EV Fleets Power Microgrids Through Storms—A New Resilience Blueprint Emerges
When Typhoon Lekima slammed into China’s eastern seaboard in August 2019, it didn’t just bring record winds and torrential rain—it left behind a trail of systemic fragility. Over 4,000 power lines failed. Nearly 6.8 million customers lost electricity. Some rural islands remained dark for more than 60 hours. In the aftermath, engineers, policymakers, and grid operators shared a sobering realization: resilience—not just reliability—had become the new imperative for modern electricity systems.
Fast-forward to late 2023, and a quiet revolution is taking shape—not in national control rooms, but in clusters of neighborhood-scale microgrids, their rooftops studded with solar panels, their substations humming with battery stacks, and their parking lots increasingly populated by electric vehicles (EVs). In these localized ecosystems, EVs are no longer just commuters’ tools—they’ve become mobile power plants, dynamic buffers, and, in moments of crisis, lifelines.
At the heart of this shift lies a groundbreaking study recently published in Power System Protection and Control, led by researchers Ke Sun, Wengang Chen, Jiajia Chen, and Wenliang Yin from the College of Electrical and Electronic Engineering at Shandong University of Technology. Their work introduces a two-stage resilience-boosting strategy for multi-microgrid systems under extreme weather, with EVs as the linchpin. Unlike prior concepts that treat EVs as static storage units—or sideline them entirely during disasters—this framework leverages their unique mobility, bidirectional energy flow (V2G), and spatiotemporal availability to orchestrate a coordinated, adaptive response across microgrid clusters.
To grasp the significance of their approach, it helps to understand the anatomy of a modern microgrid—and its vulnerabilities.
A microgrid, by definition, is a self-contained energy island: a tightly integrated mix of local generation (solar, wind, small gas turbines), energy storage, controllable loads, and smart controls. It can operate tethered to the main grid—or, crucially, decouple and run in “island mode” when external supply falters. That decoupling capability is the first line of defense against widespread outages. But isolation has its costs. A single microgrid, especially in an urban or industrial zone, may lack sufficient generation or storage to meet all its critical loads for extended periods. Its diesel backup may run dry. Its batteries may deplete under sustained demand. Its solar panels may be obscured by storm clouds or debris.
This is where multi-microgrid systems—networks of individually islandable but mutually supportive microgrids—offer a compelling upgrade. Think of them as neighborhood mutual-aid societies: when one household runs short on rice, another shares its surplus. In energy terms, surplus power from a solar-rich industrial park can flow to a residential zone straining under evening peak loads—even if the main grid is down.
But here’s the catch: during a typhoon, physical interconnections—overhead lines, underground cables—often fail. Poles bend. Transformers flood. Conductors snap. So how can microgrids share resources when the wires between them are severed?
Enter the electric vehicle—not as a novelty, but as a strategic asset.
The Shandong team’s innovation lies in recognizing that EVs are not just batteries on wheels, but mobile, dispatchable, and socially embedded energy carriers. Their availability follows human patterns: office workers charge at commercial buildings by day, return home in the evening, and leave again at dawn. Delivery vans idle at depots overnight. Fleet vehicles rotate predictably through charging hubs. These rhythms create windows of opportunity—windows that, if intelligently harnessed, can be synchronized with grid stress points.
Their two-stage strategy unfolds like a carefully choreographed emergency drill.
Stage One: Local Autonomy, Maximized
The moment a typhoon makes landfall—or when sensors detect imminent disconnection—the system shifts into resilience mode. Each microgrid’s local controller (MGCC) springs into action, re-optimizing its internal resources independently. Solar, wind, gas turbines, stationary batteries, and locally parked EVs are dispatched to minimize load shedding. No assumptions are made about external help. The goal is self-sufficiency, hour by hour.
This stage is essential for speed. Centralized coordination takes time; local autonomy does not. By acting immediately on local forecasts and real-time measurements, each microgrid buys crucial minutes—or hours—before any degradation cascades.
But autonomy has limits. In the team’s simulation, Microgrid 2 (a commercial district) faced a 202.9 kWh shortfall at 7:00 PM—its EVs had mostly departed for the night, its stationary storage was nearly exhausted, and rooftop solar had long since dimmed.
Stage Two: Mobile Mutual Aid, Activated
Here’s where the architecture shines. Once local optimization runs its course, each MGCC reports its net energy deficit or surplus—plus the available capacity and location of its remaining EVs—to a central Energy Management System (EMS). The EMS doesn’t command power flows over damaged lines. Instead, it dispatches vehicles.
Yes—dispatches vehicles.
Based on distance, remaining state-of-charge, and travel energy consumption, the EMS identifies which EVs (or, more realistically, which charging stations with idle EVs) can physically drive—or be driven—to a neighboring microgrid in need. In the simulation, Microgrid 1 (a residential area) had a surplus of idle EVs returning home after work. Microgrid 2 was just 8 km away—closer than Microgrid 3 (10 km). So the EMS prioritized EVs from MG1 to support MG2: first, all 18.5 kWh available from nearby commercial chargers, then an additional 106.24 kWh drawn from MG3’s fleet, compensating for the longer travel distance with higher remaining capacity.
