EVs and Thermal Power Unite for Smarter Grid Frequency Control
The integration of renewable energy into modern power grids is advancing at an unprecedented pace. Wind and solar generation, while essential for achieving global decarbonization targets, introduce inherent challenges due to their intermittent and variable nature. These fluctuations can destabilize grid frequency, a critical parameter for the reliable and safe operation of any electrical system. As nations strive toward ambitious climate goals, the stability of the power grid has become a focal point for innovation. A groundbreaking study by researchers at the Yichang Key Laboratory of Intelligent Operation and Security Defense of Power System at China Three Gorges University, in collaboration with State Grid International Development Co., Ltd., presents a sophisticated new strategy that leverages the untapped potential of electric vehicles (EVs) to bolster grid resilience. This research, published in the prestigious journal Power System Protection and Control, details a two-stage optimization control strategy that enables EVs to work in concert with traditional thermal power units, transforming them from passive consumers into active participants in the grid’s frequency regulation services. The findings offer a compelling blueprint for a more stable, efficient, and economically viable power system in the era of high renewable penetration.
The core challenge addressed by this research is the increasing difficulty of maintaining system frequency within strict operational limits. Traditional frequency regulation relies on thermal power plants adjusting their output to match supply and demand in real time. However, this method has significant drawbacks. Constantly ramping power plants up and down to counteract renewable fluctuations moves them away from their most efficient operating points, increasing fuel consumption, accelerating equipment wear and tear, and elevating carbon emissions. Furthermore, the physical inertia of large turbines means their response time, while adequate for slower changes, is not as instantaneous as the rapid frequency deviations that can occur with sudden drops in wind or solar output. This creates a critical gap in the grid’s ability to respond to fast, dynamic imbalances. The research team, led by Professor Cheng Shan, recognized that the growing fleet of EVs represents a massive, distributed, and highly responsive energy resource that could help bridge this gap. With over ten million pure electric vehicles on Chinese roads by the end of 2022, the aggregate battery capacity of these vehicles is staggering. The key lies in harnessing this capacity through Vehicle-to-Grid (V2G) technology, which allows bidirectional energy flow between the vehicle and the power grid. By intelligently managing the charging and discharging cycles of thousands of EVs, a fleet can act as a virtual power plant, providing fast-acting power to stabilize frequency without the need for additional fossil fuel generation.
The proposed strategy is a masterclass in system optimization, designed to maximize both technical performance and economic benefit. It operates in two distinct but interconnected stages, creating a dynamic and responsive control loop. The first stage, the “allocation proportion optimization,” is where the overarching economic and strategic decision is made. When the grid dispatch center detects a frequency deviation and calculates the required Area Regulation Requirement (ARR), this signal is sent to the frequency regulation provider, or “frequency modulation agent.” Instead of relying solely on a thermal power plant or assigning a fixed percentage of the task to an EV fleet, the agent employs a rolling optimization algorithm. This algorithm, recalculated every minute, determines the optimal split of the ARR between the thermal unit and the EV charging station. The primary objective is clear: to maximize the total profit for the regulation provider. This profit is a function of the revenue earned from the ancillary services market—comprising payments for both the available capacity and the actual energy delivered (referred to as “mileage”)—minus the operational costs incurred. The cost of using the thermal plant is modeled as a quadratic function of its power deviation, reflecting the real-world inefficiencies and wear-and-tear costs associated with moving away from its optimal load. The cost of using EVs is more nuanced, centered on the concept of battery degradation. Every charge and discharge cycle contributes to the gradual aging of a battery. The model incorporates a compensation cost for this degradation, ensuring that the economic analysis is realistic and accounts for the long-term value of the EV’s battery. By balancing these revenue streams and costs in real time, the algorithm can dynamically shift the burden of regulation to whichever resource is most economical at that precise moment, often favoring the lower-cost EV response for rapid, small adjustments.
The second stage, the “EV charging station response,” is where the high-level command from the first stage is translated into action at the individual vehicle level. This is where the strategy demonstrates its sophistication and respect for user needs. The researchers understood that a successful V2G program cannot compromise the primary function of an EV: to be ready for its owner’s next trip. To ensure this, they implemented a “zoning and prioritization” method based on each vehicle’s State of Charge (SOC). Vehicles are categorized into three distinct zones. The first zone comprises vehicles with an SOC below a critical minimum threshold. These vehicles are in an emergency charging state and are excluded from any frequency regulation service to guarantee they can meet their owners’ driving needs. The second zone includes vehicles with an SOC above the minimum but below a user-defined “expected” level. For these vehicles, the only available service is to modulate their charging power. They can slow down their charge (reducing grid demand) or speed it up (increasing grid demand) within safe limits, but they cannot discharge. This provides a safe and non-intrusive way to contribute to grid stability. The third zone is for vehicles with an SOC above the expected level. These vehicles have a “buffer” of extra charge and can participate in both upward and downward regulation. They can reduce their charging rate, stop charging, or even discharge back to the grid (V2G), providing a more powerful and flexible response. Crucially, the control system is programmed to ensure that even when discharging, the vehicle’s SOC will not fall below the expected level by the time it leaves the charging station.
