The global push for decarbonization has placed the electrification of transportation, particularly battery electric cars, at the forefront of sustainable development. A critical, yet often underappreciated, subsystem within a battery electric car is its thermal management system. This system is responsible not only for cabin comfort but, more importantly, for maintaining optimal temperatures for the high-voltage battery and power electronics, directly impacting vehicle range, performance, and longevity. As regulations phase out high-Global Warming Potential (GWP) refrigerants, natural alternatives like R290 (propane) are gaining significant traction for use in battery electric cars due to their excellent thermodynamic properties and negligible GWP. Modern architectures for battery electric cars often employ secondary loop or indirect systems, where a refrigerant loop cools a coolant (e.g., water-glycol), which then circulates to cool the battery, power electronics, and the cabin via separate heat exchangers. This enhances safety, especially with flammable refrigerants like R290, and allows for greater system modularity.
The performance and reliability of these vapor-compression circuits in a battery electric car are heavily influenced by the presence and management of lubricating oil. The compressor requires oil for lubrication, sealing, and cooling. However, a fraction of this oil circulates with the refrigerant throughout the system, a parameter quantified as the Oil Circulation Rate (OCR). An optimal OCR is a delicate balance: sufficient oil must return to the compressor to prevent mechanical failure, but excessive oil circulating in heat exchangers (condenser, evaporator) can foul surfaces, degrade heat transfer, and ultimately reduce system capacity and efficiency—critical metrics for the energy-conscious operation of a battery electric car. Determining the correct system oil charge and its corresponding OCR is therefore paramount. This task is further complicated in highly integrated, compact module designs for battery electric cars, where traditional methods for measuring OCR, such as inline oil flow meters or sampling techniques, are often impractical due to space constraints, lack of access ports, or very low refrigerant charges. This study addresses this gap by proposing and validating a novel calibration method for determining the optimal oil charge and inferring the OCR in an integrated R290 secondary loop thermal management module designed for a battery electric car.
We designed a comprehensive experimental methodology involving two complementary test benches. The first is a standard compressor calorimeter test stand, which allows for precise control and measurement of the OCR. The second is a full secondary loop system test bench replicating the architecture of a battery electric car’s thermal management system, incorporating an integrated R290 module. The core of our method involves calibrating the system oil charge by comparing a key indicator—compressor discharge temperature—between the two benches under identical compressor operating states (speed, suction/discharge pressure). By adding varying amounts of oil to the system bench and matching its compressor discharge temperature to that measured on the compressor bench at a target OCR (e.g., 3%), we can effectively deduce the system oil charge that yields the desired OCR. This approach circumvents the need for direct OCR measurement on the integrated module.
Our investigation systematically examines the influence of system oil charge on four critical aspects: 1) overall system cooling performance (capacity and COP), 2) compressor discharge temperature (used for OCR calibration), 3) compressor residual oil quantity (a key reliability indicator), and 4) the effective refrigerant charge capacity of the system. The findings provide crucial guidelines for the design and calibration of efficient and reliable R290 thermal systems for the next generation of battery electric cars.
Test Rigs and Methodologies
Compressor Unit Performance Test Bench for OCR Control
To establish a baseline and target for OCR, a dedicated compressor calorimeter test bench was utilized. Figure 1 illustrates its schematic. The core component is the R290 swing compressor with enhanced vapor injection (EVI). High-pressure discharge gas from the compressor first passes through an efficient oil separator. The separated lubricant flow is precisely measured by a mass flow meter (M2), heated or cooled to the target compressor suction temperature in a temperature-controlled bath, and then returned to the compressor suction line. The separated refrigerant vapor is measured by a separate gas mass flow meter (M1).
The refrigerant stream is then split. One branch passes through an electronic expansion valve (EEV2) for pressure reduction, while the other flows through a plate heat exchanger acting as a condenser/desuperheater, with its cooling load controlled by EEV3. The two streams are mixed in a reservoir to achieve the precise target suction temperature and pressure for the compressor, monitored by sensor PT1. A data acquisition system records all parameters. The OCR in this setup is directly calculated and controlled:
$$OCR_{comp} = \frac{\dot{m}_{oil}}{\dot{m}_{ref} + \dot{m}_{oil}} \times 100\%$$
where $\dot{m}_{oil}$ is the oil mass flow rate measured by M2 and $\dot{m}_{ref}$ is the refrigerant vapor mass flow rate measured by M1. By setting the compressor speed, suction/discharge pressures, suction superheat, and a target OCR (e.g., 3.0% ± 0.1%), the corresponding steady-state compressor discharge temperature is recorded. This temperature serves as the reference fingerprint for that specific OCR under the defined operating condition. Repeatability tests confirmed the discharge temperature remained within ±1°C for the same OCR setting.
