EV Clusters and Pumped Storage Boost Wind Integration in Rural China
In the rolling hills of Xinyang, Henan Province, a new chapter in China’s renewable energy story is quietly unfolding. As the nation pushes forward with its ambitious “thousand villages, ten thousand wind turbines” initiative, rural power grids are facing unprecedented challenges. The rapid integration of distributed wind power, while a triumph for clean energy, has strained the capacity of local distribution networks, particularly in remote areas where infrastructure has not kept pace with generation capacity. Voltage fluctuations, equipment overloads, and curtailment of wind power have become common issues, threatening both grid stability and economic efficiency.
Yet, amid these challenges, a promising solution is emerging—one that leverages the growing fleet of electric vehicles (EVs) not just as transportation tools, but as dynamic, responsive assets in the energy system. A recent study published in Shandong Electric Power by Jiang Jian, Zhang Shusen, and Xu Fengliang from State Grid Xinyang Power Supply Company presents a novel collaborative dispatch strategy that integrates EV clusters, small-scale pumped storage, and distributed wind power into a cohesive, responsive network. The research, titled Research on the Collaborative Dispatch Strategy of Distributed Resources Considering the Response of Electric Vehicle Cluster, offers a practical framework for enhancing wind power utilization while improving the economic and operational performance of rural grids.
The core of the study lies in recognizing the untapped potential of EVs as flexible loads. Traditionally, EV charging has been viewed as a burden on the grid, especially when users charge their vehicles during peak hours, exacerbating the peak-to-valley load difference. However, the authors argue that with proper incentives and intelligent control, EVs can become a powerful tool for demand response—shifting their charging patterns to absorb excess wind power during off-peak hours and reducing load when supply is tight.
But not all EV users are equally willing to participate in such programs. The willingness to respond to grid signals depends on a variety of factors, including the state of charge (SOC) of the battery and the level of financial incentive offered. To capture this complexity, the researchers developed a sophisticated model based on the Takagi-Sugeno-Kang (TSK) fuzzy logic system. This approach allows for a more nuanced understanding of user behavior, moving beyond simplistic assumptions of uniform responsiveness.
The TSK model quantifies user willingness by considering two key variables: the current SOC of the EV battery and the incentive price offered by the utility. For instance, an EV with a low SOC is more likely to accept a charging schedule, especially if the incentive is attractive. Conversely, a vehicle with a high SOC may be less inclined to respond unless the compensation is substantial. By defining fuzzy sets for “high,” “medium,” and “low” levels of SOC and incentive price, the model generates a response willingness index that reflects the probability of user participation.
What makes this model particularly effective is its ability to handle uncertainty. Instead of treating user behavior as deterministic, the researchers incorporate triangular membership functions to represent the fuzziness of human decision-making. This allows the system to account for the fact that willingness is not a fixed value but a range of possibilities. The result is a more realistic assessment of how much charging power can be flexibly adjusted at any given time.
With this quantified willingness, the researchers then define the schedulable range of EV charging power. This range is not static; it evolves throughout the day based on vehicle availability, travel patterns, and grid conditions. For example, during nighttime hours when most vehicles are parked and batteries are low, the available flexibility is high. In contrast, during morning and evening commutes, when vehicles are in use, the ability to shift charging is limited.
To test their model, the team conducted a case study on a 10 kV rural distribution network in Xinyang. The system includes three EV clusters, each with 100 vehicles, connected at different nodes, along with two distributed wind farms totaling 8 MW and a 4 MW small-scale pumped storage facility. The simulation covers a full 24-hour period, capturing the dynamic interplay between wind generation, load demand, and EV charging behavior.
The results are compelling. When the system operates without considering EV response willingness—essentially assuming all users will comply with dispatch instructions—the model overestimates the available flexibility. This leads to suboptimal scheduling and higher costs, as the system attempts to utilize charging capacity that may not actually be available. In contrast, when a fixed willingness value of 0.5 is assumed (a common simplification in many studies), the model underestimates the true potential of EVs, particularly during midday hours when wind output is high and vehicle availability is favorable.
