Optimized EV Charging Station Placement Enhances Grid Voltage Stability
As electric vehicle (EV) adoption accelerates globally, urban infrastructure planners and power system engineers face mounting pressure to integrate charging networks seamlessly into existing electricity grids. The rapid proliferation of EVs introduces new challenges, particularly in maintaining stable voltage levels within distribution networks. Uncoordinated charging behavior—especially during peak hours—can exacerbate load imbalances, strain grid components, and degrade power quality. Traditional approaches to charging station planning have largely focused on profitability, accessibility, or traffic flow optimization, often overlooking the critical interplay between transportation patterns and electrical grid performance.
Now, a groundbreaking study by Ou Jiancong and Yang Lijie from Guangxi Communications Design Group Co., Ltd. presents a novel strategy that bridges the gap between power systems and transportation networks. Published in Mechanical & Electrical Engineering Technology, their research introduces a voltage-quality-driven framework for optimal charging station siting, marking a significant shift from conventional revenue-centric models to a more holistic, system-aware approach.
The core innovation lies in the integration of EV charging behavior with both traffic mobility patterns and power grid dynamics. Rather than treating EVs as isolated loads, the model recognizes that charging demand is inherently tied to human movement—where people go, when they arrive, how long they stay, and where they choose to charge. By fusing these transportation attributes with electrical constraints, the researchers have developed a planning methodology that not only meets user demand but also actively improves grid performance.
At the heart of this strategy is the objective of maximizing average line voltage across the distribution feeder. This focus on voltage quality is particularly relevant in medium-voltage distribution systems, such as the 10 kV feeders common in Chinese urban power networks. Voltage deviations, especially drops below acceptable thresholds, can lead to equipment malfunction, reduced efficiency, and even blackouts. With EVs drawing substantial power—particularly during evening charging surges that coincide with residential peak loads—the risk of voltage instability grows significantly.
Previous studies have explored various aspects of EV infrastructure planning. Some have used queuing theory to estimate wait times at charging stations, while others have applied multi-objective optimization to balance cost, user convenience, and network expansion. Several have incorporated traffic constraints or simulated annealing algorithms to improve station placement. However, many of these models remain siloed, either focusing narrowly on the transportation side or treating the power grid as a passive backdrop.
Ou and Yang’s work stands out by explicitly modeling the coupling between the two systems. They begin by analyzing the charging characteristics of slow-charging stations—the most prevalent type in urban and residential areas. While individual slow chargers operate at modest power levels (typically 7 kW), their widespread deployment and collective impact make them a dominant factor in local grid loading. Unlike fast-charging stations, which are often located on the urban periphery and serve transient users, slow chargers are deeply embedded in daily life, primarily serving overnight charging needs in neighborhoods, parking lots, and workplaces.
To capture realistic charging behavior, the model incorporates probabilistic factors such as daily driving distance, battery capacity, and charging initiation time. Daily mileage is modeled using a log-normal distribution, reflecting the variability in commuter patterns and personal travel. Charging start times are derived from statistical distributions of departure from work, return home, and parking duration—key determinants of when and where an EV owner decides to plug in.
Crucially, the model introduces a spatial willingness function that quantifies the likelihood of an EV owner choosing a particular charging station based on proximity. This reflects real-world behavior: drivers tend to prefer stations that are geographically close and easily accessible via existing road networks. By mapping vehicle origins and potential charging sites onto a coordinate plane and calculating travel distances along actual roads, the model simulates a more accurate representation of charging demand distribution.
This spatial-temporal charging load profile is then integrated into a power system model that includes standard electrical constraints: power flow equations, current limits, voltage bounds, and equipment capacity. The optimization problem seeks to determine the number and location of charging stations that maximize the average voltage across all nodes in the feeder, while ensuring that no operational limits are violated.
To solve this complex, nonlinear optimization problem, the researchers employ an enhanced particle swarm optimization (PSO) algorithm. PSO is a population-based stochastic technique inspired by the social behavior of birds flocking or fish schooling. In this context, each “particle” represents a potential solution—a specific configuration of charging station locations and quantities. The algorithm iteratively updates these particles based on their own best-known position and the global best solution found by the swarm, guiding the search toward the optimal configuration.
The improvement over standard PSO lies in adaptive mechanisms that enhance convergence speed and avoid premature trapping in local optima. These modifications allow the algorithm to efficiently navigate the high-dimensional solution space, balancing exploration and exploitation to find robust, high-quality solutions.
