The report presents a methodology for estimating air-conditioning demand and gross solar PV generation output from half-hourly aggregated demand data. The methodology draws on Bayesian inference approaches, net aggregate load data, and data on environmental conditions to estimate the parameters of coarse thermostatically controlled load or solar PV models. Those models can then be used to estimate total thermostatically controlled (predominantly air-conditioner) load based only on temperature data or gross solar PV output based only on temperature and solar irradiance data. Preliminary validation of heating and cooling estimates are provided, using circuit-level metering data in Brisbane. Similarly, preliminary validation of gross PV estimates are provided, drawing on circuit-level metering data in Melbourne.

Initial results show air-conditioning demand and model parameters for a small set of aggregate loads in Queensland, New South Wales and Victoria. These results are complemented by preliminary findings for gross PV load for a single zone substation in Victoria.

Note that the general approach documented here has been deployed to provide estimates of air-conditioning demand across Australia. The approach to gross PV estimation has also been improved subsequent to this publication. Please refer to the Related Resources listing on this page for relevant reporting stemming from that more contemporary set of work.

Note that all results reflect estimates only.