EV and Wind Power Real-Time Bidding Strategy Unveiled for Energy-Frequency Markets
In a significant advancement for the integration of renewable energy and electric mobility into power systems, a team of researchers has introduced a novel real-time bidding strategy enabling electric vehicles (EVs) and wind power to jointly participate in energy and frequency regulation markets. The study, published in the Chinese Journal of Automotive Engineering, presents a comprehensive framework that addresses the inherent uncertainty of both EV availability and wind generation, offering a pathway to enhance grid stability while maximizing economic returns for participants.
As nations accelerate their transition toward decarbonized energy systems, the role of variable renewable sources like wind power and flexible loads such as EVs has become increasingly critical. However, the intermittent nature of wind and the unpredictable charging and discharging patterns of EVs pose substantial challenges to grid operators striving to maintain real-time balance between supply and demand. Traditional market mechanisms, often designed around predictable fossil fuel generation, struggle to accommodate these new, dynamic resources effectively. The mismatch between day-ahead market schedules and real-time physical delivery can result in costly imbalances, penalizing participants and undermining system reliability.
The research team, led by Pang Songling from the Electric Power Research Institute of Hainan Power Grid Co., Ltd., and the Smart Grid and Island Microgrid Joint Laboratory, in collaboration with Hao Ruiyi and Zhang Qian from the State Key Laboratory of Power Transmission Equipment Technology at Chongqing University, has developed a sophisticated two-level optimization model to tackle this challenge. Their approach moves beyond simple participation, focusing on the creation of a robust market mechanism that incentivizes accurate forecasting and responsive behavior through a well-defined deviation assessment system.
At the heart of their strategy is the concept of a “dynamic scheduling boundary” for EVs. Unlike static models that rely on pre-defined availability windows, this framework continuously updates the feasible range of power and energy that an EV fleet can provide. This is achieved by the EV aggregator, a central entity that manages a large number of individual vehicles, constantly collecting real-time data on each vehicle’s state of charge (SOC), its expected departure time, and its charging needs. As vehicles plug in and out of the network throughout the day, the aggregator dynamically reclassifies them into different clusters based on these parameters. This real-time classification allows the aggregator to form an accurate picture of the fleet’s collective flexibility at any given moment, a capability essential for making informed decisions in the fast-paced real-time market.
This dynamic boundary model is coupled with a rapid power allocation algorithm. Once the aggregator has determined the optimal total charging, discharging, and frequency regulation capacity for a specific time period based on market signals and its own optimization goals, it must efficiently distribute this plan across thousands of individual vehicles. The proposed algorithm ensures this distribution is done with minimal deviation from the ideal per-vehicle schedule, respecting the physical constraints of each battery, such as maximum charge and discharge rates and SOC limits, while also honoring the user’s declared charging requirements. This seamless translation from a fleet-level decision to individual vehicle commands is a key enabler for the practical implementation of Vehicle-to-Grid (V2G) technology at scale.
The core innovation of the study lies in its integrated approach to the energy and frequency regulation markets. Instead of treating these as separate entities, the model allows EVs and wind farms to bid for both energy delivery and their capacity to provide fast-responding frequency regulation services simultaneously. This dual participation is crucial because it allows these resources to capture value from multiple market streams. For instance, an EV can earn revenue not only by charging when electricity prices are low and discharging when they are high (energy arbitrage) but also by being on standby, ready to inject or absorb power within seconds to correct frequency deviations, for which it is paid a separate capacity fee.
To ensure that participants have a strong incentive to adhere to their commitments, the model incorporates a detailed deviation assessment mechanism. This is a critical component for market integrity. In real-time operations, the actual power output of a wind farm or the net power consumption of an EV fleet will inevitably deviate from the amount they bid for in the day-ahead market due to unforeseen changes in wind speed or user behavior. The proposed mechanism quantifies the cost of these deviations. If an entity over-delivers energy (e.g., the wind blows stronger than predicted), it may be compensated at a lower real-time price. If it under-delivers, it faces a higher penalty, often significantly greater than the reward for over-delivery, to discourage under-bidding and ensure system reliability. Similar penalties apply to deviations in the provision of frequency regulation capacity. This asymmetric penalty structure is designed to promote conservative and accurate forecasting.
The mathematical framework is structured as a bi-level optimization problem, reflecting the hierarchical nature of the market. At the upper level, individual market participants—the EV aggregator and the wind power supplier—act as leaders. Their primary objective is to minimize their own deviation assessment costs and operational expenses, such as battery degradation for EVs. They make their bidding decisions based on their internal models of their available resources and the current market conditions. At the lower level, the power exchange acts as a follower. Its objective is to minimize the overall system operating cost, which includes the cost of energy, the cost of procuring frequency regulation services, and the cost of balancing any remaining imbalances through more expensive reserve generators. The exchange clears the market by determining the real-time nodal marginal prices based on all the bids it receives.
