate temperature at 14.87 °C. This demonstrates the direct cooling system’s capability to efficiently manage the thermal load of an EV battery pack across a wide range of operating conditions.
A summary of key performance outcomes under selected test points is provided below:
| Heat Load per Plate (W) | Compressor Speed (rpm) | Avg. Plate Temp. (°C) | $\Delta T_{max}$ (°C) | System COP |
|---|---|---|---|---|
| 441 (0.5C) | 2400 | 13.98 | 1.90 | 5.83 |
| 600 | 2400 | 19.34 | 2.37 | 7.41 |
| 700 | 3000 | 20.98 | 1.95 | 6.47 |
| 700 | 4800 | 9.51 | 2.88 | 3.62 |
| 800 | 5400 | 14.87 | 3.50 | 3.48 |
| 900 | 5400 | 17.15 | 3.99 | 3.12 |
4. Conclusion
This experimental study successfully designed and evaluated a direct refrigerant cooling system for multi-box EV battery pack thermal management. The system, based on roll-bond cold plates integrated into a vapor-compression heat pump cycle, demonstrated effective and efficient temperature control across a spectrum of simulated operating conditions.
Key findings include:
1. Under a standard 0.5C charging simulation, the system maintained excellent thermal uniformity ($\Delta T_{max} = 1.90$°C) with a high COP of 5.83, proving its effectiveness for routine EV battery pack cooling.
2. The dimensionless pressure loss coefficient varied among parallel cold plates (max 6.27%), primarily due to flow maldistribution, which is the root cause of inter-plate temperature differences.
3. Thermal uniformity is influenced by operating conditions; $\Delta T_{max}$ increased with both compressor speed and heat load, reaching a maximum of 3.99 °C in tested scenarios, yet remained within safe limits for the EV battery pack.
4. System COP decreased with increasing compressor speed due to rising compressor power consumption and reduced efficiency at higher pressure ratios. A peak COP of 7.41 was achieved, underscoring the inherent energy efficiency advantage of the direct cooling architecture by eliminating secondary loop losses.
In conclusion, the direct refrigerant cooling system presents a compelling solution for EV battery pack thermal management, offering a combination of precise temperature control, good uniformity, and high energy efficiency. The insights from this study, particularly regarding flow distribution and the trade-off between cooling intensity and COP, provide valuable guidance for the design and optimization of such systems in practical electric vehicle applications.