A Hybrid Mean Field Framework for Aggregators Participating in Wholesale Electricity Markets

TL;DR

This paper introduces a hybrid mean-field framework combining control and game theory to model DER aggregator strategies in wholesale electricity markets, accounting for market feedback and uncertainties.

Contribution

It develops a novel hybrid MFC-MFG model that captures strategic interactions and market feedback, incorporating reinforcement learning for dynamic bidding strategies.

Findings

Reduces price volatility in the market

Improves overall market efficiency

Validates the approach with a case study on Oahu Island

Abstract

The rapid growth of distributed energy resources (DERs), including rooftop solar and energy storage, is transforming the grid edge, where distributed technologies and customer-side systems increasingly interact with the broader power grid. DER aggregators, entities that coordinate and optimize the actions of many small-scale DERs, play a key role in this transformation. This paper presents a hybrid Mean-Field Control (MFC) and Mean-Field Game (MFG) framework for integrating DER aggregators into wholesale electricity markets. Unlike traditional approaches that treat market prices as exogenous, our model captures the feedback between aggregators' strategies and locational marginal prices (LMPs) of electricity. The MFC component optimizes DER operations within each aggregator, while the MFG models strategic interactions among multiple aggregators. To account for various uncertainties, we incorporate reinforcement learning (RL), which allows aggregators to learn optimal bidding strategies in dynamic market conditions. We prove the existence and uniqueness of a mean-field equilibrium and validate the framework through a case study of the Oahu Island power system. Results show that our approach reduces price volatility and improves market efficiency, offering a scalable and decentralized solution for DER integration in wholesale markets.