To effectively plan and execute energy efficiency programs, a
sound understanding of the determinants that drive household
electricity consumption (such as floor area, average outside temperature,
and number of occupants) is needed [5]. However,
because of lack of easily-accessible, high-resolution consumption
data, underlying determinants of energy use and energy-related
behaviors have not been extensively examined before [6].
With growing deployment of smart meters and real-time home
energy-monitoring services, data that allow studying such underlying
determinants are becoming available (for examples of studies
using high-resolution consumption data, see Refs. [7e9]). However,
the methodologies to analyze the data and infer the results that can
be used to support decision making at the household level have not
yet been formalized [6].
To address that gap, this paper proposes a methodology to
analyze large data sets of residential electricity consumption to
derive insights for policy making and energy efficiency programming.
In particular, it offers a method to disaggregate the impact of
structural determinants (e.g., insulation level of the residence) from