Electricity Auction: Optimization, Market Behaviors, and Comparative Studies

University Of Connecticut, Storrs CT

Investigators

Abstract

ECS-0621936 Electricity Auction: Optimization, Market Behaviors, and Comparative Studies Deregulated electricity markets use an auction mechanism to select generation offers and their levels for energy and ancillary services. A settlement scheme is then used to determine the payments. Currently, most independent system operators (ISOs) use an auction mechanism that minimizes the total offer costs ("offer cost minimization"), but determine what consumers have to pay by using a settlement scheme that pays uniform market clearing prices (MCPs) or congestion-dependent Locational Marginal Prices (LMPs). Illustrative examples have shown that for a given set of offers, an auction mechanism that directly minimizes the payment costs ("payment cost minimization") could considerably reduce the total payment cost. This observation has led to serious discussions among stakeholders and policymakers as to which of the two auction mechanisms is more appropriate for ISOs. While methods for offer cost minimization abound, limited approaches for payment cost minimization exist in view of its distinct and fundamental difficulties. In collaboration with Southern California Edison and ISO-New England, this proposal will to develop payment cost minimization methods to include key ISO practices such as simultaneous auctions of energy and ancillary services and transmission capacity constraints. Theoretical and computational issues will be investigated. These results will then be validated by ISO-New England data though a proper construction of scaled-down markets and the innovative training of neural networks to compare the two mechanisms in terms of payment costs, production efficiency, and long-term investment signals. I Intellectual merits include the novel mathematical and computational methods to resolve the new classes of problems, including strategic behaviors of participants, and the innovative validation approach. The research has broader impacts of enabling the research, development, as well as policy-making communities to assess the merits of auction mechanisms in a quantitative manner. The research will also impact the education of students and professionals through courses, online materials, and UConn's da Vinci program for high school teachers and counselors.