Date of Completion


Embargo Period



Cutting plane methods, Electricity markets, Extended LMPs, Lagrangian relaxation, Commitment cost allocation, Monte-Carlo simulation, Ramp capabilities, Real-time dispatch, Renewable energy, Requirement Design

Major Advisor

Peter B. Luh

Associate Advisor

Yaakov Bar-Shalom

Associate Advisor

Paul Gribik

Field of Study

Electrical Engineering


Doctor of Philosophy

Open Access

Campus Access


In this dissertation, two topics in the area of wholesale electricity market optimization and economics are investigated: optimization and cost allocation for Extended Locational Marginal Prices, and ramp requirement design for renewable energy integration.

In the current U.S. electricity markets, Locational Marginal Prices (LMPs) are obtained in economic dispatch with fixed commitment decisions. The costs of committing fast-start units or dispatching them at their minimum limits may not be covered by LMPs and significant uplift payments may thus be needed. To appropriately reflect these costs, Extended LMPs (ELMPs) were established as the optimal Lagrangian multiplier of the dual of the unit commitment and economic dispatch (UCED) problem. It is important to obtain ELMPs for the investigation and demonstration of their economic features. Nevertheless, solving the dual problem with multiplier optimality and computational efficiency has been a challenge. We develop a subgradient simplex cutting plane method by innovatively using subgradients and simplex tableaus to efficiently obtain ELMPs. For the market implementation of ELMPs, an approximate ELMP (aELMP) model has then been developed, and an important design issue is to allocate commitment costs of fast-start units based on their commitment and dispatch, so that the resulting aELMPs effectively approximate ELMPs. We derive allocation guidelines based on Karush-Kuhn-Tucker (KKT) conditions by decomposing the ELMP model into individual-hour problems resembling the aELMP model.

With the national efforts to increase the penetration of intermittent renewable energy, new issues emerge in the U.S. electricity markets. Several ISOs are instituting a new product on ramp capabilities to manage the resulting operational challenges of maintaining real-time power balance. Ramp capabilities have been required ten minutes ahead based on the Gaussian-sigma rule without considering the costs, e.g., 2.5 sigma for 99% confidence level. However, different from traditional reserves, the ramp product is used to manage both net load variations (foreseeable changes) and uncertainties (unforeseeable changes) and is regularly deployed through economic dispatches every five minutes. The design of the product can thus be subtle. We analyze and design the ramp capability product by formulating ramp requirement constraints in a reliable way and minimizing the cost through simulation-based optimization.