A Data-Driven Modeling Approach for Commercial Building Energy Savings in the Midwest Using Change-Point Regression Analysis

Author

• Patrick Daniel McFarland (0009-0001-7605-9039), University of Dayton

Date of Award

5-9-2026

Degree Name

M.S. in Renewable and Clean Energy

Department

Department of Mechanical and Aerospace Engineering Graduate Renewable and Clean Energy Engineering

Advisor/Chair

Andrew Chiasson

Abstract

Commercial buildings account for a significant portion of total the United States energy consumption which makes them a primary candidate for energy reduction strategies and implementation. The challenge is being able to accurately predict how applying energy conservation measures (ECMs) affects building health and performance. Complex and detailed simulations require large amounts of data analysis, modeling skills, and extensive knowledge of ECM implementation to obtain reliable energy saving results. EnergyPlus and OpenStudio are used to simulate a medium-sized office building in the Midwest and apply three- and five-parameter change-point regression models to evaluate the impact of energy conservation measures under varying weather conditions. By combining detailed simulations with simplified regression modeling, the work aims to provide a practical approach for building managers to quantify energy savings and support informed decision-making in commercial energy efficiency projects. The model uses TMY3 and Design Day (DDY) weather files to capture both typical and extreme climate conditions for accurate annual performance evaluation. Several OpenStudio measures are applied to generate simulation outputs and export meter data for regression analysis. Python scripts are utilized as data-processing tools to condense meter data into digestible excel summaries and produce change-point regression plots. The change-point regression models demonstrate how the ECMs have a positive impact on load reduction and energy savings.

Keywords

Energy, Engineering, Environmental Engineering, Mechanical Engineering, Sustainability

Comments

OCLC No. 1591830136

Rights Statement

Copyright 2026, author.

Recommended Citation

McFarland, Patrick Daniel, "A Data-Driven Modeling Approach for Commercial Building Energy Savings in the Midwest Using Change-Point Regression Analysis" (2026). Graduate Theses and Dissertations. 7642.
https://ecommons.udayton.edu/graduate_theses/7642