Electric Vehicles Power Grid Recovery in Disaster Scenarios
As the world accelerates its shift toward electrified transportation, the role of electric vehicles (EVs) is expanding far beyond personal mobility. A groundbreaking study conducted by researchers at Xi’an Shiyou University reveals how EV charging stations can transform into critical assets for power grid resilience, particularly during disasters that disrupt electricity supply. This innovative research, published in Electric Power Construction, introduces a two-stage recovery strategy that leverages EVs not just as energy consumers but as dynamic, responsive energy sources capable of stabilizing unbalanced distribution networks after catastrophic failures.
The integration of renewable energy sources and the rapid proliferation of electric vehicles have fundamentally altered the landscape of modern power systems. While these advancements support decarbonization goals and enhance energy sustainability, they also introduce new complexities into grid operations. One of the most pressing challenges is the disruption of three-phase balance in distribution networks. Traditional power grids were designed under the assumption of balanced loads across all three phases. However, with the increasing penetration of distributed generation (DG) units such as solar photovoltaics and wind turbines, coupled with the unpredictable charging patterns of EVs, this balance is frequently compromised. The resulting three-phase imbalance can lead to inefficient power flow, increased line losses, equipment stress, and, in extreme cases, widespread outages.
When a natural disaster—such as a storm, earthquake, or wildfire—strikes, the vulnerability of these already-stressed networks becomes even more apparent. Conventional fault recovery strategies often fail to account for the dynamic nature of modern loads and generation, relying instead on static models that assume balanced conditions. This limitation reduces their effectiveness in real-world scenarios where imbalances are the norm rather than the exception. In response to this gap, Li Yingliang, Bai Boxu, Zhu Qi, Yang Yi, and Li Fei from the School of Electronic Engineering at Xi’an Shiyou University have developed a novel approach that redefines how power systems can self-heal after major disruptions.
Their methodology, detailed in Electric Power Construction, proposes a two-stage fault recovery framework specifically designed for unbalanced distribution networks. The first stage focuses on minimizing recovery costs by determining optimal islanding configurations following a fault. An “island” in this context refers to a portion of the grid that becomes electrically isolated from the main network but continues to operate autonomously, powered by local generation resources such as DG units and EV charging stations. The goal is to restore power to as many customers as possible while maintaining system stability and adhering to operational constraints.
What sets this strategy apart is its emphasis on the active participation of EV charging stations in the recovery process. Unlike conventional loads, EVs possess inherent energy storage capabilities. When connected to the grid through smart charging infrastructure, they can function as distributed energy resources (DERs), discharging stored energy back into the network when needed. This bidirectional energy flow, often referred to as vehicle-to-grid (V2G) technology, enables EVs to act as mobile energy reservoirs that can be dispatched during emergencies.
In the second stage of the proposed model, the researchers shift focus from mere restoration to optimization. They construct a collaborative scheduling framework aimed at maximizing the net operational benefit of the distribution network. This involves balancing multiple factors: the depreciation and maintenance costs of DG units and EV charging stations, the financial penalties associated with load shedding (i.e., intentionally cutting power to certain users), and the revenue generated from selling electricity. By integrating these variables into a unified optimization model, the system can dynamically adjust the output of each resource to achieve the most economically efficient outcome.
A key innovation in this work is the explicit consideration of three-phase imbalances in both modeling and control. Most existing studies treat distribution networks as single-phase equivalents, simplifying analysis but sacrificing accuracy. This approximation fails to capture the true behavior of real-world systems, where phase-specific imbalances can lead to voltage fluctuations, neutral conductor overloading, and protective relay malfunctions. The team from Xi’an Shiyou University addresses this shortcoming by employing a three-phase power flow model that accounts for asymmetrical loading and fault conditions.
To validate their approach, the researchers applied it to a modified version of the IEEE 33-node test system—a widely used benchmark in power system analysis. They enhanced the standard model by incorporating three EV charging stations at nodes 7, 13, and 27, each capable of delivering up to 300 kW of dispatchable power. Additionally, they added 500 kW photovoltaic and wind generation units at nodes 17 and 30, respectively, to simulate a realistic mix of renewable energy sources. The simulation scenario involved a simulated disaster occurring at 8:00 a.m., causing simultaneous faults on lines 5, 13, 20, and a single-phase short circuit on line 24—conditions that would severely challenge any conventional recovery system.
Three distinct scenarios were analyzed to evaluate the performance of the proposed method. In the first scenario, only traditional islanding techniques were used, without any coordinated dispatch of EVs or DG units. The results showed significant load shedding, particularly on the phase affected by the short circuit, due to insufficient generation capacity and lack of phase-specific control. The total cost of recovery was high, driven largely by penalties for unmet demand.
In the second scenario, the researchers implemented their two-stage strategy on a balanced network approximation. Here, EV charging stations were actively managed to support islanded sections of the grid. The optimized dispatch led to a noticeable reduction in load shedding and lower overall recovery costs. However, because the model assumed balanced conditions, it could not fully address the asymmetry introduced by the C-phase fault on line 24.
