How Probabilistic Wind Power Prediction can Help in Optimising Operations
It is now widely accepted that wind power forecasts can help utilities and transmission system operators (TSOs) integrate highly variable wind power into their operations. These forecasts are known as ‘deterministic’ predictions and are often based on numerical weather prediction model wind forecasts, which are then optimised and converted to deterministic power forecasts. Now, additionally, forecast providers are beginning to provide ‘probabilistic’ forecasts that can quantify the uncertainty of the prediction. These probabilistic forecasts are traditionally produced by running the models with slightly different scenarios, producing an ensemble of possible predictions. Newer methods are providing ways to accurately quantify the uncertainty with fewer ensemble members or even by creating an ensemble from a single deterministic run – an analog ensemble. Such methods to quantify the uncertainty are making it possible to further optimise wind power integration.
By Dr Sue Ellen Haupt and Dr Luca Delle Monache, National Center for Atmospheric Research, USA