Prospective Students






Data Downloads



Markets and Investment

Market Power:
We have shown that traditional indicators of oligopoly and market power are inapplicable to electricity markets. We developed a game-theoretic model of generators bidding their supply in uniform price auctions. In California, as few as three firms could have jointly set the market price of power more than 90% of the time in 2000. The PJM and New York systems appear to have much more competitive market structures.
Contact: Dmitri Perekhodstev, Seth Blumsack, Lester Lave

Autonomous Agents that Learn in Power Markets:
We have developed a Java-based simulation facility modeling a distributed environment with different clearing mechanisms and different learning algorithms. We find that even “dumb” agents drive the market to monopolistic pricing in a few auctions. We study software agents as traders, including learning, strategies, demand response, and tests for market failures.
Contact: Kong-Wei Lye, Sarosh Talukdar

Transmission Investment Incentives:
We have argued that Locational Marginal Pricing (LMP) does not provide proper incentives for transmission investment, although LMPs do provide incentives to not wheel power over the most congested lines. Instead of relying on LMPs alone, we find that a two-part tariff for transmission, with a megawatt-hour-mile charge and an LMP charge provides more optimum incentives.
Contact: Lester Lave, Jay Apt, Dmitri Perekhodstev

Valuation of Ancillary Services:
We show that the value of ancillary services from hydro facilities can be quantified and is substantial, and we have developed a method of setting internal pricing for ancillary services.
Contact: Dmitri Perekhodstev, Lester Lave

Generation Portfolio Optimization:
We have developed a decision support framework for strategic investment decision-making for generation assets based on efficient frontier portfolio theory.
Contact: Jon Mayes, David Rode, Paul Fischbeck, Jay Apt, Lester Lave

Risk Analysis and Electricity Project Capital Structures:
We have shown that the ability to tie operational risks (such as fuel prices, maintenance expenses, regulatory compliance costs, etc.) directly to impacts on capital structure provides a unifying mechanism for understanding firm value and provides insight into areas in which nonintuitive negative consequences may arise from well-intentioned actions.
Contact: David Rode, Paul Lewis, Paul Fischbeck

Monte Carlo Methods for Appraisal and Valuation as applied to Nuclear Power Plants:
For large-scale, technically unique projects, such as chemical and power plants, and old industrial practices, the four standard appraisal methods are insufficient. To include political, technical, and economic risks in an appraisal, we have developed a Monte Carlo simulation method. Probability distributions are used to model the appropriate uncertainty. The use of Monte Carlo methods and the modeling of future decisions decreased the worth of a nuclear generation plant by 28% as compared to a standard income capitalization method.
Contact: David Rode, Paul Fischbeck, Steven Dean

Risk Management in Electricity Markets:
We examine risk management for power suppliers, bidding strategies of power suppliers with the application of game theory, analysis of markets power, and evaluation of alternative market models ranging from totally centralized dispatch to decentralized bidding. We have constructed complex dynamic models of the electric power system in a restructured environment based upon the principles of control theory, and having information, technical, and policy loops. The control system uses performance metrics as inputs, and disturbances are introduced into the system.
Contact: Zhiyoung Wu, Marija Ilic

Return to Research Page