Train, K.E. (2003), Discrete Choice Methods with Simulation, Cambridge University
Press, Cambridge.
Appendix: Data assembly
Our data is drawn from a sample of 2530 single-family home owners, surveyed
in 2005 as part of the German Residential Energy Consumption Survey. The
data contain a location identifier for each household, which is measured at the
municipal level. The data additionally contain socioeconomic and dwelling char-
acteristics, including whether the household received an energy audit and which
retrofit measure was implemented within the last 10 years, if any. Four different
retrofit measures (and their combinations) are surveyed: roof insulation, facade
insulation, windows replacement, and heating-equipment replacement.
4.1 Energy Savings
The computation of energy savings are based on engineering relationships and are
measured as the decline of the building’s annual primary energy demand following
a retrofit. We first reconstruct the size of the building shell using computer
aided design. This reconstruction, which combines information on the area of
living space, the number of stories, and simplifying assumptions concerning the
building form, allows us to derive the extent of the heat-transmitting surface
and the required heating power. Following the relationships provided by the
respective technical standards set by the German Institute for Standardization,
the demand for primary energy can be expressed as:
(11) Q = (qh (Hτ) + qw) * ep,
where Q is the building’s primary energy demand, QH is the demand for space
heating, and QW is the energy demand for hot water, all under standardized
26