Measuring plant level energy efficiency when production activities are not homogeneous: The ENERGY STAR energy performance indicator


Journal Article

Energy efficiency is the comparison of actual energy used to the lowest amount of energy required to perform some level of service or produce some amount of product. Energy intensity, also called specific energy consumption (SEC), is the simple normalization of energy use to a measure of service or production. This energy intensity can then be compared to the "best practice." This simple description belies the complexities that lie in industrial activity. When there is no single measure of activity for the denominator, the energy intensity may not be well defined. There is a substantial literature that examines what the "best measure" of production to use for various sectors; physical volumes, market value, value-added, etc. Creating aggregate measures of activity, e.g. total dollars, is often an inadequate approach. When sectors are aggregated, the changing mix of production in various sectors influences the aggregate energy intensity index. The difference between the observed and "best practice" is the level of (in)efficiency. The estimate of "best practice" can be an optimum configuration derived from engineering, possibly a notion of the theoretical minimum, or comparison to a real world application with the lowest observed energy intensity, i.e. "best observed practice." Since production/activity mix can be quite varied, it may be difficult to find a real world application that is sufficiently similar to the observed plant to conduct the needed comparison. This reflects an oft-expressed view that "every plant is unique." A statistical model which provides a functional relationship between the level of energy use and the level and type of various production activities, material inputs quality, and external factors, e.g. climate, is one solution to these problems. The stochastic frontier regression estimates the lowest observed plant energy use, given these factors. This statistical model also provides a distribution of (in)efficiency across the industry, which allows the user to answer the hypothetical but very practical question, "How would my plant compare to everyone else in my industry, if all other plants were similar to mine?" The result is a tool that can be used by corporate and plant energy managers to estimate the energy efficiency of their portfolio of plants. This paper presents recent empirical results prepared for ENERGY STAR for corn refining. This sector typifies the issue of product mix impact on plant energy use. Accounting for these factors allows construction of a normalized distribution of energy efficiency for the entire sector and a percentile performance score, for any individual plant. The results of the analysis can be used by energy managers in these various sectors to assess plant energy performance, set targets, and track progress. © 2005 ACEEE Summer Study on Energy Efficiency in Industry.

Duke Authors

Cited Authors

  • Boyd, GA

Published Date

  • December 1, 2005

Published In

  • Proceedings Aceee Summer Study on Energy Efficiency in Industry

Citation Source

  • Scopus