Energy efficiency is currently a hot topic on all regional, national, and global stages. Accurate measurements on how energy is being used over a period of time can improve performance of the drinking water utility substantially and reduce energy consumption. Nevertheless, the drinking water industry does not have a specific benchmarking practice to evaluate its energy performance of the system. Therefore, there are no standards to compare energy use between water utilities that have a variety of system characteristics. The goal of this research is to develop quantitative approach to select energy benchmarking parameters of the water system, so the drinking water utilities can use those parameters to improve their energy efficiency. In addition to a typical benchmarking of drinking water utilities, the energy benchmarking can specifically compare energy efficiency of a utility with other utilities nationwide.

The research developed a regression model based on the statistical representation of the energy use and descriptive characteristics of the drinking water utilities data throughout the U.S. Methodologies to eliminate singularity and multicollinearity from collinear survey dataset are discussed. The all possible regressions were chosen as parameters selection methodology to identify a subset of most significant parameters, i.e. system characteristics, that can mathematically correspond to energy use across different utilities. As a result, the energy benchmarking would be able to calculate the predicted total energy use of the system from given system characteristics.