Statistical Analysis of Historical State-Level Residential Energy Consumption Trends

David B. Belzer and Katherine A. Cort, Pacific Northwest National Laboratory

Keywords

Abstract

Obtaining an accurate picture of the major trends in energy consumption in the nation’s stock of residential buildings can serve a variety of national and regional program planning and policy needs related to energy use. This paper employs regression analysis and uses a variable degree-day approach with historical data to provide some insight into overall changes in the thermal integrity of the residential building stock by state.

Although national energy use intensity estimates exist in aggregate, these numbers shed little light on what drives building consumption, as opposing influences are hidden within the measurement (e.g., households may contain more appliances that increase energy use while shell improvements reduce it). This study addresses this issue by estimating changes in the reference temperatures that best characterize the existing residential building stock on a state basis. Improvements in building thermal integrity are reflected by declines in the heating reference temperature, holding other factors constant. Heating degree-day estimates to various reference temperatures were computed from monthly average temperature data for approximately 350 climatic divisions in the U.S. A simple cross-sectional analysis is employed to try to explain the differential impacts across states. Among other factors, this analysis considers the impact that the relative growth in the number of residential buildings has had on natural gas consumption in residential buildings. This paper describes the methodology used, presents results, and suggests directions for future research.