Posters

Abstract

Based on the Author's recent experience, primarily but not exclusively in the Nordic region, there is a growing body of evidence, primarily from operational wind farm SCADA data, that suggests conventional wake models with typical parametrisation that are used extensively in the wind industry under-predict the magnitude of wake losses. It is postulated that this is, at least in part, due to prevailing stable atmospheric conditions leading to wake effects persisting for greater distances downstream than typically modelled in the onshore environment, although other factors, such as cumulative induction zone effect (CIZE), also termed blockage effect, may also be relevant. The magnitude of wake loss under-prediction can be significant and may be particularly relevant within large development clusters. During energy yield assessment reconciliation exercises, adjustments are considered for all factors that can be quantified from operational data, such as wind resource and turbine availability; however, there are other factors, such as wind flow model error and turbine power performance, that cannot be quantified. As such, there is some degree of uncertainty regarding the level of bias that may be attributed to wake loss modelling in isolation. Nonetheless, evidence, including wake model validation, suggests that wake loss under-prediction is a significant contributing factor. Based on the individual wake model results and a reconciliation assessment, the level of uncertainty associated with wake modelling is appraised to advise appropriate uncertainty allocations in energy yield assessment.