Abstract

This thesis creates a model that serves as an evaluation tool for testing decarbonization strategies on energy-related projects at Minnesota State University, Mankato, by estimating Energy Use Intensity for different types of spaces across campus. These EUIs can be used to predict future energy consumption when space usage changes and also serve as a baseline for dynamic energy modeling. The methodology used was a regression-based spreadsheet model proposed by Dr. Shreshth N. and originally applied at the Massachusetts Institute of Technology. EUIs were estimated across five different scenarios, and for electricity, steam, and chilled water. Three criteria were used to evaluate scenarios: model performance, validation of regression assumptions, and agreement with campus-level data. Scenario 1 produces the best estimates, but the results were incomplete; therefore, the next best options selected were scenario 3 for electricity and chilled water, and scenario “1 on 2” for steam. This model showed a better fit when evaluated at the building level compared to the campus level. At the building level, the maximum percentage error between measured and observed energy use was 3%. At the campus level, the errors for electricity and steam were higher but still below 15%, which falls within the recommended threshold. In contrast, the error for chilled water was 53%. These errors mainly resulted from the small dataset size and inconsistent metering. To improve this, future work should focus on improving residual normalization and testing non-linear regression to explore more complex energy consumption behaviors.

Advisor

Patrick Tebbe

Committee Member

Dennis Soltis

Committee Member

Namyong Lee

Date of Degree

2025

Language

english

Document Type

Thesis

Degree

Master of Science (MS)

Program of Study

Mechanical Engineering

Department

Mechanical and Civil Engineering

College

Science, Engineering and Technology

Download
Request Accessible Version

Share

COinS
 

Rights Statement

In Copyright