Towards Objective Evaluation of the Accuracy of Marginal Emissions Factors

Abstract

A growing number of policymakers and voluntary actors seek to reduce CO2 or other emissions by intentionally changing the timing of flexible electricity load. In practice, this requires selecting a model to estimate what is often called the marginal emissions factor (MEF). Multiple different models to estimate MEFs exist, but there is no single objective ground truth MEF dataset with which to compare the relative accuracy of two or more such models. Thus, historically MEF validation has been forced to use subjective tests which rely on the very models they evaluate. Here we propose a technique to instead gain insight into the relative accuracy of different MEF models without relying on the assumptions of any MEF models. It works by isolating indirect model claims that can be objectively tested. This method also requires first applying a taxonomy to clarify the objectives of different types of MEF models, so that accuracy is rigorously defined. We then demonstrate the application of this technique in a comparative analysis of a few sample MEF models in the US, using 2021 data.

Keywords: demand side management (DSM), economic dispatch, carbon optimization, marginal emissions rates, causal modeling, Life cycle assessment, carbon accounting, renewable energy siting