Rural EV Charging and Solar Power Integration Optimized in New Study
A groundbreaking study published in Northeast Electric Power Technology has introduced a comprehensive planning framework for electric vehicle (EV) charging stations and distributed photovoltaic (PV) systems in rural power grids. The research, led by ZHANG Hui from Shenyang Institute of Engineering and State Grid Chengde Power Supply Company, in collaboration with WANG Cunxu and WANG Liang, offers a strategic solution to one of the most pressing challenges in China’s clean energy transition: how to expand EV infrastructure into rural areas while maintaining grid stability and economic efficiency.
As China pushes forward with its ambitious “dual carbon” goals—aiming for peak carbon emissions by 2030 and carbon neutrality by 2060—the electrification of transportation has become a national priority. While urban centers have seen rapid deployment of EV charging networks, rural regions have lagged behind. This disparity is not merely a matter of convenience; it reflects deeper technical, economic, and infrastructural challenges. Rural power grids, typically designed for low-density, low-load agricultural and residential use, are ill-equipped to handle the sudden surge in demand that widespread EV adoption could bring. Moreover, the integration of renewable energy sources like rooftop solar adds another layer of complexity due to their intermittent nature.
The study directly addresses these challenges by proposing a novel, two-tiered optimization model that simultaneously considers economic viability and grid security. At its core, the model is built on a deep understanding of rural user behavior, a factor often overlooked in conventional urban-centric planning approaches. By incorporating travel chain theory—a method that maps the sequence and timing of daily trips—the researchers were able to simulate realistic EV charging patterns in rural settings. Unlike city dwellers who often commute during morning and evening rush hours, rural residents tend to travel later in the morning and earlier in the afternoon, primarily for purposes such as visiting family, attending markets, or engaging in agricultural activities. This distinct travel behavior significantly influences when and where charging demand occurs.
To capture the inherent randomness and variability in both EV usage and solar generation, the team employed Monte Carlo simulation and Latin hypercube sampling. These advanced statistical techniques allowed them to generate a wide range of possible scenarios, ensuring that the planning model is robust against uncertainty. For solar power, the researchers used k-means clustering to identify five typical daily output patterns, each with its own probability of occurrence. This approach enables planners to account for different weather conditions and seasonal variations without overcomplicating the model.
The heart of the study is its multi-objective bilevel programming model. The upper level of the model focuses on minimizing the total annual societal cost, which includes not only the direct expenses of building and maintaining charging stations but also the indirect costs borne by users and the power grid. These indirect costs are often ignored in traditional planning but are critical for a holistic assessment. For instance, the time users spend traveling to and waiting at charging stations represents a real economic burden, especially in sparsely populated areas where distances are greater. Similarly, the strain that EV charging places on the grid—manifested as increased power losses and voltage fluctuations—has financial implications that must be internalized in the planning process.
The lower level of the model shifts the focus from economics to technical performance. It aims to minimize active power losses and voltage deviations across the distribution network. This dual objective ensures that the grid remains stable and efficient even under high EV penetration and fluctuating solar output. The inclusion of voltage stability as a key criterion is particularly important for rural grids, which often operate near the lower limits of acceptable voltage levels. Sudden load increases from EV charging can push these systems out of compliance, leading to equipment damage and poor power quality for consumers.
To solve this complex optimization problem, the researchers developed a hybrid algorithm that combines the strengths of Voronoi diagrams and an improved particle swarm optimization (PSO) technique. Voronoi diagrams, a geometric tool that partitions space into regions based on proximity to a set of points, are used to define the service areas of each charging station. This ensures that every demand point is assigned to the nearest available station, creating a natural and efficient spatial distribution. Meanwhile, the improved PSO algorithm handles the global search for the optimal solution. By dynamically adjusting the inertia weight and acceleration coefficients, the algorithm avoids getting trapped in local optima—a common pitfall in traditional PSO methods—and converges more quickly and reliably to the best possible configuration.
The practical application of this model was demonstrated through a case study in a newly developed rural district covering approximately 60 square kilometers. The area, served by a modified 33-node distribution system, features 49 road network nodes and a mix of residential, agricultural, and commercial land use. After running extensive simulations, the model determined that the optimal number of charging stations for this region is five. Fewer stations would lead to excessive user travel time and congestion, while more stations would increase construction and operational costs beyond the point of diminishing returns.
