CO₂ Heat Pump Systems in EVs: New Insights on Charge Optimization
As electric vehicles (EVs) continue to gain traction across global markets, one persistent challenge remains at the forefront of automotive engineering: how to maintain cabin comfort and battery efficiency in extreme weather without sacrificing driving range. The problem is especially acute in cold climates, where heating demands can significantly drain battery power, leading to what consumers commonly refer to as “range anxiety.” In response, automakers and researchers have increasingly turned to advanced thermal management systems, with trans-critical CO₂ heat pump technology emerging as a leading solution. Unlike traditional refrigerants such as R134a—phased out due to high global warming potential (GWP)—CO₂ offers an environmentally friendly alternative with excellent thermodynamic performance, particularly in low-temperature environments.
Recent research conducted by Wang Congfei from the Chinese Association of Refrigeration, in collaboration with Jia Fan, Yin Xiang, and Cao Feng from Xi’an Jiaotong University, has shed new light on the critical role of refrigerant charge volume in the performance and stability of CO₂-based thermal systems in electric vehicles. Published in a recent issue of a leading engineering journal, their study explores how varying operating conditions affect the optimal refrigerant charge and reveals potential instability risks when systems operate outside ideal charge ranges.
The work addresses a key gap in current EV thermal system design. While CO₂ heat pumps are known for their efficiency and environmental benefits, their behavior under real-world driving conditions—characterized by fluctuating ambient temperatures, variable airflow, and dynamic load demands—has not been fully understood. This complexity is compounded by the fact that CO₂ operates in a transcritical cycle, where the gas cooler (functioning similarly to a condenser) does not involve phase change in the same way as subcritical systems. As a result, system performance becomes highly sensitive to refrigerant charge levels, which must be carefully managed to avoid inefficiencies or even system failure.
At the heart of the study is a comprehensive simulation model developed using CT-Suite, a platform widely used for dynamic thermal system analysis. The researchers constructed a detailed model of a full-scale vehicular CO₂ heat pump system, incorporating components such as a compressor, dual indoor heat exchangers, an outdoor heat exchanger, a liquid receiver with internal heat exchange functionality, a battery chiller, electronic expansion valves (EXVs), and a four-way reversing valve to switch between cooling and heating modes. This architecture allows the system to handle cabin climate control, defrosting, defogging, and battery thermal regulation—functions essential for year-round EV usability.
To ensure the accuracy of their simulations, the team validated the model against experimental data obtained from a physical test bench configured identically to the simulated system. Performance metrics such as coefficient of performance (COP) and heat load were compared across multiple operating conditions. The results showed that simulation outputs deviated by less than ±10% from actual measurements, confirming the model’s reliability for predictive analysis. This level of fidelity is crucial when exploring nuanced phenomena like refrigerant migration and charge sensitivity, where small errors can lead to misleading conclusions.
One of the primary findings of the research centers on how external environmental factors influence the minimum refrigerant charge required for stable operation. The team systematically analyzed the effects of three key variables: ambient temperature, indoor air volume, and outdoor wind speed. These parameters were chosen because they directly reflect real-world driving scenarios—such as city driving with frequent stops (low outdoor wind speed), highway cruising (high wind speed), or adjusting cabin airflow for passenger comfort.
In heat pump mode—used for cabin heating—the study found that the minimum required refrigerant charge increases with rising ambient temperature. Specifically, when outdoor temperatures climbed from -10°C to 5°C, the optimal charge increased by 18.6%. This may seem counterintuitive at first glance, as one might expect less heating demand in milder cold. However, the explanation lies in the thermodynamic behavior of CO₂ under transcritical conditions. As ambient temperature rises, the optimal discharge pressure also increases to maintain efficient heat rejection in the gas cooler. Higher pressure leads to greater refrigerant density in the low-pressure side of the system, particularly in the evaporator (the outdoor heat exchanger in heating mode), thereby increasing the total mass of refrigerant needed to sustain stable operation.
Similarly, outdoor wind speed was found to have a pronounced effect. When wind speed increased from 1.0 m/s to 6.0 m/s at 0°C, the required charge rose by 18.9%. Enhanced airflow improves heat transfer in the outdoor unit, allowing more effective evaporation and higher system mass flow rate. However, this also means more refrigerant accumulates in the evaporator, necessitating a higher overall charge to prevent underfilling and subsequent performance degradation. This finding has direct implications for vehicle design, especially for models intended for high-speed operation or regions with consistently strong winds.
Indoor air volume, meanwhile, showed a more moderate but still significant impact. In heat pump mode, reducing indoor airflow led to a higher minimum charge requirement, with a 6.16% increase observed when airflow dropped from 6.0 kg/min to 1.0 kg/min at 5°C. Lower airflow reduces heat absorption in the indoor heat exchanger (now functioning as the condenser), which in turn affects the balance of refrigerant distribution between high- and low-pressure sides. The system compensates by requiring more refrigerant in the low-pressure loop to maintain adequate suction conditions for the compressor.
In contrast, the behavior in cooling mode follows a different pattern. Here, the minimum charge requirement decreases as ambient temperature rises. For instance, when temperature increased from 25°C to 40°C, the required charge dropped by 7.03%. This is because higher ambient temperatures elevate the optimal discharge pressure, increasing the temperature at the gas cooler outlet. At the same time, the evaporating pressure (on the indoor side) tends to decrease, reducing the density of refrigerant in the low-pressure section. With less refrigerant mass needed in the evaporator, the total system charge can be reduced without risking instability.
