An evaluation of the ORNL residential energy use model

Abstract

This report provides an evaluation of the architecture, empirical foundation, and applications of the Oak Ridge National Laboratory (ORNL) residential energy use model. A particular effort is made to identify the strengths and shortcomings of the model for alternative uses, and to identify areas where model structure and empirical support could be upgraded. Concrete suggestions are made for improvements in model logic, strengthening the empirical basis for behavioral and technical parameters, and reducing the biases in the model arising from aggregation. The overall conclusion is that the model has the potential to provide adequate forecasts of the aggregate impacts at a regional or national level of policies whose effects on households are relatively homogeneous. There are a number of model changes which would be relatively easy to implement, and which should substantially improve forecasts of this sort. On the other hand, the aggregate architecture of the ORNL model makes it fundamentally unsuitable for applications to geographical areas smaller than DOE regions, or to policies which have a heterogeneous impact on households.