This zoning system is coupled with a clear hierarchy of response priorities, which is essential for efficient and predictable operation. When the grid needs to reduce power (a downward regulation command), the system first looks to the vehicles in the third zone, asking them to reduce their charging power. If more reduction is needed, it then asks the second zone vehicles to reduce their charging. Only if these measures are insufficient will the system request the third zone vehicles to actively discharge power back to the grid. This priority ensures that the most valuable service—actual power generation from the battery—is used only when absolutely necessary, minimizing battery degradation. The reverse priority applies for upward regulation: the system first stops or reduces any discharging from the third zone, then increases the charging power of the second zone, and finally increases the charging power of the third zone. Within each zone, the required power adjustment is distributed proportionally among the participating vehicles based on their available flexibility, ensuring a fair and balanced load on the fleet. This granular control, managed by a central aggregator that communicates with each smart V2G charger, allows for a highly responsive and precise modulation of the entire charging station’s power profile.
To validate the effectiveness of this two-stage strategy, the research team constructed a detailed simulation model of a two-area interconnected power system, a standard benchmark in power system analysis. This model included a thermal power unit and an EV charging station in each area, subjected to realistic load and wind power fluctuations to simulate the frequency disturbances common in a renewable-rich grid. The simulation compared three different scenarios. The first, a baseline, used only the thermal power units for regulation. The second used a fixed 50/50 split of the regulation task between the thermal units and the EV stations. The third was the proposed dynamic, rolling optimization strategy. The results were unequivocal. The scenario using only thermal units showed the largest frequency deviations, with a peak-to-peak swing of over 0.18 Hz. The fixed-split scenario improved this significantly, but the proposed dynamic strategy achieved the smallest frequency deviations by a clear margin. The root mean square (RMS) value of the frequency error, a key metric for overall performance, was reduced by more than 50% compared to the baseline and by a further 25% compared to the fixed-split method. Similarly, the deviations in the power flowing between the two interconnected areas—a critical measure of grid stability—were also minimized under the new strategy. This superior technical performance stems from the EVs’ ability to respond almost instantaneously to high-frequency components of the grid’s deviation, a task for which their power electronics are ideally suited, while the slower-responding thermal units handle the longer-term, bulk adjustments.
The economic analysis presented in the study is equally compelling. The simulation calculated the total revenue for the frequency regulation provider under each strategy. The baseline scenario, relying solely on the thermal plant, yielded the lowest profit. The fixed-split scenario, by incorporating EVs and thus reducing the costly thermal plant adjustments, increased the total profit. However, the proposed dynamic optimization strategy generated the highest profit of all. This is because the algorithm continuously seeks the most cost-effective way to meet the regulation demand. By preferentially using the lower-cost EV response for frequent, small adjustments, it significantly reduces the power deviation cost of the thermal plant. The data showed a dramatic reduction in this cost under the new strategy. While the strategy also incurs a higher battery degradation compensation cost—because the EVs are being used more intensively and intelligently—the savings on the thermal plant side far outweigh this expense. The net result is a substantial increase in total profit, demonstrating that this is not just a technically superior solution but a financially sound one for market participants. This economic incentive is crucial for driving the adoption of such strategies by utilities and aggregators.
Beyond the immediate technical and economic benefits, this research has profound implications for the future of transportation and energy. It represents a significant step toward a truly integrated energy ecosystem. By proving that EVs can be a reliable and profitable source of grid services, it strengthens the value proposition of electrified transportation. It moves the conversation beyond just reducing tailpipe emissions to one where the EV becomes an active, revenue-generating asset for its owner and a critical component of grid infrastructure. For EV owners, the promise of financial compensation for battery degradation provides a tangible incentive to participate in V2G programs, addressing a major barrier to user adoption. For grid operators, it offers a new, fast-acting tool to manage the volatility of renewable energy, enhancing grid reliability and potentially deferring the need for expensive new transmission or generation infrastructure. The strategy’s focus on user needs, by protecting the minimum and expected SOC, is key to its practicality and long-term sustainability. It ensures that grid services are provided without inconveniencing the vehicle owner, a fundamental requirement for any large-scale V2G deployment.
The success of this strategy hinges on the maturity of several enabling technologies. Widespread deployment of smart, bidirectional V2G chargers is a prerequisite. These chargers must be able to receive control signals from an aggregator, precisely manage the vehicle’s charging current, and communicate the vehicle’s SOC and available flexibility back to the central system. A robust and secure communication network is also essential to handle the real-time data flow between thousands of vehicles, charging stations, and the grid operator. The role of the EV aggregator is central to this model. This entity acts as the intermediary, managing the fleet, running the complex optimization algorithms, and bidding into the ancillary services market on behalf of the vehicle owners. The research from China Three Gorges University provides a powerful algorithmic foundation for such an aggregator. As battery technology continues to improve, with longer lifespans and faster charging capabilities, the economics of V2G will only become more attractive. Future research, as the authors suggest, will likely focus on refining the economic models to create even more equitable revenue-sharing schemes between the aggregator and the EV owners, further boosting participation.
In conclusion, the work by Cheng Shan and his colleagues presents a comprehensive and highly effective solution to one of the most pressing challenges of the modern power grid. Their two-stage optimization strategy, which dynamically allocates frequency regulation tasks between thermal power and EV fleets while meticulously respecting user charging needs, is a significant advancement in the field. It successfully demonstrates that EVs are not just a load on the grid but a powerful, flexible, and economically valuable resource. By reducing frequency response bias and increasing the economic benefits for service providers, this strategy paves the way for a more resilient, efficient, and sustainable energy future. It is a clear example of how intelligent control systems can unlock the hidden potential of existing technologies to solve complex problems, turning the challenge of renewable integration into an opportunity for innovation.
Cheng Shan, Li Fengyang, Liu Weiwei, Li Mao, Wang Can, Yichang Key Laboratory of Intelligent Operation and Security Defense of Power System(China Three Gorges University), State Grid International Development Co., Ltd., Power System Protection and Control, DOI: 10.19783/j.cnki.pspc.230747