Smart PV-DC Charging System Achieves 100% Load Satisfaction in Beijing Office Trial
In a significant advancement for sustainable urban energy integration, a new intelligent charging strategy developed by researchers from State Grid Beijing Electric Power Research Institute and Tsinghua University has demonstrated the potential to fully power electric vehicle (EV) fleets using only rooftop solar energy. The study, conducted at an office building in Beijing, showcases a groundbreaking approach to aligning solar photovoltaic (PV) generation with EV charging demand, eliminating reliance on the external power grid during daylight hours and significantly enhancing renewable energy utilization.
As global cities accelerate their transition toward carbon neutrality, the convergence of building-integrated renewables and transportation electrification presents both immense opportunities and complex technical challenges. Solar power, while abundant and clean, is inherently intermittent and often peaks during midday—when electricity demand in commercial buildings may be moderate but EV charging demand is typically low. Conversely, EV charging demand often spikes in the evening, when solar generation has ceased, placing additional strain on the grid and increasing reliance on fossil-fuel-based power sources. This temporal mismatch has long been a critical barrier to maximizing the environmental benefits of solar-powered EV charging.
The research team, led by Yifeng Ding, Shuang Zeng, Baoqun Zhang, Liyong Wang, Chang Liu from State Grid Beijing Electric Power Research Institute, and Zhi Fu, Ji Zhang from Tsinghua University’s School of Architecture, has addressed this challenge through an innovative PV-direct current (PV-DC) intelligent charging system. Their work, published in Electric Power, introduces an orderly charging strategy that dynamically adjusts EV charging power in real time based on available solar generation and vehicle battery status, achieving unprecedented levels of self-consumption and grid independence.
The core innovation lies in the system’s use of DC bus voltage as a real-time signal for energy balance. Unlike conventional AC charging systems that require multiple power conversions and often operate on fixed charging schedules, this DC-coupled architecture enables a more efficient and responsive energy flow. The system integrates rooftop PV panels, bidirectional AC/DC converters for grid interaction, building loads, and smart DC charging stations—all connected to a shared 375-volt DC bus. By leveraging the DC bus voltage as a control parameter, the system creates a decentralized yet coordinated energy management framework that requires minimal communication infrastructure.
The control logic is elegantly simple yet highly effective. When solar generation is abundant, the DC bus voltage rises, signaling the charging stations to increase their output power. Conversely, when solar output declines due to cloud cover or the setting sun, the bus voltage drops, prompting the chargers to reduce their power draw. This voltage-based control eliminates the need for complex forecasting algorithms or centralized control systems, making the solution both robust and scalable.
What sets this strategy apart is its integration of vehicle-specific parameters into the charging decision process. The system continuously monitors each EV’s state of charge (SoC) and maximum allowable charging power. Vehicles with lower SoC are given priority during periods of limited solar availability, ensuring that those most in need of energy receive preferential treatment. This prioritization is achieved through a dynamic power allocation algorithm that modulates the charging curve based on both bus voltage and SoC, creating a fair and efficient distribution of available solar energy across multiple vehicles.
The experimental setup was deployed at a commercial office building in Beijing, featuring a 20-kilowatt-peak (kWp) rooftop PV system and two 6.6-kilowatt smart charging stations. The building’s occupants, primarily office workers, were invited to plug in their EVs during the workday, with the understanding that charging would be free and powered entirely by solar energy. This “plug-and-leave” model is particularly well-suited to workplace charging scenarios, where vehicles remain parked for extended periods, providing ample time for solar energy to accumulate and charge the batteries.
Over the course of the study, the system was tested under a variety of weather conditions, from clear, sunny days to overcast, low-irradiance scenarios. On a typical sunny day, the PV system generated 71.8 kilowatt-hours (kWh) of electricity, with a peak output of 13.7 kW. During this period, two EVs were charged, receiving 30.6 kWh and 13.9 kWh respectively. Crucially, the system achieved a 100% load satisfaction rate, meaning that all charging energy came directly from the on-site solar array, with zero reliance on the external grid. The excess solar energy—amounting to 63% of total generation—was exported to the grid, demonstrating the system’s ability to not only meet local demand but also contribute surplus clean energy to the broader power network.
Even under less favorable conditions, the system performed remarkably well. On a cloudy day with only 22.1 kWh of total solar generation, the intelligent charging strategy still managed to meet 100% of the EVs’ charging needs without drawing from the grid. The photovoltaic self-consumption rate reached 86%, and when accounting for the small amount of energy used by the control systems themselves, the effective utilization of solar power climbed to 94%. This resilience under variable weather conditions underscores the robustness of the voltage-based control strategy, which can adapt seamlessly to rapid changes in solar output.
