Prediction of the tidal turbine power fluctuations from the knowledge of incoming flow structures

Type Article
Date 2022-05
Language English
Author(s) Druault Philippe1, Germain GregoryORCID2
Affiliation(s) 1 : Sorbonne Université, CNRS, Institut Jean Le Rond d’Alembert, F-75005 Paris, France
2 : Ifremer, Marine Structure Laboratory, 150 quai Gambetta, 62200 Boulogne-sur-mer, France
Source Ocean Engineering (0029-8018) (Elsevier BV), 2022-05 , Vol. 252 , P. 111180 (17p.)
DOI 10.1016/j.oceaneng.2022.111180
WOS© Times Cited 7
Keyword(s) Turbine power fluctuations, Large scale flow structures, Stochastic estimation, Proper orthogonal decomposition, Fourier analysis

After positioning a 1:20 scaled model of a three-bladed horizontal-axis turbine in the wake of a wall-mounted cylinder, synchronized turbine performance and flow measurements are carried out to investigate the relationship between the incoming flow field and the turbine power fluctuations. The Linear Stochastic Estimation (LSE) is used to predict the turbine output fluctuations from the knowledge of the Large Scale flow Structures (LSS) embedded in the incoming turbulent flow. LSS extraction by Fourier analysis or Proper Orthogonal Decomposition shows that LSS are responsible for the main unsteady variations of the power fluctuations, especially their highest amplitudes. The RMS of turbine output fluctuations are entirely due to the LSS. It is also demonstrated that whatever the nature of the incoming turbulent flow is, the low frequency filtering process coupled with the LSE method allows the recovering of at least 90% of the turbine power RMS. Furthermore, the low-frequency spectral content of the turbine power fluctuations is very well predicted, especially the frequency peaks. A preliminary LSE application is performed in order to predict the instantaneous turbine output fluctuations at more than 85% confidence level, from only three velocity signals measured in front of the turbine.

Full Text
File Pages Size Access
17 5 MB Access on demand
Author's final draft 34 7 MB Open access
Top of the page