The objective of this paper is to present a methodology and pertinent equations for performing an engineering uncertainty analysis on the savings due to a particular energy conservation measure (ECM) as compared to a baseline system. Though this study is geared toward building cooling systems, the methodology can be applied with minor modification to most building measurement and verification programs that involve performing measurements of the system, identifying a regression model, and using the model to extrapolate for future behavior once the ECM has been implemented. The methodology covers the case in which a nested model is used, as when shortterm data are used to develop a model for building secondary thermal loads, which is then used to drive a model for chiller electricity use. Hence, the methodology treats measurement errors, model internal prediction uncertainty, and model extrapolation bias uncertainty in the framework of a nested model approach. A notable feature of this paper is that a nomograph, consisting of six separate interlinked graphs, has been generated based on the equations presented herein, whereby a user can graphically determine the final uncertainty in savings by selecting appropriate values of the various sources of uncertainty. The use of this nomograph is explained by means of an example.

Units: Dual