The interaction between these two levels is complex and iterative. The actions of the upper-level players influence the market-clearing prices set by the lower-level exchange, which in turn feed back and influence the optimal bidding strategies of the participants. To solve this computationally challenging problem, the researchers employed the Karush-Kuhn-Tucker (KKT) conditions, effectively transforming the bi-level problem into a single-level mixed-integer linear program. This methodological choice makes the model tractable for real-world application, allowing for timely decision-making in a 15-minute market interval.
The paper presents a detailed case study to validate the effectiveness of the proposed strategy. The simulation, conducted over a 24-hour period with 96 fifteen-minute intervals, uses realistic parameters for EV fleets and wind generation profiles. The results are compelling. The dynamic scheduling boundary model successfully captures the ebb and flow of EV availability, with the fleet’s power and energy capacity dropping to zero as vehicles leave their charging stations in the morning and recovering as they return in the evening. This dynamic view is far more accurate than a static, day-ahead estimate.
For wind power, the strategy demonstrates intelligent adaptation to market conditions. During periods of high frequency regulation prices, the wind farm strategically bids to provide upward regulation capacity, positioning its power output below its maximum forecasted capability to retain headroom for rapid increases. Conversely, when energy prices are high and its power output is near its peak, it focuses on bidding its energy deviation, accepting that its ability to provide downward regulation is limited by its physical ramping constraints. The analysis shows that wind power often over-generates compared to its day-ahead schedule, turning potential imbalance penalties into a source of additional revenue when the real-time price is favorable.
The impact on EVs is equally profound. The model shows that EVs can engage in profitable energy arbitrage without compromising user needs. By charging extra when wind power is over-generating and electricity prices are low, and discharging this stored energy during peak-price hours, EV owners can maximize their financial benefit. Crucially, the strategy ensures that this arbitrage activity is performed within the bounds of the vehicle’s dynamic scheduling boundary, guaranteeing that the car will be sufficiently charged by its owner’s departure time. Furthermore, the inclusion of battery degradation costs in the optimization prevents excessive cycling, promoting the long-term health of the vehicle’s battery.
The study also highlights the synergistic relationship between EVs and wind power. The flexibility of the EV fleet acts as a natural buffer for the variability of wind generation. When wind power exceeds forecasts, EVs can absorb the excess energy by charging more. When wind power is lower than expected, EVs can discharge to help meet demand. This mutual support reduces the overall system imbalance, lowering the need for expensive balancing services from conventional power plants and contributing to a more stable and efficient grid.
The implications of this research extend far beyond the theoretical. It provides a practical blueprint for how grid operators and market designers can create a more inclusive and resilient electricity market. By establishing clear rules for deviation penalties and enabling fast, accurate decision-making through dynamic modeling, the strategy lowers the barrier to entry for distributed energy resources. It transforms EVs from passive loads into active, revenue-generating grid assets, which can significantly improve the business case for V2G technology and encourage greater consumer adoption.
For wind farm operators, the strategy offers a more sophisticated way to monetize their generation, moving beyond simple energy sales to participate in higher-value ancillary services. This diversification of revenue streams can improve the financial viability of wind projects, especially in markets with high penetration of renewables where energy prices can be volatile.
The work also underscores the importance of advanced data analytics and real-time communication infrastructure. The success of the dynamic boundary model depends on the reliable and timely flow of data from individual EVs to the aggregator. This necessitates robust cybersecurity measures and standardized communication protocols, which are key components of a modern smart grid.
While the current model represents a significant leap forward, the authors acknowledge areas for future exploration. The study focuses on a single EV aggregator; a more complex market with multiple competing aggregators would introduce new strategic dynamics. The inclusion of other flexible resources, such as distributed battery storage and demand response from commercial buildings, could further enhance system flexibility. Moreover, the development of effective incentive mechanisms to encourage EV owners to participate in such programs remains a critical social and economic challenge.
In conclusion, the research by Pang Songling, Zhao Yunan, Li Linwei, Ma Lihong, Fan Kaidi, Hao Ruiyi, and Zhang Qian presents a groundbreaking framework for the real-time integration of electric vehicles and wind power into energy and frequency regulation markets. By combining dynamic scheduling, rapid power allocation, and a well-calibrated deviation assessment mechanism within a rigorous bi-level optimization model, they have created a strategy that not only ensures real-time power balance but also maximizes the economic benefit for both resource owners and the power system as a whole. This work is a vital step toward a future where the variability of renewable energy is not a liability but a managed characteristic, and where the millions of electric vehicles on the road become a cornerstone of a flexible, sustainable, and economically efficient power grid.
Pang Songling, Zhao Yunan, Li Linwei, Ma Lihong, Fan Kaidi, Hao Ruiyi, Zhang Qian, Electric Power Research Institute of Hainan Power Grid Co., Ltd., Smart Grid and Island Microgrid Joint Laboratory, State Key Laboratory of Power Transmission Equipment Technology, Chongqing University, Chinese Journal of Automotive Engineering, DOI: 10.3969/j.issn.2095‒1469.2024.06.08