Statistical modelling in energy efficiency benchmarking

Benchmarking is an important tool to promote the efficient use of energy in commercial buildings. Benchmarking models are mostly constructed in a percentile table, which is relevant to energy use when normalized only with floor area and temperature. We describe a multiple regression approach for benchmarking the energy efficiency through the relationship between energy use intensities (EUIs) and the explanatory factors. These EUIs are then normalized by removing the effect of deviances in the significant explanatory factors. The normalized EUIs give an empirical cumulative distribution of the normalized EUI and the percentile table could be used for benchmarking the energy efficiency of a commercial unit. A practical study on commercial buildings is presented to illustrate the development and the use of the energy efficiency benchmarking method.