An Inverse Model to Predict and Evaluate the Energy Performance of Large Commercial and Institutional Buildings

1 Present address: Energy Systems Laboratory, Texas A&M University, College Station, TX 77843-3123, USA. ABSTRACT A new method for predicting and evaluating the energy performance of large commercial and institutional buildings is developed, as an alternative to using existing comprehensive energy simulation programs like DOE-2 and BLAST, or oversimplified tools like analyzing monthly utility bills. The steps are convenient and practical in their potential use by the energy analysts. The method mainly groups the Stepwise Multiple Linear Regression (SMLR) to empirically model the dynamic building thermal performance; Fourier Series to predict (extrapolate) internal loads; the Monte-Carlo Simulation to deal with the prediction of internal loads probabilistically; and a new approach in normalizing weather conditions. The method is capable of predicting the building energy performance in the pre-retrofit phase, identifies appropriate energy conservation measures, and estimates potential energy savings.