Electric Vehicles’ Spatial Flexibility Boosts Grid Capacity, New Study Finds
As the global energy transition accelerates, electric vehicles (EVs) are no longer viewed merely as transportation tools but as dynamic assets capable of reshaping the way power systems operate. A groundbreaking study published in Power System Technology reveals that the spatial flexibility of EV charging—how and where drivers choose to charge—can significantly enhance the capacity of distribution networks to integrate renewable energy.
The research, led by Huang Mengqi, Li Yonghui, Yang Jun, and Wang Mengke from the Hubei Engineering and Technology Research Center for AC/DC Intelligent Distribution Network and the School of Electrical Engineering and Automation at Wuhan University, introduces a novel computational framework that accounts for the spatial decision-making behavior of EV owners. This model marks a shift from traditional assessments of grid capacity, which have largely focused on temporal demand patterns, to a more holistic approach that considers the geographic distribution of charging activities.
For years, grid planners have grappled with the challenge of integrating high levels of distributed generation—such as rooftop solar and small wind turbines—alongside the rapidly growing fleet of electric vehicles. Both sources introduce variability and uncertainty into the grid. Solar output fluctuates with weather and time of day, while EV charging demand depends on human behavior, travel patterns, and vehicle availability. When combined, these factors can lead to voltage instability, line overloads, and inefficient power flows, especially in low-voltage distribution networks not originally designed for bidirectional energy exchange.
Most existing models for assessing grid carrying capacity—the maximum amount of distributed generation and EV load a network can handle without violating operational limits—have treated EV charging as a fixed demand tied to home locations. However, this assumption fails to capture the reality of modern EV usage. With the rise of real-time charging navigation apps, drivers are increasingly guided to charging stations based on availability, pricing, and even grid conditions. This behavior transforms EV charging into a spatially flexible resource, one that can be strategically directed to support grid stability.
The team from Wuhan University recognized this shift and sought to quantify its impact. “We realized that EV owners don’t just charge wherever they park,” said Huang Mengqi, the lead author. “They make decisions based on convenience, cost, and increasingly, digital guidance from navigation systems. This creates an opportunity to align charging behavior with grid needs.”
Their model, grounded in distributionally robust optimization (DRO), is designed to handle the dual uncertainties of renewable generation and EV demand. Unlike stochastic optimization, which requires precise probability distributions, or traditional robust optimization, which can be overly conservative, DRO uses historical data to construct a confidence set of possible probability distributions. This allows the model to balance economic efficiency with operational reliability, avoiding both underutilization of grid capacity and excessive risk of constraint violations.
A key innovation in the study is the integration of spatial schedulability into the optimization framework. The researchers defined a “spatially schedulable” EV as one whose owner is willing to travel a certain distance beyond their immediate destination to charge at a different station. This willingness is influenced by factors such as the availability of charging spots, the state of the vehicle’s battery, and compensation for additional travel distance. By modeling this flexibility, the researchers were able to simulate how charging demand could be redistributed across multiple stations to alleviate local congestion and improve overall system performance.
The study was conducted using a modified IEEE 33-node distribution system, a standard test case in power systems research. The network was equipped with various grid-supporting devices, including microturbines, energy storage systems, on-load tap changers, static VAR compensators, and capacitor banks. The researchers simulated four typical daily scenarios based on historical wind and solar data, each representing different weather and load conditions.
Results showed that incorporating spatial schedulability increased the grid’s photovoltaic hosting capacity by 3.1%—from 5.12 MW to 5.28 MW—compared to a scenario where EVs charged only at the nearest station. While wind capacity slightly decreased due to the finite overall hosting capability, the total integration of renewable energy improved. More importantly, system-wide operational costs dropped significantly. The model reduced electricity purchases from the main grid, lowered network losses, and minimized curtailment of renewable generation.
One of the most striking findings was the redistribution of charging demand. Without spatial flexibility, certain charging stations—particularly those near residential areas—became overloaded during peak hours, while others remained underutilized. When spatial scheduling was enabled, the model shifted some charging activity to less congested stations, effectively smoothing the load profile. For example, during midday hours, EVs that would have charged at stations near shopping centers were redirected to nearby office parks, where demand was lower and solar generation was high. This not only reduced stress on the grid but also increased the local consumption of solar power, enhancing energy self-sufficiency.
