Load forecasting in pre-dispatch and STPASA
AEMO must produce load forecasts for each region for the following timeframes:
- Each day for the day ahead – pre-dispatch forecast
- Each day for the period two to seven days ahead – short term projected assessment of system adequacy (STPASA) forecast
These forecasts are produced in accordance with the Spot Market Operations Timetable.
Eight calendar days are required to fully cover the seven trading days of the STPASA forecast period.
The Demand Forecasting System (DFS) generates the 10%, 50% and 90% Probability of Exceedance (POE) load forecasts used in pre-dispatch and STPASA. The DFS generates load forecasts automatically every half-hour and, under normal circumstances, does not require manual intervention.
The key inputs into the DFS are:
- Historical actual metered loads
- Real-time actual metered loads (SCADA data from immediately preceding intervals)
- Historical and Forecast Weather data (temperature and humidity)
- Non-scheduled wind generation forecasts (from the Australian Wind Energy Forecasting System (AWEFS))
- Rooftop PV generation forecasts (from the Australian Solar Energy Forecasting System phase 2 (ASEFS2))
- Type of day (weekday/weekend), school holidays, public holidays and daylight savings information.
- Mandatory Restrictions (MR)/RERT schedules.
For more information on load forecasting in pre-dispatch and STPASA refer to the document below.
AEMO uses forecasts from weather forecast providers as an input into the Demand Forecasting System (DFS). AEMO assessed the performance of its weather providers over the summer 2018-19, winter 2019 and summer 2019-20 periods, with the reports provided below. AEMO is committed to collaboration with the meteorological industry to continue improving weather forecasting techniques for Australia's energy markets.