To quantify the advantages of their approach, the researchers conducted a comparative analysis using Monte Carlo simulations. They modeled the same charging scenario using a traditional constant-power charging method, where EVs draw power at their maximum rated rate (6.6 kW) regardless of solar availability. In this conventional setup, the charging demand created a significant morning peak, forcing the system to import up to 10.9 kW from the grid to meet the instantaneous power requirements. Meanwhile, during the solar noon peak, when generation was highest, the EVs had already partially charged, leading to a large amount of curtailed or exported solar energy. The result was a photovoltaic self-consumption rate of only 43.2% and a maximum grid export of 11.5 kW.
In stark contrast, the intelligent charging strategy achieved a self-consumption rate of 61.3%, representing a 42% improvement over the traditional method. More importantly, the maximum power exported to the grid was reduced to just 5.3 kW, a 54% decrease. This dramatic reduction in peak export is a critical benefit for grid operators, as it mitigates the risk of overvoltage and instability in distribution networks that are not designed to handle large, unpredictable injections of distributed solar power. By smoothing both the import and export profiles, the system enhances grid stability and reduces the need for costly infrastructure upgrades.
The implications of this research extend far beyond a single office building in Beijing. As urban areas worldwide grapple with the dual challenges of decarbonizing transportation and integrating high levels of renewable energy, this PV-DC charging strategy offers a practical and scalable solution. The technology is particularly well-suited for commercial and industrial facilities, where large rooftop spaces are available for solar installation and employee parking lots provide a natural aggregation point for EVs. Shopping malls, logistics centers, and public institutions could all adopt similar systems to create self-sustaining energy ecosystems that reduce operating costs and carbon footprints.
Moreover, the success of this project highlights the importance of system-level design in the energy transition. Rather than treating solar panels, EVs, and the grid as separate entities, the researchers have demonstrated the value of an integrated approach that optimizes the interactions between these components. The DC architecture itself is a key enabler, as it reduces conversion losses and simplifies control. In a conventional AC system, solar DC power must be inverted to AC to feed the grid or power AC loads, and then converted back to DC to charge an EV battery—a process that can incur losses of 10% or more. By keeping the energy in DC form from generation to consumption, the system achieves higher overall efficiency.
The user experience is another critical factor in the success of such systems. The “set it and forget it” nature of the charging process—where drivers simply plug in their vehicles and receive a full charge by the end of the workday—removes the behavioral barriers that often hinder the adoption of smart charging technologies. There is no need for users to schedule charging sessions or respond to price signals; the system operates autonomously in the background, optimizing energy use without requiring any action from the driver. This seamless integration into daily routines is essential for widespread adoption.
The research also opens up new possibilities for future development. The team notes that while the current system performs exceptionally well, there is room for further optimization. For instance, incorporating time-of-day parameters into the control algorithm could help to spread charging more evenly throughout the day, preventing situations where vehicles are fully charged by midday, leaving excess afternoon solar energy to be exported. Additionally, the system could be expanded to include battery storage, allowing surplus solar energy to be stored for use during evening hours or on days with low generation. This would further increase self-consumption and grid independence, moving closer to the goal of a truly net-zero energy building.
Another promising avenue is the integration of vehicle-to-building (V2B) capabilities. While the current system focuses on solar-to-vehicle (S2V) energy flow, future iterations could allow EVs to discharge back to the building during peak demand periods, effectively turning parked vehicles into mobile energy storage units. This bidirectional capability would enhance the building’s resilience and provide additional revenue streams through participation in demand response programs.
The economic benefits of such a system are also significant. By reducing or eliminating electricity purchases for EV charging, building owners can achieve substantial cost savings. At the same time, the revenue from selling excess solar energy to the grid can further improve the financial viability of the investment. As the cost of solar panels and power electronics continues to decline, the payback period for these systems is expected to shorten, making them increasingly attractive to businesses and institutions.
From a policy perspective, this research provides valuable evidence to support the development of regulations and incentives that promote integrated energy solutions. Governments and utilities can use these findings to design programs that encourage the adoption of smart charging infrastructure, particularly in commercial and public sectors. Standards for DC microgrids and building-integrated energy systems may need to be updated to accommodate these new technologies, ensuring interoperability and safety.
The success of this project also underscores the importance of collaboration between utilities, academic institutions, and industry. The partnership between State Grid Beijing Electric Power Research Institute and Tsinghua University exemplifies how practical engineering expertise and academic research can come together to solve real-world problems. Such collaborations are essential for accelerating the deployment of innovative technologies that are critical to achieving global climate goals.
In conclusion, the work of Ding, Zeng, Zhang, Wang, Liu, Fu, and Zhang represents a major step forward in the integration of renewable energy and electric transportation. Their PV-DC intelligent charging strategy has proven capable of meeting 100% of EV charging demand using only on-site solar power, while significantly improving photovoltaic self-consumption and reducing grid impact. The results, published in Electric Power (DOI: 10.11930/j.issn.1004-9649.202305101), provide a compelling blueprint for the future of sustainable urban energy systems. As cities around the world seek to reduce their carbon emissions and build more resilient infrastructure, this research offers a practical, scalable, and highly effective solution that brings the vision of a clean energy future one step closer to reality.