Secondary Loop System Test Bench
A complete secondary loop system test bench was constructed in an environmental chamber to emulate the thermal management system of a battery electric car. The schematic is shown in Figure 2. It consists of three main circuits:
- Refrigerant (R290) Circuit (Integrated Module): This is the core integrated module, featuring the EVI scroll compressor, a water-cooled condenser (LCC), a liquid receiver/drier, a plate-type chiller evaporator (Chiller), an internal heat exchanger (economizer), and two electronic expansion valves (EXV1 for main circuit, EXV2 for injection circuit). All refrigerant-side components are tightly integrated with minimal internal piping to reduce volume and potential leak points, a common design goal for compact battery electric car applications.
- High-Temperature Coolant Circuit: This circuit simulates the waste heat from the battery electric car’s powertrain (e.g., motor, inverter). A water-glycol mixture is heated by a controlled heater and pumped through the LCC, where it rejects heat to the condensing R290. The flow rate and inlet temperature are precisely controlled.
- Low-Temperature Coolant Circuit: This circuit simulates the cooling demand of the battery electric car’s cabin (via a HVAC unit) and/or the battery. Chilled water-glycol from the Chiller is pumped to a cooling load unit (a heater with precise power control). The flow rate and inlet temperature to the Chiller are controlled to set the desired cooling load and evaporating conditions.
The system is controlled to mimic real operation in a battery electric car. EXV1 modulates to maintain a target compressor suction superheat. EXV2 modulates to control the superheat at the economizer outlet (injection port). The compressor speed is set constant for a given test. The heating power on the high-temperature circuit controls the condensing pressure, while the cooling load on the low-temperature circuit controls the evaporating pressure.
Key components used in the system bench are listed in Table 1.
| Component | Key Parameters / Specifications |
|---|---|
| Electric Compressor | 34cc EVI Scroll, Speed Range: 800 – 8500 rpm |
| Water-Cooled Condenser (LCC) | Dimensions: 180mm (L) x 70mm (W) x 89mm (H) |
| Plate Evaporator (Chiller) | Dimensions: 180mm (L) x 70mm (W) x 89mm (H) |
| Electronic Expansion Valve | Orifice Diameter: 2.1 mm, Step Range: 0-576 pulses |
| Receiver/Drier | Internal Volume: 100 ml, Desiccant: 25 g |
| Coolant Pump | Rated Power: 200 W, Max Head: 18m @ 25 L/min |
System Oil Charge and OCR Calibration Method
For the integrated module on the system bench, direct OCR measurement is not feasible. Our calibration method is as follows:
- Step 1 – Target OCR Definition on Compressor Bench: On the compressor bench, set the operating condition to a representative point: Compressor speed = 5000 rpm, Discharge pressure = 1.9 MPa ± 0.003 MPa, Suction pressure = 0.47 MPa ± 0.004 MPa, Suction superheat = 9°C ± 1°C. Set the OCR to the design target (e.g., 3.0% ± 0.1%) and record the stable compressor discharge temperature (e.g., 84.0°C). This is the target discharge temperature.
- Step 2 – Oil Charge Variation on System Bench: On the secondary loop system bench, ensure the integrated module is clean and evacuated. Charge it with a specific mass of PAG oil (e.g., 50 g, 60 g, 70 g, 80 g, 90 g). Then, charge with R290 refrigerant.
- Step 3 – System Bench Operation at Matching Condition: Operate the system bench, controlling the compressor speed, LCC conditions (water flow and temperature), and Chiller conditions to achieve the exact same steady-state compressor suction pressure, discharge pressure, and suction superheat as in Step 1. This typically involves adjusting the water temperatures and flows on the high and low-temperature circuits.
- Step 4 – Discharge Temperature Comparison and OCR Inference: Once the system is stable (typically after 2 hours), record the compressor discharge temperature. Compare it to the target discharge temperature from the compressor bench.
- If the system bench discharge temperature is significantly higher than the target, it suggests insufficient oil is returning to the compressor (low OCR), leading to poorer internal sealing and cooling.
