Decision Support Algorithm for Evaluating Carbon Dioxide Emissions from Electricity Generation in the United States

This article presents an algorithm to aid practitioners in determining the most appropriate method to estimate carbon dioxide emissions from an electricity load. Applications include sustainability assessments of products, processes, energy efficiency improvements, changes in generation infrastructure, and changes in electricity demand. Currently, there is no consensus on appropriate methods for calculating greenhouse gas emissions resulting from specific electricity loads. Previous research revealed significant differences in emissions when different methods were used, a situation that could result in divergent sustainability or policy recommendations. In this article, we illustrate the distribution of emissions estimates based on method characteristics such as region size, temporal resolution, average or marginal approaches, and time scales. Informed by these findings, a decision support algorithm is presented that uses a load's key features and an analyst's research question to provide recommendations on appropriate method types. We defined four different cases to demonstrate the utility of the algorithm and to illustrate the variability of methods used in previous studies. Prior research often employed simplifying assumptions, which, in some cases, can result in electricity being allocated to the incorrect generating resources and improper calculation of emissions. This algorithm could reduce inappropriate allocation, variability in assumptions, and increase appropriateness of electricity emissions estimates.