To validate their approach, Ou and Yang applied it to a real-world-inspired 8-node 10 kV distribution feeder within the China Southern Power Grid system. The test network includes typical residential load profiles, with peak demand occurring in the early evening—a period that overlaps heavily with anticipated EV charging activity. The total EV load was set at 70 kW, reflecting a moderate but non-negligible penetration level.
Two scenarios were compared: one using the proposed voltage-optimized placement strategy, and another representing conventional planning practices where stations are distributed more uniformly across available nodes (nodes 3 through 7). In the optimized case, the solution concentrated stations at nodes 2, 4, and 6, with 4, 3, and 3 units respectively.
The results were striking. Under the proposed method, the average feeder voltage reached 10.12 kV, compared to 10.01 kV under the traditional approach. This 0.11 kV improvement may seem small in absolute terms, but in the context of distribution system operation, it represents a meaningful enhancement in power quality and system headroom. It translates to better voltage regulation, reduced losses, and increased resilience against further load growth.
More importantly, this gain was achieved without compromising service coverage or user accessibility. The optimized layout still met the full charging demand, demonstrating that improved grid performance does not require sacrificing user convenience. In fact, by aligning station placement with natural traffic patterns and electrical sensitivity, the solution enhances both system efficiency and user experience.
The implications of this research extend beyond technical optimization. It offers a new paradigm for infrastructure planning—one that recognizes the interdependence of energy and transportation systems in the age of electrified mobility. As cities worldwide strive to decarbonize their transport sectors, they must also ensure that the electricity grid can support this transformation reliably and sustainably.
Urban planners, utility operators, and policy makers can leverage this methodology to make more informed decisions about where to invest in charging infrastructure. Instead of reactive deployment driven by demand spikes or political considerations, a coordinated, data-driven approach enables proactive grid management. This is especially valuable in areas with aging infrastructure or limited upgrade budgets, where maximizing the value of each new installation is critical.
Moreover, the model’s emphasis on voltage quality aligns with broader smart grid objectives. Maintaining stable voltage levels supports the integration of other distributed energy resources, such as rooftop solar and battery storage. It also reduces stress on transformers and cables, extending equipment lifespan and lowering maintenance costs. In this way, optimized EV charging becomes not just a necessity, but an opportunity to strengthen the overall grid.
The study also highlights the importance of interdisciplinary collaboration. Effective EV infrastructure planning requires expertise in transportation engineering, electrical systems, behavioral modeling, and computational optimization. Ou Jiancong and Yang Lijie’s background in both highway design and power system planning uniquely positions them to bridge these domains. Their work exemplifies how cross-sector knowledge can yield innovative solutions to complex urban challenges.
Looking ahead, the framework could be expanded in several directions. Future versions might incorporate dynamic pricing signals, allowing the model to respond to real-time electricity markets. It could also account for bidirectional charging (vehicle-to-grid), where EVs not only draw power but also supply it back to the grid during peak periods. Integration with renewable generation forecasts would enable coordinated scheduling that maximizes the use of clean energy.
Additionally, the model could be adapted to different urban contexts—dense metropolitan cores, suburban neighborhoods, or rural communities—each with distinct travel patterns and grid configurations. With sufficient data, machine learning techniques could further refine the probabilistic models of driver behavior, making predictions even more accurate.
Another promising avenue is the integration of equity considerations. While the current model focuses on technical performance, future iterations could include social objectives, such as ensuring fair access to charging facilities across different neighborhoods or income groups. This would support more inclusive and just energy transitions.
From a regulatory standpoint, the findings suggest that incentives for EV charging infrastructure should be tied not just to the number of stations deployed, but also to their system-wide impacts. Performance-based subsidies could reward developers who locate stations in ways that benefit grid stability, rather than simply maximizing customer traffic.
In conclusion, Ou Jiancong and Yang Lijie’s research represents a significant step forward in the intelligent integration of electric vehicles into urban power systems. By redefining the goal of charging station planning—from profit maximization to voltage optimization—they offer a more sustainable and resilient vision for the future of transportation electrification. Their method demonstrates that with the right tools and perspective, EVs can be part of the solution to grid challenges, rather than a source of new problems.
As cities continue to expand their EV charging networks, adopting such system-aware planning strategies will be essential. The transition to electric mobility is not merely about replacing internal combustion engines with batteries; it is about reimagining how energy and transportation systems interact. This study provides a practical and scalable framework for doing exactly that—ensuring that the roads of the future are not only cleaner, but also smarter and more reliable.
Ou Jiancong, Yang Lijie. Optimized EV Charging Station Placement Enhances Grid Voltage Stability. Mechanical & Electrical Engineering Technology. DOI: 10.3969/j.issn.1009-9492.2024.07.046