Comparing Methods for Estimating the Change in Energy from Wind Farm Flow Control
Автор: IEA Wind Task 44 Talks
Загружено: 2025-07-23
Просмотров: 59
Work package 2 presentation by Eric Simley, NREL from February 2025.
In this talk, we compare two methods for estimating the change in energy from a wind farm flow control toggle experiment using NREL's FLASC software: the "power ratio" method and the "expected power" method based on the 2020 TNO report "Active Wake Control Validation Methodology." We demonstrate the two methods using the SMARTEOLE wake steering dataset. We investigate the sensitivity of the estimated energy gain and its uncertainty to the wind speed bin size, finding greater sensitivity in the expected power method. Additionally, we compare the two methods' approaches to uncertainty quantification (UQ), finding that the bootstrapping approach of the power ratio method yields an uplift uncertainty that closely matches the uncertainty from the expected power method, which is based on propagation of standard errors. Lastly, we investigate the block bootstrapping method commonly used for UQ with temporally correlated time series data. Results indicate that a block size of 8 hours produces roughly twice the amount of uncertainty as standard bootstrapping, suggesting that both standard bootstrapping and the standard error-based expected power UQ methodology may significantly underestimate the uplift uncertainty.
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