Crucially, the model accounts for the energy cost of mobility itself. Each EV’s usable energy is reduced by an estimated consumption factor per kilometer (e.g., 153–232 Wh/km, depending on model). This ensures that a vehicle doesn’t arrive at its destination with an empty battery—rendering it useless for grid support. It’s a detail many theoretical models overlook but that, in practice, separates feasible plans from fantasy.
The results were striking. Compared to isolated microgrid operation, this EV-enabled mutual aid boosted the system’s resilience index—a metric quantifying the area between full supply and actual delivered power over time—by 6.7%. Over a full 24-hour typhoon scenario, the improvement held consistently across fault-onset times, with peaks exceeding 10% during overlapping evening demand surges and diminishing solar output.
But resilience isn’t just about keeping the lights on—it’s about doing so affordably. Outages cost economies billions. Diesel generators burn expensive fuel. Emergency repairs strain municipal budgets.
Here again, the strategy delivers. In normal (non-emergency) operation, the same coordinated framework—where EVs charge during off-peak hours, discharge during peaks, and trade power across microgrids—reduced total system operating costs by 4.09%, or roughly ¥541 (US$75) per day in their three-microgrid test case. That may sound modest, but scaled across hundreds of microgrid clusters in a single city, the savings become transformative.
And savings aren’t hoarded centrally. The team applied the Shapley value—a Nobel Prize-winning concept from cooperative game theory—to allocate the “emergent benefits” fairly. Rather than splitting savings equally, Shapley rewards each microgrid based on its marginal contribution to the coalition: how much worse off the group would be without it. In their simulation, the industrial microgrid (MG3), with its abundant solar surplus, received the largest share of benefits (¥290.69), followed by the commercial (MG2, ¥155.92) and residential (MG1, ¥94.56) zones. This fairness mechanism is vital: without it, high-contributing members may defect, undermining the entire system.
Critically, the researchers didn’t stop at ideal conditions. They stress-tested their model against internal failures—like a simultaneous gas-supply disruption that knocked out microturbines. Even then, the EV-aided strategy outperformed traditional methods, though resilience dropped predictably during evening peaks when solar faded and demand surged. They also modeled longer-duration outages (3–5 hours), finding that EV support remains effective for roughly 3 hours before state-of-charge depletion begins to erode gains—a sobering but realistic boundary that highlights the need for complementary solutions (e.g., hydrogen backup, demand response).
So what does this mean for the future of urban resilience?
First, it redefines the role of transportation electrification. EV adoption is often framed in terms of emissions reduction or oil independence. This work shows it’s also a grid-hardening strategy. Cities investing in EV infrastructure aren’t just building charging stations—they’re laying the groundwork for decentralized, shock-absorbing energy networks.
Second, it elevates microgrids from niche experiments to core infrastructure. The Department of Energy estimates that microgrids could prevent 80% of outage-related economic losses—if deployed at scale. This research provides a blueprint for how to scale them intelligently: not as siloed islands, but as federated communities, linked by data and mobility more than copper.
Third, it underscores the value of anticipatory design. Most grid upgrades are reactive: replace a failed pole, bury a vulnerable line. But typhoons, wildfires, and ice storms are no longer anomalies—they’re recurring events. Building resilience in advance, using assets already in place (like parked EVs), is far cheaper—and faster—than scrambling post-disaster.
Of course, real-world implementation faces hurdles. Not all EVs support bidirectional charging (though standards like ISO 15118-20 are accelerating adoption). Consumer acceptance of automated V2G dispatch remains uncertain—though pilot programs in California and Japan show strong willingness when tied to financial incentives or emergency preparedness. And cybersecurity for vehicle-to-grid communications must be ironclad.
But the pieces are falling into place. Ford’s F-150 Lightning can power a home for days. Volkswagen’s ID. models ship with V2G-ready hardware. In the UK, Octopus Energy offers “Powerloop” tariffs paying drivers to export battery power during peaks. In Japan, Nissan’s “Vehicle-to-Home” systems helped sustain households after the 2011 earthquake.
What the Shandong team adds is the systemic layer: the algorithms, the coordination protocols, the fairness mechanisms that turn individual capabilities into collective strength.
Imagine a near-future city during a Category 3 storm. As winds shear transmission towers and flood substations, microgrids island themselves automatically. Local controllers fire up gas turbines, tap rooftop solar, discharge home batteries. Then, as dusk falls and batteries wane, fleets of EVs—autonomous or human-driven—begin moving. A delivery van from the industrial park rolls into the downtown core, plugs in, and feeds 50 kWh into a hospital microgrid. A commuter’s sedan, parked at a shopping mall, exports power to keep refrigeration running at a pharmacy. A municipal shuttle, idle at city hall, supports emergency comms at the fire station.
No new wires are laid. No diesel trucks rumble through flooded streets. The energy was already there—distributed, dormant, waiting for a signal.
That’s not science fiction. It’s an operational model, validated in simulation, ready for pilot deployment.
And it starts with seeing an electric car not just as a way to get from A to B, but as a node in a living, breathing grid—one that, when the storm hits, doesn’t just survive. It adapts.
Sun Ke, Chen Wengang, Chen Jiajia, Yin Wenliang
College of Electrical and Electronic Engineering, Shandong University of Technology, Zibo 255000, China
Power System Protection and Control, Vol. 51, No. 24, Dec. 16, 2023
DOI: 10.19783/j.cnki.pspc.230668