This thesis investigates the implication of empirical grounded electric power market facts using multiple methodologies, including market and contract design, analytic method, statistical method, and agent-based simulation method. Basically, this thesis focuses on centrally-managed electric power markets.

European and U.S. electricity sectors have undergone substantial restructuring over the past twenty years. They have devolved from highly regulated systems operated by vertically integrated utilities to relatively decentralized systems based more fully on market valuation and allocation mechanisms.

These restructuring efforts have been driven by a desire to ensure efficient energy production and utilization, reliable energy supplies, affordable energy prices, and effective rules and regulations for environmental protection. In keeping with the latter goal, a dramatic change is taking place in energy mixes: namely, a rapid penetration of variable energy resources combined with a movement away from traditional thermal generation.

Variable energy resources (VERs) are renewable energy resources, such as wind and solar power, whose generation cannot be closely controlled to match changes in load or to meet other system requirements. Consequently, the integration of VERs tends to increase the volatility of net load (ie, load minus as-available generation) as well as the frequency of strong ramp events. Flexibility in service provision by other types of resources then becomes increasingly important to maintain the reliability and efficiency of power system operations.

To accommodate increased VER penetration, TSOs and ISOs have introduced major changes in their market rules and operational procedures. These changes have included new product definitions to enhance load-following capability (eg, ramping products), revised market eligibility requirements to encourage greater VER participation, and the introduction of capacity markets in an attempt to ensure sufficient thermal generation as a backstop for the intermittency of VER generation.

Also, CO2 emission issues are increasing important in electric power markets. In the U.S., the largest source of CO2 emissions is the electricity sector, which was responsible for 32% of total emissions in 2012. The Obama Administration proposed a Clean Power Plan in June 2014; nationwide, by 2030, this plan would achieve approximately 30 percent of CO2 emission reduction relative to 2005 CO2 emission levels in the power sector. There are several important issues arising from carbon mitigation options such as a carbon tax imposition and increase penetration of VERs need to be resolved.

Chapter 2 introduces standardized energy and reserve contracts with swing (flexibility) in their contractual terms to resolve key issues that have arisen for centrally-managed wholesale electric power markets with increased penetration of renewable energy resources. Concrete examples are used to demonstrate how the trading of these standardized contracts can be supported by linked forward markets in a manner that permits efficient real-time balancing of net load subject to system and reserve-requirement constraints. Comparisons with existing wholesale electric power markets are given, and key policy implications are highlighted.

Chapter 3 extends the system pattern short-term forecasting method for power systems to incorporate non-dispatchable renewable energy, thus permitting the forecasting of CO2 emissions along with the forecasting of prices, line congestion, and other system variables. It also develops an empirically-based system pattern transition matrix permitting a dynamic extension of the method. The practical usefulness of the resulting extended forecasting method is illustrated by means of a 5-bus test system based on data from the Midcontinent Independent System Operator (MISO).

Chapter 4 develops a 9-zone test system based on MISO data for application in electric power market studies. This 9-zone test system models MISO's original seven midwestern zones together with two recently-incorporated southern zones operating over a 15-line AC transmission grid. Generators with different fuel types and capacities submit hourly supply offers in a day-ahead market (DAM). Load-serving entities submit hourly demand bids into a day-ahead market in the form of a 24-hour regional load profile. This 9-zone test system allows a wide range of sensitivity studies. To illustrate the capabilities of the 9-zone test system, this study undertakes a comparative study of DAM Locational Marginal Price (LMP) outcomes for MISO prior to and after the integration of the two southern zones by conducting test-bed simulations for 7-zone and 9-zone test cases based on MISO data.

Chapter 5 analyzes how the imposition of a carbon tax and the increased penetration of wind power in such markets could impact CO2 emissions and other key outcomes, such as energy dispatch, energy prices, market participant profits (by fuel type), and government tax revenues. Another innovation of this part is that the effects of increases in a carbon tax and wind power penetration are studied jointly. It is shown, for example, that CO2 emissions decrease from 0.23% to 6.17% as the carbon tax and the degree of wind penetration are systematically varied from a base case of zero tax and zero wind. The profits of coal- and oil-fired generation systematically decrease with increases in the carbon tax and/or increases in wind penetration, but the profits of other types of generation exhibit a more complex response. Comparisons with current MISO conditions are also given, and key policy implications are discussed.