Wind energyProject Wind Forecasting for the NEM
This is the final report for Fulcrum 3D’s Wind Forecasting for the NEM project. Fulcrum3D installed four Sodar systems, two at Taralga and one each at Clements Gap and Crowlands.
The objective of this project was to provide high accuracy short-term power forecasts utilising peripheral sodars and state-of-the-art software at three Pacific Hydro wind farms: Clements Gap (SA), Crowlands (VIC) and Taralga (NSW).
The key driver for this project was to reduce Causer Pays charges which can be enacted when a semi-scheduled generator is unable to meet the generation target set by the Australian Energy Market Operator (AEMO) dispatch system. In 2018 AEMO enabled semi-scheduled generators to submit their own five-minute forecasts through their MP5F API. It was expected that self-forecasts that are more accurate than AEMO’s forecasts would lower the solar farm’s Causer Pays penalties.
Fulcrum3D installed Sodars upstream of the wind farm prevailing wind directions: two at Taralga and one each at Clements Gap and Crowlands. A detailed power model for each wind farm was developed using real time data inputs from power station SCADA. These models are used to calculate the current unconstrained generation capability, produce five-minute wind farm output forecasts and submit them to AEMO via their MP5F API.
The project’s core achievement was demonstrating that it is possible to satisfy AEMO’s accuracy requirements and achieve zero Causer Pays charges at wind farms via MP5F with generators operating under normal conditions.
Our forecasts are now being used in dispatch at all three wind farms and are consistently outperforming AWEFS. The project met its contracted outcomes.