Uncertainty evaluation of an alternative energy yield assessment method for repowering projects
Résumé
The prediction of a wind plant production over its lifetime is a critical component of the plant development decision process. Current industry model are called "greenfield AEP prediction" and rely on local climate onsite measurement and wind flow models. The model also includes power curve models and plant production losses models. The error in production prediction exhibits a certain heteroskedasticity with the project parameters (number of turbines, etc...) and the error standard deviation typically ranges from 6% to 12%. In this poster, we present an alternative predictor that applies to wind plant repowering. It consist of a learning from the existing plant production data. We demonstrate that significant precision improvement can be achieved.
Origine : Fichiers produits par l'(les) auteur(s)