A Framework for Improving the Cost-Effectiveness of DSM Program Evaluations

The prudence of utility demand-side management (DSM) investments hinges on their performance, yet evaluating performance is complicated because the energy saved by DSM programs can never be observed directly but only inferred. This study frames and begins to answer the following questions:

1. How well do current evaluation methods perform in improving our confidence in the measurement of energy savings produced by DSM programs?
2. In view of this performance, how can we best allocate limited evaluation resources to maximize the value of the information they provide?

We review three major classes of methods for estimating annual energy savings: tracking database (sometimes called engineering estimates), end-use metering, and billing analysis and examine them in light of the uncertainties in current estimates of DSM program measure lifetimes. We assess the accuracy and precision of each method and construct trade-off curves to examine the costs of increases in accuracy or precision. We demonstrate several approaches for improving evaluations for the purpose of assessing program cost effectiveness. The methods can be easily generalized to other evaluation objectives, such as shared savings incentive payments.