Development and analysis of residential change-point models from smart meter databy Perez, KX; Cetin, K; Baldea, M; Edgar, TF
As access to residential energy use data becomes more widely available, it is possible to identify significant energy consumers and provide guidance on mitigating such large loads. In hotter climates, such as Texas, air-conditioning (AC) systems are important contributors to overall residential electricity demand. Providing a quick, simple and effective framework to describe and compare electricity demand patterns between different hotises is valuable to identify potential candidates fot peak load reduction and overall energy use mitigation. In this study, we evaluate the application of daily change-point models to describe the demand patterns of residential AC systems for 45 actual houses in Austin, TX during 2013. While previous research regarding change -point models has been focused on monthly data for commercial buildings, this study extends its application to daily residential energy use. The resulting models describe a behavior where energy consumption with relation to outdoor dry-bulb temperature is negligible up until a change -point, after which AC energy use increases linearly and results in an “energy slope.” An analysis of the neighborhood shows the distribution of the AC “energy slopes” is left-skewed and centered on 0.08 kW per degrees C dry bulb temperature. Energy audit information found eight house characteristics to be correlated with a higher energy slope. A subsequent parametric analysis using data from the energy simulation software BEopt confirmed the direction of the correlation. This work provides a screening tool to compare energy demand patterns of houses and target houses with the largest magnitude of energy slopes for future energy audits. (C)2016 Published by Elsevier B.V.