Abstract
Shared energy storage (SES) offers a promising "storage-as-a-service" model to improve energy efficiency and reduce costs. However, establishing a fair business model for diverse participants remains a challenge. This paper proposes an optimal scheduling method based on asymmetric Nash bargaining to manage cooperative operations among power generation users, demand-side consumers, and microgrid prosumers. By decomposing the problem into social cost minimization and equitable benefit distribution, the model ensures that rewards align with individual contributions. An adaptive ADMM algorithm is utilized to solve the equilibrium while protecting user privacy. Numerical results demonstrate a 41.3% reduction in total social costs and improved fairness compared to symmetric models.
Introduction
As the global energy landscape shifts toward renewable sources, the need for flexible energy storage becomes critical. While individual storage systems can balance local supply and demand, their high capital costs and often low utilization rates pose significant barriers to adoption. Shared Energy Storage (SES) presents a collaborative alternative, allowing multiple users to access a centralized storage resource. Despite its potential, the adoption of SES is hindered by the absence of a transparent and equitable business model. Existing models often fail to account for the varying contributions of different user types, leading to suboptimal participation and instability within the cooperative alliance.
The Cooperative Framework
To address these challenges, this research introduces a comprehensive framework for SES operation involving a diverse set of stakeholders.
As illustrated in the framework above, the system comprises four primary entities:
- SES Operator: Manages the physical storage assets and coordinates charging/discharging.
- Generation-side Users: Renewable energy plants (e.g., wind and solar) that provide surplus energy.
- Demand-side Consumers: Industrial or commercial users with controllable loads.
- Microgrid Prosumers: Entities that both consume and generate energy, often utilizing local resources like diesel engines.
The framework facilitates energy and information exchange between these users, the SES operator, and the main grid, aiming to maximize collective benefits through coordinated scheduling.
Methodology: Asymmetric Nash Bargaining
The core of the proposed solution is an asymmetric Nash bargaining model. Unlike traditional symmetric bargaining, which treats all participants equally, the asymmetric approach assigns "bargaining power" based on quantitative metrics of contribution:
- Marginal Contribution: The specific reduction in total social cost attributable to a single user's participation.
- Energy Interaction: The net volume of energy a user exchanges with the SES system.
- Peak-time Contribution: A measure of how much a user helps stabilize the grid during periods of peak demand.
The problem is solved using a two-stage decomposition. The first stage focuses on scheduling—optimizing the physical operation of the battery to minimize the total social cost. The second stage handles benefit distribution, determining the service fees and subsidies required to ensure that the resulting profits are shared fairly according to each user's bargaining power.
To solve this complex optimization while maintaining data privacy, the researchers employed an Adaptive Penalty Factor Alternating Direction Method of Multipliers (ADMM). This distributed algorithm allows participants to reach an equilibrium without sharing sensitive internal operational data, such as specific load profiles or fuel costs. The adaptive penalty factor dynamically adjusts during computation, significantly accelerating convergence compared to standard ADMM implementations.
Results and Discussion
The model was validated using a numerical simulation based on real-world data from Jiangsu Province, China. The results highlight several key advantages of the asymmetric cooperative approach:
- Significant Cost Savings: The total social cost for the alliance dropped by 41.3% compared to a non-cooperative scenario, falling from 48,130.8 ¥ to 19,873.4 ¥.
- Enhanced Fairness: By utilizing asymmetric bargaining, the model rewarded high-contributing users more effectively. For instance, a key renewable energy provider achieved a profit of 7,930.1 ¥, nearly double what they would have received under a standard "equal split" model. This ensures that the most valuable participants have a strong incentive to remain in the alliance.
- Computational Efficiency: The adaptive ADMM algorithm achieved a 37.8% reduction in solution time, demonstrating its suitability for real-time or near-real-time operational scheduling.
Conclusion
This research provides a robust mathematical and economic foundation for the "storage-as-a-service" model. By recognizing the diverse contributions of different energy users through asymmetric Nash bargaining, the proposed framework ensures both economic efficiency and long-term alliance stability. Future work will focus on incorporating stochastic programming to account for the inherent uncertainty in renewable energy forecasts and integrating detailed power flow analysis to ensure the physical feasibility of the proposed schedules within local distribution grids.
Summary updated