The third and most comprehensive scenario applied the full two-stage strategy to the actual unbalanced network. In this case, the model allowed for independent control of each phase at every EV charging station. This capability proved crucial: the station at node 27, responsible for powering the downstream section affected by the C-phase fault, increased its output specifically on the C phase to compensate for the imbalance. This targeted response prevented excessive voltage deviation and ensured more equitable power delivery across all phases.
The comparative analysis revealed compelling advantages of the proposed method. Relative to the traditional approach (Scenario 1), the two-stage strategy reduced total recovery costs by 9,186.7 yuan and decreased the total amount of shed load by 2,807.02 kW. Even when compared to the balanced-network version of the strategy (Scenario 2), the full three-phase control offered superior performance in terms of load restoration and voltage stability. Although the operational revenue was slightly lower due to the complexity of phase-specific dispatch, the overall cost savings and improved reliability justified the approach.
Beyond the quantitative improvements, the study highlights a paradigm shift in how we think about grid resilience. Rather than viewing EVs solely as loads that strain the system during peak hours, this research positions them as integral components of a flexible, adaptive energy infrastructure. By enabling EVs to participate in grid support functions—especially during emergencies—we unlock a vast, distributed reservoir of energy that can be mobilized precisely when and where it is needed most.
This concept aligns with broader trends in smart grid development, including the rise of microgrids, advanced metering infrastructure, and artificial intelligence-driven energy management systems. However, the practical implementation of such strategies requires more than just technological readiness. It demands supportive regulatory frameworks, market mechanisms that incentivize consumer participation, and robust cybersecurity measures to protect against potential threats.
One of the critical enablers of this vision is the evolution of charging infrastructure. Modern EV chargers, especially those at commercial or fleet facilities, are increasingly equipped with bidirectional power electronics and communication interfaces. These features allow them to respond to grid signals, modulate their charging rate, or even export power during emergencies. As the number of EVs on the road continues to grow—projected to exceed 200 million globally by 2030—the aggregate energy storage capacity they represent will become a strategic national asset.
The research also underscores the importance of accurate modeling and simulation in developing effective grid management strategies. By moving beyond simplified single-phase representations and embracing the complexity of three-phase unbalanced systems, engineers can design solutions that are not only theoretically sound but practically viable. The use of advanced optimization tools, such as mixed-integer programming and convex relaxation techniques, enables the solution of complex, multi-objective problems within computationally feasible timeframes.
Moreover, the integration of EVs into disaster recovery planning has implications for urban resilience and emergency preparedness. In the aftermath of a major event, access to electricity is essential for communication, medical care, refrigeration, and basic lighting. By deploying EV fleets as mobile power sources, cities can enhance their ability to maintain critical services even when the central grid is compromised. This capability is particularly valuable in remote or underserved areas where grid infrastructure may be less robust.
From a policy perspective, the findings suggest that governments and utilities should consider incorporating V2G capabilities into their grid modernization plans. Incentive programs, such as time-of-use pricing, demand response compensation, and emergency service agreements with EV fleet operators, could accelerate adoption. Furthermore, standards for interoperability—such as ISO 15118 and IEEE 2030.5—are essential to ensure that different manufacturers’ equipment can communicate and coordinate effectively.
The environmental benefits of this approach should not be overlooked. By optimizing the use of existing distributed energy resources, including renewable generation and EV batteries, the need for fossil-fueled backup generators can be reduced. This contributes to lower greenhouse gas emissions and improved air quality, especially in densely populated urban centers.
Looking ahead, the next frontier in this domain may involve the integration of other flexible loads and storage systems, such as home energy management systems, building HVAC controls, and stationary battery storage. Combining these resources into a holistic, multi-vector energy network could further enhance system resilience and efficiency.
In conclusion, the work by Li Yingliang and his colleagues at Xi’an Shiyou University represents a significant step forward in the quest for smarter, more resilient power systems. By reimagining electric vehicles not just as transportation devices but as active participants in grid operations, they offer a compelling vision of a future where energy is not only clean and efficient but also inherently reliable. As climate change increases the frequency and severity of extreme weather events, such innovations will be essential to ensuring that our critical infrastructure can withstand the shocks of the 21st century.
The full study, titled “Self-healing and Optimal Operation of Unbalanced Distribution Network Based on Electric Vehicle Charging Station,” was published in Electric Power Construction, Vol. 45, No. 5, May 2024, with DOI: 10.12204/j.issn.1000-7229.2024.06.004. The research was supported by the National Natural Science Foundation of China, the National Key R&D Program of China, and the Natural Science Basic Research Program of Shaanxi Province. Li Yingliang, Bai Boxu, Zhu Qi, Yang Yi, and Li Fei are affiliated with the School of Electronic Engineering at Xi’an Shiyou University, Xi’an, China.