Date of Award


Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Engineering and Public Policy


Ines Lima Azevedo

Second Advisor

M. Granger Morgan

Third Advisor

Jay Apt

Fourth Advisor

Dean Kamen


There is growing interest in reducing the environmental and human-­‐health impacts resulting from electricity generation. Renewable energy, energy efficiency, and energy conservation are all commonly suggested solutions. Such interventions may provide health and environmental benefits by displacing emissions from conventional power plants. However, the generation mix varies considerably from region to region and emissions vary by the type and age of a generator. Thus, the benefits of an intervention will depend on the specific generators that are displaced, which vary depending on the timing and location of the intervention.

Marginal emissions factors (MEFs) give a consistent measure of the avoided emissions per megawatt-­‐hour of displaced electricity, which can be used to evaluate the change in emissions resulting from a variety of interventions. This thesis presents the first systematic calculation of MEFs for the U.S. electricity system. Using regressions of hourly generation and emissions data from 2006 through 2011, I estimate regional MEFs for CO2, NOx, and SO2, as well as the share of marginal generation from coal-­‐, gas-­‐, and oil-­‐fired generators. This work highlights significant regional differences in the emissions benefits of displacing a unit of electricity: compared to the West, displacing one megawatt-­‐hour of electricity in the Midwest is expected to avoid roughly 70% more CO2, 12 times more SO2, and 3 times more NOx emissions.

I go on to explore regional variations in the performance of wind turbines and solar panels, where performance is measured relative to three objectives: energy production, avoided CO2 emissions, and avoided health and environmental damages from criteria pollutants. For 22 regions of the United States, I use regressions of historic emissions and generation data to estimate marginal impact factors, a measure of the avoided health and environmental damages per megawatt-­‐ hour of displaced electricity. Marginal impact factors are used to evaluate the effects of an additional wind turbine or solar panel in the U.S. electricity system. I find that the most attractive sites for renewables depend strongly on one’s objective. A solar panel in Iowa displaces 20% more CO2 emissions than a panel in Arizona, though energy production from the Iowa panel is 25% less. Similarly, despite a modest wind resource, a wind turbine in West Virginia is expected to displace 7 times more health and environmental damages than a wind turbine in Oklahoma.

Finally, I shift focus and explore the economics of small-­‐scale cogeneration, which has long been recognized as a more efficient alternative to central-­‐station power. Although the benefits of distributed cogeneration are widely cited, adoption has been slow in the U.S. Adoption could be encouraged by making cogeneration more economically attractive, either by increasing the expected returns or decreasing the risks of such investments. I present a case study of a 300-­‐kilowatt cogeneration unit and evaluate the expected returns from: demand response, capacity markets, regulation markets, accelerated depreciation, a price on CO2 emissions, and net metering. In addition, I explore the effectiveness of feed-­‐in tariffs at mitigating the energy-­‐price risks to cogeneration projects.

Included in

Engineering Commons