Estimates of rooftop PV capacity, by region and postcode, is publicly available in Australia. However, data on
individual households’ rooftop photovoltaic (PV) capacity is not publicly available. This is valuable in price
comparison and research, for example analysing the impact of rooftop solar in wholesale markets and on networks.
We develop a model to estimate an individual household’s rooftop PV capacity, using data on the household’s
estimated annual rooftop PV exports to the grid and its volume of annual grid electricity purchases. The model
relies on simulated data of hypothetical rooftop PV systems, which are then used to estimate relationships between
the variables. The model was found to reliably predict PV capacity in a test of 124 households where PV capacity
was known. The model is useful in applications that require estimation of the relationship between annual
measures of household grid electricity purchases, rooftop PV exports and rooftop PV capacity. This includes
research into the impact of rooftop PV on wholesale market and network charges, and electricity price comparison.
The model development and testing approach used here can be replicated in other locations but data on rooftop
PV capacity for households in those locations will be needed for model verification.