- If the system bench discharge temperature is significantly lower than the target, it suggests excessive oil is circulating (high OCR), altering the compression process and mixture properties.
- If the system bench discharge temperature is within a narrow band (e.g., ±1°C) of the target, it implies the OCR in the system is very close to the target OCR set on the compressor bench. The corresponding system oil charge is thus calibrated for that OCR.
Additional Test Procedures
System Performance Tests: To evaluate the impact of oil charge on cooling performance, tests were conducted at a high-load condition relevant for battery electric car cabin cooling: Ambient = 25°C, Compressor speed = 8500 rpm, LCC water inlet = 49°C, LCC water flow = 25 L/min, Chiller water inlet = 27.5°C, Chiller water flow = 15 L/min. Steady-state cooling capacity and COP were measured for different oil charges.
Compressor Residual Oil Test: This test ensures the compressor retains enough oil for lubrication during low-speed operation common in a battery electric car. The system is run at a low compressor speed (800 rpm) with mild cooling loads for 2 hours. The system is then shut down, the refrigerant is carefully recovered, and the compressor is detached and weighed to determine the amount of oil retained inside it. The requirement for the specific compressor used was a minimum of 50 g.
Refrigerant Charge Determination Tests: The effect of oil charge on the optimal refrigerant charge was investigated. For a fixed oil charge, the system refrigerant charge was gradually increased while monitoring performance indicators (compressor discharge pressure, suction superheat, EXV opening, LCC subcooling) under both cooling and heating mode conditions. The “charge plateau”—the range of refrigerant mass providing stable, near-optimal performance—was identified.
Data Processing and Uncertainty Analysis
Key performance parameters were calculated from measured data:
Coolant-side Heat Transfer Rate (Cooling/Heating Capacity):
$$Q = \frac{V_{coolant} \cdot \rho \cdot c \cdot (t_{in} – t_{out})}{60}$$
where $Q$ is the heat transfer rate (W), $V_{coolant}$ is the volumetric flow rate (L/min), $\rho$ is the coolant density (kg/m³), $c$ is the coolant specific heat (J/(kg·K)), and $t_{in}$, $t_{out}$ are the inlet and outlet temperatures (°C) of the heat exchanger (Chiller for cooling, LCC for heating).
System Coefficient of Performance (COP):
$$COP = \frac{Q_c}{W_{comp}}$$
where $Q_c$ is the Chiller cooling capacity (W) and $W_{comp}$ is the compressor electrical input power (W).
Compressor Isentropic Efficiency:
$$\eta_{cs} = \frac{h_{dis, is} – h_{suc}}{h_{dis, act} – h_{suc}}$$
where $h_{dis, is}$ is the specific enthalpy at the discharge pressure and the suction entropy (isentropic compression), $h_{suc}$ is the specific enthalpy at the suction state, and $h_{dis, act}$ is the specific enthalpy at the actual discharge state.
Compressor Volumetric Efficiency:
$$\eta_{cv} = \frac{\dot{m}_{ref, act}}{\rho_{suc} \cdot V_{disp} \cdot N / 60} = \frac{Q_h / (h_{dis, act} – h_{lcc, out})}{\rho_{suc} \cdot V_{disp} \cdot N / 60}$$
where $\dot{m}_{ref, act}$ is the actual refrigerant mass flow rate, estimated from the LCC heat rejection $Q_h$ and the enthalpy difference across it, $\rho_{suc}$ is the suction gas density (kg/m³), $V_{disp}$ is the compressor displacement (m³), $N$ is the compressor speed (rpm), and $h_{lcc, out}$ is the enthalpy at the LCC outlet.
The uncertainty of the measurements was analyzed based on the accuracy of the sensors. The specifications of key instruments are listed in Table 2.
| Instrument | Range & Accuracy |
|---|---|
| Temperature Sensor (PT100) | -50 to 200°C, ±0.15°C |
| Water Flow Meter | 2 to 50 L/min, ±0.5% of reading |
| Pressure Transducer | 1 to 30 bar, ±1% of full scale |
| High Voltage Power Analyzer | 1000V, 50A, ±0.1% of reading for power |
| Precision Mass Scale (oil) | ±0.5 g |
Using the root-sum-square method for error propagation, the maximum uncertainties for system cooling capacity and COP were calculated to be 1.11% and 1.48%, respectively, which are acceptable for this comparative study.
