Masters Theses

Date of Award

5-1997

Degree Type

Thesis

Degree Name

Master of Science

Major

Engineering Science

Major Professor

Richard Jendrucko

Committee Members

Esteban E. Walker, William S. Johnson

Abstract

In this study, a large quantity of industrial manufacturers' energy consumption and waste generation data compiled by the University of Tennessee Industrial Assessment Center was analyzed for inherent trends and relationships. This effort first required the amassing of the energy and waste data generated through over 20 years of service to regional manufacturing facilities, then organization of the amassed data into an onsite database. Previous efforts to compile this type of data for statistical analysis provided some preliminary insights, but the earlier investigatory analyses did not reach the depth of analysis this work undertook in categorizing and analyzing different types of information as well as examining the data for relationships, including nonlinear functions. Specifically, the objectives of this work were to 1) create the database to facilitate easier access to and analysis of the UTIAC-generated data and 2) perform several different types of analyses in search of trends and relationships within the data.

The characterization analyses performed on the energy and waste data revealed some notable information and trends. The two most interesting included the following. The first is that the implementation rate of the recommended measures (both energy and waste related) varies moderately from state to state in which the assessments were performed, with Tennessee and South Carolina clients reporting an implementation rate of approximately 50% and Kentucky, North Carolina and Virginia clients reporting an implementation rate of approximately 60%. The second is that the implementation rate for all recommendations increases significantly and almost linearly with an increase in the number of energy recommendations made per report up to 9 (from 0% at 1 recommendation per report to about 60% at 9).

The simple linear, curvilinear and multiple linear regression analyses performed on the data were limited to a sampling of the analyses that could be done, simply because of the sheer quantity of program data amassed. However, five investigatory questions were propounded with the overriding theme of which factors most strongly relate to plant energy consumption, waste generation and recommendation implementation rate. In order as posed, the results yielded the following.

The analysis of question #1 revealed a strong, positive relationship for all SIC groups between annual sales and annual electrical consumption, with five specific SIC groups also showing moderate to highly significant, similar relationships. Question #2's analysis exposed a modest, curvilinear relationship for all SIC groups between annual sales and the UTIAC- recommended annual energy cost savings, along with several, modest SIC- specific linear and curvilinear relationships between these variables. The investigation of question #3 revealed a mediocre, positive relationship between annual production hours and annual energy consumption for all industry types, with the same results for six specific SIC groups. Question #4's analysis showed that there is practically no relationship between the number of recommendations made per report and the total UTIAC-recommended energy conservation or waste generation quantities, outside of some modest correlations within three specific SIC groups. Lastly, the investigation of question #5 exposed that, for all industrial manufacturing types as well as all the specific ones, there is almost no association between implementation rate and either the capital cost of implementation or payback period.

Most certainly, these findings show that there are significant, inherent relationships within the various types of the industrial data compiled, and the thorough analysis of this data can provide valuable insights into the factors affecting plant energy consumption, waste generation and recommendation implementation. In addition, the results can be used to prioritize future program work for maximum client and program benefit.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS