Skip to content Skip to navigation

A Method to Analyze Large Data Sets of Residential Electricity Consumption to Inform Data-Driven Energy Efficiency

TitleA Method to Analyze Large Data Sets of Residential Electricity Consumption to Inform Data-Driven Energy Efficiency
Publication TypeWorking Paper
Year of Publication2012
AuthorsKavousian, A, Rajagopal, R, Fischer, M
Date Published06/2012
Publication Languageeng
KeywordsCenter for Integrated Facility Engineering, CIFE, Data-driven energy efficiency, Energy Efficiency, Modeling, Residential electricity consumption modeling, Smart meter data analysis, Stanford University, Structural and behavioral determinants of energy consumption
AbstractEffective demand-side energy efficiency policies are needed to reduce residential electricity consumption and its harmful effects on the environment. The first step to devise such polices is to quantify the potential for energy efficiency by analyzing the factors that impact consumption. This paper proposes a novel approach to analyze large data sets of residential electricity consumption to derive insights for policy making and energy efficiency programming. In this method, underlying behavioral determinants that impact residential electricity consumption are identified using Factor Analysis. A distinction is made between long-term and short-term determinants of consumption by developing separate models for daily maximum and daily minimum consumption and analyzing their differences. Finally, the set of determinants are ranked by their impact on electricity consumption, using a stepwise regression model. This approach is then applied on a large data set of smart meter data and household information as a case example. The results of the models show that weather, location, floor area, and number of refrigerators are the most significant determinants of daily minimum (or idle) electricity consumption in residential buildings, while location, floor area, number of occupants, occupancy rate, and use of electric water heater are the most significant factors in explaining daily maximum (peak) consumption. The results of the models are compared with those of previous studies, and the policy implications of the results are discussed.
PDF Link
Citation Key1116