Indoor airflow in cooling mode also plays a critical role. Increasing airflow enhances cooling capacity but reduces the amount of refrigerant required in the evaporator due to faster heat exchange and lower refrigerant hold-up. The study recorded a 7.85% reduction in minimum charge when indoor airflow was raised from 1.5 kg/min to 6.0 kg/min at 35°C. This inverse relationship underscores the importance of adaptive control strategies that can adjust refrigerant management based on real-time cabin conditions.
Outdoor wind speed in cooling mode had the smallest relative impact among the three variables, with a 2.27% increase in minimum charge when wind speed rose from 1.5 m/s to 6.0 m/s at 25°C. Nevertheless, the trend remains upward: improved airflow enhances heat rejection in the gas cooler, lowering discharge temperature and increasing refrigerant density in the high-pressure side. This results in slightly higher refrigerant retention in the outdoor unit, thus demanding a marginally larger total charge.
These findings collectively highlight a fundamental challenge in EV thermal system design: there is no single “optimal” charge level that works across all conditions. Instead, the ideal charge is dynamic, shifting in response to environmental and operational variables. This poses a significant challenge for system engineers, who must design systems that can accommodate these fluctuations without compromising efficiency or safety.
The issue becomes even more complex when considering non-optimal charge scenarios—situations where the actual refrigerant amount deviates from the ideal range. The researchers discovered that undercharged or overcharged conditions can lead to system instability, particularly during transient operations such as startup or mode switching. In one simulated case, an overcharged system in heat pump mode led to liquid refrigerant being drawn into the compressor suction line—a condition known as “slugging” that can cause mechanical damage and drastically reduce efficiency.
The simulation revealed a dangerous feedback loop: as liquid entered the compressor, discharge temperature dropped sharply, causing the cabin air temperature to fall below the setpoint. In response, the control system increased compressor speed in an attempt to raise output, while the electronic expansion valve opened further to reduce discharge pressure. However, these actions only worsened the problem by increasing mass flow and exacerbating liquid carryover. The system eventually reached a state of control instability, with the compressor running at maximum speed (8,000 rpm) and the EXV fully open, yet failing to stabilize.
This scenario illustrates a critical insight: in non-ideal charge conditions, the usual control logic—where valve position regulates pressure and compressor speed regulates temperature—can break down. The underlying cause is a shift in the relationship between control inputs and system outputs due to abnormal refrigerant distribution. When too much liquid accumulates in the wrong part of the circuit, traditional feedback mechanisms no longer produce the expected responses.
To address this, the researchers proposed a novel mitigation strategy: dynamically adjusting the target discharge pressure based on system state. In the simulation, they intentionally raised the pressure setpoint from 8.6 MPa to 9.5 MPa, which caused the EXV to close and reduced refrigerant flow into the evaporator. This allowed the system to gradually clear the excess liquid from the suction line. Once stable operation was restored, the pressure setpoint could be safely lowered back to normal levels. This approach effectively decouples control from charge sensitivity by temporarily altering the system’s operating point to correct refrigerant distribution.
The implications of this research extend beyond academic interest. For automakers, it provides actionable data for designing more robust thermal management systems. One immediate application is in the sizing of the liquid receiver—the component responsible for storing excess refrigerant and ensuring proper charge distribution. By understanding how charge demand varies with ambient conditions, engineers can optimize receiver volume to minimize weight and cost while maintaining performance across a wide range of climates.
Additionally, the findings support the development of smarter control algorithms that can adapt to changing conditions in real time. Rather than relying on fixed setpoints, future systems could use sensor data—such as ambient temperature, vehicle speed, and cabin airflow—to predict optimal charge requirements and adjust control parameters accordingly. Machine learning models trained on datasets like those generated in this study could enable predictive thermal management, further enhancing energy efficiency and passenger comfort.
From a sustainability standpoint, the advancement of CO₂-based systems aligns with global efforts to reduce greenhouse gas emissions. With a GWP of just 1—compared to thousands for some synthetic refrigerants—CO₂ represents a truly low-impact solution. Moreover, its natural origin and non-flammability make it safer and more sustainable than many alternatives currently under consideration.
However, widespread adoption still faces hurdles. CO₂ systems operate at much higher pressures than conventional refrigerants, requiring stronger materials and more robust sealing technologies. Manufacturing costs remain higher, and service infrastructure is still limited. Yet, as this research demonstrates, the technical challenges are being systematically addressed through rigorous modeling and experimentation.
The study also emphasizes the importance of interdisciplinary collaboration in advancing EV technology. The team brought together expertise in refrigeration engineering, thermodynamics, and system modeling—fields that must converge to solve the complex challenges of modern mobility. Their work exemplifies the kind of rigorous, evidence-based research needed to drive innovation in the automotive sector.
Looking ahead, the next frontier may involve integrating thermal management with broader vehicle energy systems. For example, waste heat from the powertrain or battery could be used to preheat the CO₂ circuit in cold starts, reducing the load on the heat pump. Similarly, during regenerative braking, excess thermal energy could be stored or redirected to optimize system efficiency.
In conclusion, the research conducted by Wang Congfei, Jia Fan, Yin Xiang, and Cao Feng offers a significant contribution to the field of electric vehicle thermal management. By quantifying the impact of environmental and operational variables on refrigerant charge requirements and identifying strategies to mitigate instability in non-ideal conditions, their work provides a solid foundation for the next generation of high-efficiency, climate-resilient EVs. As the automotive industry continues its transition toward electrification, studies like this will play a crucial role in ensuring that comfort, efficiency, and sustainability go hand in hand.
Wang Congfei, Jia Fan, Yin Xiang, Cao Feng. Xi’an Jiaotong University and Chinese Association of Refrigeration. Published in International Journal of Refrigeration. DOI: 10.3969/j.issn.0253-4339.2024.04.059