Meridian Energy Australia (MEA) and the University of Melbourne are developing innovative techniques to improve 5 minute ahead, Unconstrained Intermittent Generation Forecasts (UIGF) for the Australian Energy Market Operator (AEMO).
This will be done at the 131 MW Mt Mercer Wind Farm in Victoria and the 70 MW Mt Millar Wind Farm in South Australia, which are both owned and operated by MEA. The project will develop a real-time wind forecasting system, which will optimise information from a number of data sources and forecast models. Both proven and yet to be developed techniques will be used to provide forecasts to AEMO for both wind farms to help AEMO predict how much energy will be produced in order to better balance electricity supply with demand.
How the project works
As the transition to cleaner energy solutions continues, the number of variable renewable resources in the overall energy mix will increase. This creates challenges for AEMO to accurately balance supply and demand across the entire system in real time.
This wind forecasting demonstration project will first develop a methodology that will allow MEA to provide an accurate, 5-minute ahead forecast of the output from the Mount Mercer Wind Farm and Mount Millar Wind Farm. It will do this by combining data-driven and physics-based modelling, relevant weather phenomena such as wind, rain and thunderstorms and the onset of calm night-time conditions that affect the amount of wind that will be generated. The benefits to the National Electricity Market (NEM) of these improved forecasts will also be investigated.
Project: Meridian Energy Australia Wind Forecasting Demonstration Project Lessons Learnt 6
This report outlines lessons that Meridian Energy Australia and the University of Melbourne have recently learnt regarding clustered wind turbine-based forecasting and wind turbine cross correlations for the Mt. Mercer wind farm.
Area of innovation
The wind forecasting demonstration project is innovative due to its use of a combination of data sources and new algorithms. This includes existing atmospheric data available from measurement devices currently installed at the site, along with a scanning light detection and ranging system (LIDAR) and other equipment. This data will be incorporated into new algorithms to calculate a 5-minute ahead forecast for both wind farms. This forecast will be sent to AEMO via an application program interface tool also developed as part of this project, and the market benefits of improved forecasts will be reviewed.
Improved short-term forecasting of semi-scheduled generators should improve the operation of the National Electricity Market (NEM). Several studies around the world, including one undertaken at the University of Melbourne, suggest that once wind and solar exceed 30-40% of the overall generation mix (as expected in Australia within the next 10 to 20 years), energy forecasts (in MW) become more uncertain than that of demand.
This project will therefore first implement, and then quantify, the benefits to the NEM of more accurate, short-term wind forecasting from two wind farms. Improved forecast performance could result in less uncertainty, and therefore enable the market operator to schedule all generators more efficiently. This, in turn, could lower costs for consumers and lower greenhouse gas emissions from the whole system, relative to a less well forecasted system with the same installed renewable capacity.
The project will also develop an interactive website that will allow the public and user groups to observe the output from Mt Mercer Wind Farm in real time and compare it to the forecast for the same timeframe. The public and user groups will also be able to simultaneously view conditions at the site via the web cameras that will be installed at locations around the site. Some subsets of data are expected to be made available to researchers and school groups for further study as the project progresses.
A trial of forecasting technology has been launched to help predict the future output from wind and solar farms, which varies depending on the weather and time of day.