Document Type

Conference Paper

Publication Date


Publication Source

Proceedings of the Industrial Energy Technology Conference


Energy in manufacturing facilities is used for direct production of goods, space conditioning, and general facility support such as lighting. This paper presents a methodology for statistically analyzing plant energy use in terms of these major end uses. The methodology uses as few as 60 data points that are relatively easy for most plants to obtain. Multivariable change-point models of electricity and natural gas use as functions of outdoor air temperature and production data are then developed. The statistical models can be used to predict energy use for energy budgeting, measure savings, determine cost structures, and diagnostic purposes. Moreover, in many cases, the statistical models are able to subdivide plant energy use into facility, space-conditioning and production-related components. These breakdowns suggest the savings potential from reducing non-production and space-conditioning energy use. A detailed case study example is used to demonstrate the method and discuss interpretations of the results.

Document Version

Published Version


This document is provided for download by permission of the publisher. Permission documentation is on file.


Texas A&M University

Place of Publication

Houston, TX