Abating Carbon Dioxide Emissions from Electric Power Generation: Model Uncertainty and Regulatory Epistemology

Computational modeling of natural, economic, and technological systems is a primary analytical methodology in US energy and environmental regulation. Validating or otherwise evaluating such models and analyzing the uncertainties involved in their regulatory applications have become both more important and more challenging. This paper reviews these issues in the context of an important recent example involving energy, the US Environmental Protection Agency's (EPA's) development of regulations to reduce carbon dioxide emissions from electric power plants using a numerical model of the US electric power system. Following a summary of background information about greenhouse gas abatement policy, the paper discusses the agency's general computational model evaluation philosophy; the history of, and current practices in, energy model evaluation; the specific model used by the EPA and its application to carbon dioxide regulation; and the concept of fundamental model uncertainty and its significance for this modeling domain.