Session: Dynamic flexibility, risk-awareness, and mechanisms for price-quantity alignment in electricity markets
Chair: Andy Sun
Cluster: Optimization Applications (Communication, Energy, Health, ML, ...)
Talk 1: Dynamic transmission capacity adjustment in large-scale electricity markets
Speaker: Baptiste Rabecq
Abstract: In this talk, we present a new optimization model for the real-time scheduling of a power grid that incorporates the so-called dynamic line ratings (DLR), where transmission lines' capacity can be adjusted in real-time according to the ambient weather conditions and the line flows through differential equations. We develop a new decomposition method that decouples AC optimal power flows from temperature dynamics with guaranteed global convergence. We will present a computational study to demonstrate the benefits of dynamic flexibility from DLR in a large-scale power grid. This is joint work with Thomas Lee and Andy Sun.
Talk 2: Handling real-time volatility in the SCUC computation
Speaker: Daniel Bienstock
Abstract: Under the typical real-time energy markets setup, any real-time load in excess of forecast must be paid for using the real-time LMPs. This introduces a form of risk faced by load-serving entities (and hence, by the public) that real-time volatility in generation, or in loads, will give rise to significant excess payments. Here, "real-time volatility" describes deviations from expected values which take place in short time frames -- a phenomenon that has been observed in power systems with high renewable penetration, or intelligent loads, especially under peak conditions. We describe an adjustment to the SCUC computation that is explicitly aware of potential volatility conditions during peak hours of the forthcoming day. The output of this updated SCUC has the same format as that for current SCUC; it is the nature of the computed commitments that will change. We will present computational experiments and describe the algorithm. Joint work with Y. Dvorkin, C. Guo, R. Mieth and J. Wang.
Talk 3: Mechanisms for Ensuring Price and Quantity Agreements in Electricity Day-Ahead Markets
Speaker: John Birge
Abstract: Many electricity markets include day-ahead energy markets as a mechanism to commit generators which require fixed costs for setup and have operating constraints. Based these markets on deterministic forecasts inherently cannot match both prices and quantities with the expected real-time values. This talk will discuss mechanisms that build stochastic representations of the market and how these can be solved to obtain efficient generator commitments that match day-ahead prices and quantities with their real-time expectations.
