Analysis of Energy Consumption of Duplex Residences in College Station, Texas

Abstract

This paper characterizes the variability of energy consumption due to a series of construction, occupant, and weather-related effects in duplex residences in College Station, Texas. In this paper, spline regression was used to estimate cooling efficiency, heating efficiency, base load, modified cooling balance point temperature, and modified heating balance point temperature for monthly billed kWh against average daily outside temperature. These estimates were used to predict normalized annual consumption. Best subsets regression and multiple regression were used to explore the relationship between energy consumption and construction, occupant, and weather-related factors. The sample for this paper was I40 duplex residences which used only electricity for cooling and heating, and had one year minimum occupancy in College Station, Texas. The spline regression models with weather-related factors achieved adjusted R~ values averaging 0.82. Construction, occupant, and weather-related factors were determined to be components of energy consumption. In the final modeling, construction, occupant, and weather-related factors accounted for 93% of the variance for the normalized annual consumption of duplex residences. The findings showed there was a significant relationship between normalized annual consumption and year built, thermostat setting, cooling efficiency, heating efficiency, base load, modified cooling balance point temperature, and modified heating balance point temperature.