The economic implications are equally significant. Although the model introduced a small compensation cost for EV owners who traveled farther to charge, this expense was more than offset by savings in grid operation. The total operational cost decreased by over $150 compared to the inflexible charging scenario. “The compensation is a small price to pay for much greater system efficiency,” noted Li Yonghui. “It’s a classic case of a small incentive leading to a large collective benefit.”
The researchers also compared their DRO-based model against deterministic and traditional robust optimization approaches. The deterministic model, which assumes perfect knowledge of future conditions, produced higher renewable capacities but at the risk of frequent constraint violations under real-world uncertainty. The robust model, while safe, was overly conservative, limiting renewable integration by up to 2.1% compared to the DRO approach. The proposed model struck a balance, offering a realistic and economically viable solution that accounts for uncertainty without sacrificing performance.
Another critical aspect of the study was the role of supporting infrastructure. The team found that devices like energy storage systems (ESS) and microturbines had a more pronounced impact on grid capacity than voltage regulation equipment such as capacitor banks. ESS units, in particular, played a dual role: they stored excess solar energy during the day and discharged it during evening peaks, reducing reliance on the main grid. When combined with spatially flexible EV charging, their effectiveness was further amplified. “Storage and flexible demand are synergistic,” explained Yang Jun. “One shifts energy in time, the other in space. Together, they create a more resilient and efficient system.”
The findings have immediate relevance for utility planners, policymakers, and charging infrastructure developers. As cities expand their EV charging networks, the location and capacity of stations should not be determined solely by traffic patterns or real estate availability. Instead, a coordinated approach that considers grid constraints and renewable integration potential is needed. The study suggests that charging navigation platforms could be integrated with grid management systems, allowing them to guide drivers toward stations that are not only convenient but also beneficial for the network.
Moreover, the research underscores the importance of data-driven planning. By leveraging historical data on EV behavior, weather patterns, and grid performance, utilities can build more accurate models of future capacity. The use of 1-norm and ∞-norm constraints in the DRO framework ensures that the model remains robust even with limited data, making it applicable to regions with less mature monitoring systems.
The work also opens new avenues for demand response programs. Rather than simply offering time-based incentives, utilities could introduce spatial incentives—rewarding drivers for charging at underutilized stations during periods of high renewable output. Such programs could be seamlessly integrated into existing navigation apps, providing a user-friendly interface for grid participation.
While the study focused on a specific test system, its methodology is scalable and adaptable. The core idea—that human decision-making in space can be modeled and optimized for grid benefit—applies to any urban or suburban distribution network with growing EV adoption. As autonomous vehicles and vehicle-to-grid (V2G) technologies mature, the potential for spatial and temporal coordination will only increase.
However, the authors caution that their model represents a first step. It assumes that EV owners are willing to accept some deviation from their preferred charging location, which may not hold true for all demographics. Future research could explore behavioral models that account for individual preferences, trip purposes, and psychological factors. Additionally, the current model does not fully integrate temporal flexibility—such as delayed charging or V2G discharging—alongside spatial decisions. A unified spatio-temporal optimization framework could unlock even greater grid benefits.
The study also highlights the need for regulatory and market reforms. For spatial flexibility to be fully realized, charging operators, grid operators, and mobility service providers must share data and coordinate strategies. Current market structures often silo these functions, limiting the potential for system-wide optimization. Policymakers may need to establish new frameworks that incentivize collaboration and value grid-supportive behaviors.
In conclusion, the research by Huang Mengqi, Li Yonghui, Yang Jun, and Wang Mengke demonstrates that the way we think about EV charging must evolve. It is not just a load to be managed but a flexible resource that, when properly guided, can strengthen the grid and accelerate the clean energy transition. By embracing the spatial dimension of EV demand, utilities and planners can unlock hidden capacity, reduce costs, and build a more sustainable power system.
The implications extend beyond technical modeling. They touch on the broader theme of human-centered energy systems—where technology, behavior, and infrastructure converge to create smarter, more resilient communities. As EV adoption continues to rise, the ability to harness the spatial intelligence of millions of drivers could become one of the most powerful tools in the energy transition toolkit.
Huang Mengqi, Li Yonghui, Yang Jun, Wang Mengke, Hubei Engineering and Technology Research Center for AC/DC Intelligent Distribution Network, School of Electrical Engineering and Automation, Wuhan University, Power System Technology, DOI: 10.13335/j.1000-3673.pst.2024.1425