Results and Discussion
Impact of Oil Charge on System Cooling Performance
The cooling performance of the R290 secondary loop system, a crucial subsystem for battery thermal management and cabin comfort in a battery electric car, was significantly affected by the total oil charge. The results are summarized in Figure 4. Both the system cooling capacity and COP exhibited a non-monotonic relationship with oil charge, forming a distinct peak.
At a low oil charge of 50 g, the system showed the poorest performance. This is attributed to insufficient lubrication and sealing within the compressor, leading to increased internal leakage (blow-by), which reduces the effective mass flow rate of refrigerant. This is corroborated by the lower volumetric efficiency at this point. The increased friction and internal losses also manifest as a higher compressor discharge temperature (as will be shown later). The refrigerant-oil mixture in the evaporator (Chiller) may have had a slightly higher enthalpy difference, but the severely reduced mass flow rate resulted in the lowest net cooling capacity.
As the oil charge increased to 70 g, performance peaked. The compressor volumetric efficiency reached its maximum, indicating optimal sealing with minimal internal leakage. Sufficient oil return ensured good lubrication, reducing mechanical losses. While the presence of oil in the evaporator slightly reduces the specific enthalpy difference of the refrigerant (due to the oil’s thermal properties and potential fouling), this negative effect was outweighed by the positive effect of maximizing refrigerant mass flow through improved compressor efficiency. Consequently, both cooling capacity and COP were optimized.
Further increasing the oil charge to 80 g and 90 g led to a decline in performance. The volumetric efficiency decreased again. With excessive oil in circulation, the increased viscosity of the refrigerant-oil mixture raises the pressure drop throughout the system, effectively increasing the compressor pressure ratio for the same heat exchanger conditions. This increases compressor work. More critically, excessive oil coating the internal surfaces of the condenser (LCC) and evaporator (Chiller) acts as a thermal insulation layer, significantly degrading heat transfer efficiency. This degradation forces the compressor to work harder to meet the same cooling load, lowering the COP. The reduction in cooling capacity at 90 g oil charge, despite a potentially higher mass flow, confirms that the heat transfer penalty becomes dominant.
Table 3 quantifies the performance metrics at different oil charges under the high-load cooling test.
| Oil Charge (g) | Cooling Capacity, Qc (W) | COP | Comp. Vol. Eff., ηcv | Comp. Isen. Eff., ηcs |
|---|---|---|---|---|
| 50 | 10,920 | 1.58 | 0.895 | 0.548 |
| 60 | 11,280 | 1.67 | 0.921 | 0.562 |
| 70 | 11,421 | 1.71 | 0.932 | 0.566 |
| 80 | 11,305 | 1.69 | 0.925 | 0.560 |
| 90 | 11,193 | 1.65 | 0.918 | 0.555 |
This performance trend underscores the critical need to optimize oil charge in the thermal system of a battery electric car to maximize range and efficiency.
Calibration of Oil Charge via Compressor Discharge Temperature
The results of the OCR calibration method are presented in Figure 6. On the compressor bench, with all operating parameters fixed and the OCR controlled at 3.0%, the steady-state discharge temperature was 84.0°C (the target).
On the system bench, as the oil charge was increased from 50 g to 90 g, the measured compressor discharge temperature under identical compressor operating states showed a clear decreasing trend. This aligns with the known effects of oil: it provides internal cooling for the compressor and alters the specific heat of the gas mixture being compressed.
The key finding is that at system oil charges of 70 g and 80 g, the discharge temperatures were 84.5°C and 83.5°C, respectively. Both values are within the ±1.0°C band of the target 84.0°C. According to our calibration logic, this indicates that when the integrated module is charged with 70-80 g of oil, the inherent OCR during steady-state operation is approximately 3%. At 50 g and 60 g oil charge, the discharge temperature was much higher (e.g., ~96°C at 50g), signaling a lower OCR and potential oil starvation at the compressor. At 90 g, the temperature was lower (~93°C), indicating a higher OCR.
This method successfully provides a practical way to determine the system oil charge required to achieve a target OCR in an integrated module where direct measurement is impossible. For the design of this specific R290 system for a battery electric car, an oil charge of 70 g is selected from this range, as it also corresponds to the peak performance identified earlier.
