Activities-specific disaggregation of hourly residential electricity consumption and smart household appliances

We analyze aggregated historical data that give habitual residential electricity consumption for a middle size county in the Southeastern United States. Habitual consumption is the consumption which does not depend on weather conditions. It can be represented as a sum of habitual consumptions. Each of the summands is a consumption corresponding to a habit, i.e., corresponding to a group of correlated activities that involve a specific group of appliances. Groups of the activities related to different summands are chronologically independent. Analysis of aggregated historical data reveals this structure of the habitual consumption. We conclude that measures that support chronological independence of habitual activities are important for demand-side management. Among these measures are automation and recombination (into new habits) of the activities through an updated stock of smart appliances.