The total annual societal cost under this optimal configuration was calculated to be 5.29 million yuan. This figure breaks down into several components: 7.4 million yuan for construction and maintenance (amortized over 20 years), 3.67 million yuan for user travel time, 1.05 million yuan for voltage deviation penalties, and 3.88 million yuan for grid power losses. While some of these individual costs may seem high, the key insight is that their sum is minimized, reflecting a balanced trade-off between different stakeholders’ interests. For example, adding a sixth station might reduce user waiting time, but the additional construction cost and grid losses would outweigh the benefits, leading to a higher total cost.
The spatial layout of the five stations reveals a clear pattern of service area optimization. The largest service zone, colored red in the study’s figures, covers 22 demand points and requires 18 charging units, reflecting the high concentration of activity in that part of the district. In contrast, the smallest zone, colored blue, serves only two demand points and requires just seven charging units. This granular, data-driven approach ensures that resources are allocated where they are most needed, avoiding the inefficiencies of a one-size-fits-all deployment strategy.
Perhaps the most significant contribution of the study lies in its treatment of grid stability. The researchers found that without proper reactive power optimization, the integration of EVs and solar panels could lead to severe voltage fluctuations and substantial power losses. In their simulation, the unoptimized system experienced a maximum active power loss of 0.386 megawatts and a total daily loss of 4.471 megawatt-hours. After applying the proposed optimization algorithm, these figures were reduced to 0.271 megawatts and 2.849 megawatt-hours, respectively—a reduction of nearly 37 percent. This improvement not only saves energy but also extends the life of grid equipment and improves the quality of service for all customers.
Voltage stability showed similar gains. Before optimization, the voltage at node 18 dipped below 0.93 per unit, falling outside the acceptable range of 0.93 to 1.05 per unit. After optimization, the voltage was raised to 0.95 per unit, eliminating the risk of under-voltage conditions. The overall voltage deviation across the network was reduced by more than 40 percent, from 36.638 per unit-hours to 21.739 per unit-hours. These results demonstrate that with the right planning and control strategies, rural grids can accommodate high levels of distributed energy resources without compromising reliability.
The success of the improved PSO algorithm was also evident in its convergence behavior. Unlike the standard PSO, which quickly stagnated in a suboptimal solution, the enhanced version continued to refine its search throughout the iteration process. This ability to escape local optima and explore the solution space more thoroughly is crucial for solving real-world problems with many variables and constraints. The combination with Voronoi diagrams further enhanced the algorithm’s performance by providing a structured way to explore different spatial configurations.
This research has important implications for policymakers, utility companies, and EV manufacturers. For governments, it provides a scientifically sound methodology for allocating public funds to rural charging infrastructure. Instead of relying on arbitrary quotas or political considerations, planners can now use data-driven models to identify the most cost-effective locations and capacities for new stations. For utilities, the study offers a roadmap for upgrading rural grids to handle future loads, including not only EVs but also other forms of distributed generation. And for automakers, it signals that the EV market is expanding beyond cities, creating new opportunities in underserved regions.
The study also highlights the importance of interdisciplinary collaboration in addressing complex energy challenges. The team brought together expertise in power systems engineering, transportation modeling, and computational optimization to create a holistic solution. This integrative approach is increasingly necessary as the boundaries between different sectors—electricity, transportation, and information technology—blur in the era of smart grids and connected vehicles.
Looking ahead, the researchers suggest several avenues for future work. One is to incorporate demand response mechanisms, allowing EV charging to be shifted to times when solar generation is high and grid load is low. Another is to consider battery storage systems at charging stations, which could further smooth out power flows and provide backup power during outages. Additionally, the model could be extended to include other types of distributed energy resources, such as wind turbines or biomass generators, making it applicable to a wider range of rural environments.
In conclusion, the study by ZHANG Hui, WANG Cunxu, and WANG Liang represents a significant step forward in the planning of rural EV infrastructure. By combining behavioral modeling, stochastic simulation, and advanced optimization techniques, they have developed a practical and effective tool for balancing economic and technical objectives. Their work not only advances the academic literature but also provides actionable insights for practitioners working to build a more sustainable and equitable energy future. As China continues its journey toward carbon neutrality, studies like this will play a vital role in ensuring that no community is left behind in the clean energy revolution.
ZHANG Hui, WANG Cunxu, WANG Liang, Northeast Electric Power Technology, DOI: 10.19759/j.nep.2024.07.002