Impact of Oil Charge on Compressor Residual Oil
The residual oil test is vital for ensuring the long-term reliability of the compressor in a battery electric car, which experiences diverse operating profiles including prolonged low-speed running. The results are shown in Figure 7.
As expected, the amount of oil retained in the compressor after a low-speed run increased with the total system oil charge. With a 50 g system charge, the residual oil was only about 45 g, which is below the compressor manufacturer’s minimum requirement of 50 g. This poses a risk of inadequate lubrication during extended low-load operation. At 60 g system charge, the residual oil was borderline at approximately 52 g. For system oil charges of 70 g and above, the residual oil quantity was sufficient, measuring 65 g, 75 g, and 85 g for 70g, 80g, and 90g charges respectively.
This result provides the lower bound for the acceptable oil charge. While a 60 g charge might seem acceptable from a residual oil perspective, it corresponds to a lower OCR and poorer cooling performance (as seen in Table 3). Therefore, combining the residual oil requirement with the performance and OCR calibration data strongly points to 70 g as the optimal charge, ensuring both reliability and efficiency for the battery electric car thermal system.
Interaction Between Oil Charge and Refrigerant Charge
In the compact integrated module of a battery electric car, the internal volume is fixed. Adding more oil (a liquid largely retained in the compressor sump and receiver) displaces volume that would otherwise be occupied by refrigerant. This interaction was experimentally characterized.
For a system with a 90 g oil charge, the refrigerant charge characteristic test revealed a performance plateau starting at 170 g of R290 and ending at 245 g. For the optimal system with a 70 g oil charge, the plateau started at 185 g and ended at 245 g. This demonstrates that reducing the oil charge by 20 g allowed the system to reach its optimal performance with 15 g less refrigerant. The end of the plateau (overcharge condition) remained the same, as it is governed by the total internal volume of the system.
The relationship can be approximated as a linear displacement for the onset of optimal charge:
$$\Delta m_{ref, plateau-start} \approx -0.75 \cdot \Delta m_{oil}$$
This has practical implications for the total greenhouse gas footprint (via refrigerant charge) and cost. Minimizing the refrigerant charge of R290, a flammable fluid, is also a safety priority in battery electric cars. Therefore, the optimal 70 g oil charge not only yields the best performance and OCR but also enables a lower minimum refrigerant charge, aligning with multiple design objectives for sustainable and safe battery electric cars.
Conclusion
This study presents a comprehensive experimental investigation into determining the optimal lubricant charge and oil circulation rate for a highly integrated R290 secondary loop thermal management system designed for battery electric cars. The major findings and contributions are summarized as follows:
- Novel Calibration Method: A practical method was developed and validated to determine the system oil charge that yields a target Oil Circulation Rate (OCR) in integrated modules where direct OCR measurement is infeasible. The method correlates the compressor discharge temperature of the system under test with that of the same compressor on a calibrated test bench operating at a known, controlled OCR under identical compressor state conditions.
- Optimal Oil Charge: For the specific R290 integrated module studied, an oil charge of 70 grams was identified as optimal. This charge resulted in:
- Peak System Performance: Maximum cooling capacity (11,421 W) and COP (1.71) under high-load conditions.
- Target OCR: A compressor discharge temperature indicating an OCR very close to the design target of 3%, ensuring proper compressor lubrication and minimizing heat exchanger fouling.
- Adequate Residual Oil: More than 50 g of oil retained in the compressor during low-speed operation, meeting reliability requirements.
- Reduced Refrigerant Charge: Enabled a lower optimal R290 charge, benefiting safety and environmental impact.
- Performance Trade-offs: Both insufficient and excessive oil charge degrade the performance of the battery electric car thermal system. Low oil charge increases internal compressor leakage and raises discharge temperatures, while high oil charge increases system pressure drops, degrades heat exchanger effectiveness, and lowers compressor volumetric efficiency.
- System Interaction: The oil charge directly impacts the required refrigerant charge. A reduction in oil charge allows for a corresponding reduction in the minimum refrigerant needed to reach optimal performance, which is a critical design consideration for managing the total flammable fluid inventory in a battery electric car.
The methodologies and findings of this work provide essential guidelines for engineers developing efficient, reliable, and safe R290-based thermal management systems for next-generation battery electric cars. The calibration approach solves a key practical challenge in characterizing integrated modules, paving the way for the wider adoption of low-GWP refrigerant technology in the automotive industry.