Literature Review of Uncertainty of Analysis Methods (Inverse Model Toolkit), Report to the Texas Commission on Environmental Quality

Abstract

This report reviews the reported uncertainty of ASHRAE’s Inverse Model Toolkit (IMT) analysis method and the linear, and change-point linear algorithms that it uses by reviewing the published literature on the related accuracy of IMT and its algorithms versus other well-accepted statistical analysis tools, such as SAS. This report begins with a review of the history of the IMT, and the linear and change-point linear models. Then it reviews the published comparisons of the IMT and other analysis software, relying heavily on the accuracy testing that was performed as part of ASHRAE’s Research Project 1050-RP. It also includes a detailed description of the basic algorithms and an example of the IMT weather-normalization analysis. In summary, from the literature it was found that the algorithms in the IMT almost exactly reproduce the same regression analysis one would get by running any one of the programs that it was compared against (i.e., usually to several significant digits). Therefore, it can be concluded that the IMT is accurate, when it is called upon to perform weather normalized regressions for modeling building energy use.