2939 -- Regression Based Process Energy Analysis System

The results of an investigation are presented to determine which weather, production, and time-related parameters exert significant influence on installation energy consumption for the U.S. Army Armament, Munitions and Chemical Command (AMCCOM) using regression analysis methods . Based on data gathered at AMCCOM HQ, potentially significant weather and production/mission parameters are identified, and Process Energy Analysis Systems are developed for each installation using regression analysis methods on a monthly data base for the period FY75 through FY82 (October 1974 through September 1982). The regression model for AMCCOM shows that aggregate energy consumption in general depends on heating degree-days, production level, and labor force strength. At individual installations, additional important parameters include cooling degree-days and facility changes over time. The model was applied to actual FY84 data and predicted total energy consumption to within 3% to 6% of actual consumption. Results of this effort will be used to forecast energy consumption and establish energy guidelines throughout AMCCOM.

Units: I-P