A game-theoretical approach to variable renewable generator bidding in wholesale electricity markets

摘要

We develop a two-part modeling framework to schedule bids for merchant variable renewable energy (VRE) generators participating in the day-ahead wholesale electricity market. The first part is an optimization model that maximizes revenue by optimally selling power into the marketplace. The second is a non-cooperative game-theory model in which the Nash equilibrium is the revenue-maximizing bidding strategy for all VRE market participants. The two models are calibrated using three months of consecutive demand, generation, and market clearing price (MCP) data from the California wholesale market. Analytical results provide three important conclusions for VRE operators, policymakers, and system operators alike: (1) there are periods when VRE operators can increase revenue by curtailing power output; (2) such market outcomes occur with only moderate increases in California's renewables penetration (from present-day 20% to 26%); and (3) VRE operators can increase revenue by forming aggregations, i.e. coupling bids between multiple merchant VRE.