FORECASTING FOR UTILITY-SCALE WIND FARMS - THE POWER MODEL CHALLENGE

As the penetration of wind energy continues to increase around the world, with a trend towards large utility-scale wind farms (greater than 100 MW), effective wind energy forecasting will become increasingly important. Previous work by GH has estimated the trading benefit of high quality short-term forecasting to be (sic)7/MWh. Depending on market conditions, for a 100MW wind farm with a capacity factor of 30%, this equates to an estimated annual trading benefit of up to (sic)1.8m. To date, a number of studies have focused on the mathematical modelling techniques for forecasting the production from wind farms, looking predominantly at the task of predicting the meteorological conditions at the site. This paper focuses on the final stage of the forecasting process, conversion from a meteorological forecast to a power production forecast. This challenge is particularly significant for utility-scale wind farms, where the simple application of a turbine manufacturer's power curve is insufficient to capture the true behaviour and interaction of the wind turbines over the whole site. A simple power model can be responsible for introducing mean absolute errors of the order of 10% of capacity in the final power forecast. Using more advanced power modelling methods, the potential error introduced by the power model can be reduced to around 2% of capacity. For a 100MW wind farm, GH estimates the increase in annual trading revenue when using an advanced power model to be (sic)180,000.