FORECASTING METHODOLOGY INFORMATION PAPER - 2016 NATIONAL ELECTRICITY FORECASTING REPORT - Published: July 2016

To better understand the key drivers of consumption by such customers, and accommodate their outlook in the forecast, AEMO surveyed and interviewed the largest industrial consumers in the NEM.27 The structured survey approach helped AEMO to understand how these customers would be likely to respond to changing market conditions (such as changing electricity prices), and to analyse the potential eff. [...] The impact of appliance fuel switching was reported in AEMO’s 2015 NGFR42 and the approach is discussed in the 2015 NGFR Methodology Information Paper.43 The following adjustments were made to convert the loss in gas load to a forecast gain in electricity consumption: 50% of the forecast reduction in gas consumption attributed to gas hot water heating was assumed. [...] The forecast demand components, PV generation, and battery charging were then assembled together to produce numerous 20-year, half-hourly profiles of delivered demand.48 In the last stage of the analysis, the maxima and minima of the delivered demand were extracted from each simulated half-hourly pattern and used to build probability distributions. [...] The model expressed the relationship between demand and predictor variables related to the day of the year (including whether it is a public holiday or not) and the weather standard. [...] While the normalised PV generation used to estimate actual values was tuned on the geographic distribution of rooftop PV installations at the historical time, in the simulation stage of the MD analysis the PV generation was reweighted, to better represent the current distribution of the panels.