The dynamic willingness model, however, strikes the right balance. It shows that during the early morning hours (1:00–5:00), when wind generation exceeds local demand, EV users are less willing to respond due to already high charging activity. As a result, the system relies more heavily on the pumped storage facility to absorb excess power by pumping water uphill. But from 9:00 to 15:00, when wind output remains strong but EV usage drops, user willingness peaks, allowing the system to significantly increase charging and fully utilize available wind power.
This synergy between EVs and pumped storage is a key finding of the study. While pumped storage provides fast, reliable power regulation, EVs offer a distributed, scalable form of demand response. Together, they create a “wind-storage-load” coordinated operation mode that enhances grid flexibility and reduces reliance on costly curtailment.
The economic benefits are equally significant. The study compares three scenarios: no consideration of willingness, static willingness, and dynamic willingness. In the first scenario, the total cost—including wind curtailment penalties and incentive payments to EV users—is highest, as the system overestimates available flexibility and incurs unnecessary expenses. The static willingness model performs better, but still falls short of optimal efficiency. The dynamic model, by accurately capturing user behavior, achieves the lowest total cost, reducing incentive payments by nearly 23% compared to the static case.
Moreover, the integration of pumped storage further amplifies these savings. In both peak-shaving and valley-filling conditions, the presence of pumped storage allows the system to shift more of the balancing burden away from EVs, thereby reducing the need for high incentive payments. In the reverse peak scenario—where wind generation is high at night and low during the day—the combined system reduces EV incentive costs by over 39%. In the normal peak scenario—where wind and load trends align—the savings are slightly lower but still substantial, at around 33%.
These findings have important implications for the future of rural electrification in China and beyond. As distributed wind and solar continue to expand, the challenge of grid integration will only grow. Traditional solutions, such as building new transmission lines or relying on fossil-fueled peaking plants, are expensive and environmentally unsustainable. The approach outlined in this study offers a more elegant, cost-effective alternative—one that turns the challenge of variability into an opportunity for innovation.
The success of this strategy depends on several enabling factors. First, it requires accurate data on EV usage patterns, battery states, and user preferences. This data can be collected through smart charging systems and vehicle telematics, which are becoming increasingly common in modern EVs. Second, it relies on effective communication between the grid operator and EV owners, ensuring that incentives are transparent and timely. Third, it demands advanced optimization algorithms capable of handling the complexity of multi-resource coordination.
The researchers used MATLAB’s fmincon function with an interior-point method to solve the optimization problem, demonstrating the feasibility of implementing such models in real-world applications. However, they also acknowledge that further work is needed to scale the approach to larger systems and to incorporate additional variables, such as weather forecasts, traffic conditions, and market price signals.
One of the most notable aspects of this research is its grounding in real-world conditions. Unlike many academic studies that rely on idealized assumptions, this work is based on actual grid topology, load profiles, and wind generation data from a rural area in Henan. This practical orientation increases the credibility and applicability of the findings, making them relevant not only to Chinese utilities but also to power systems in other developing regions facing similar challenges.
The policy implications are clear. To unlock the full potential of EVs as grid resources, governments and utilities must move beyond passive charging infrastructure and embrace active demand management. This includes developing incentive programs that reflect real user behavior, investing in smart grid technologies, and fostering public awareness of the benefits of flexible charging.
In Xinyang, where mountains meet rivers and wind meets water, the future of energy is being shaped by innovation and collaboration. The integration of wind, storage, and electric vehicles is not just a technical achievement—it is a vision of a smarter, more resilient, and more sustainable power system. As the world transitions to a low-carbon future, the lessons from this rural Chinese grid may offer valuable insights for communities everywhere.
The study underscores a fundamental shift in how we think about energy. No longer are consumers simply passive recipients of power; they are active participants in a dynamic, interconnected system. Electric vehicles, once seen as a challenge to grid stability, are now emerging as a key enabler of renewable energy integration. By understanding and respecting user behavior, utilities can design more effective, equitable, and efficient energy systems.
As Jiang Jian, Zhang Shusen, and Xu Fengliang demonstrate, the path to a cleaner energy future lies not in building bigger power plants, but in making smarter use of the resources we already have. In the quiet villages of Henan, a revolution is charging forward—one electric vehicle at a time.
Jiang Jian, Zhang Shusen, Xu Fengliang, State Grid Xinyang Power Supply Company, Shandong Electric Power, DOI: 10.20097/j.cnki.issn1007-9904.2024.06.001