Experimental analysis of analytical wake models for wind turbines
Author
Salgado, Luis Garcia
Term
4. semester
Education
Publication year
2017
Submitted on
2017-05-27
Pages
68
Abstract
Når vindmøller står samlet i en vindmøllepark, påvirker de hinanden gennem luftstrømmen. Møller forrest skaber en våge (wake) af langsommere og mere turbulent luft bag sig. Den lavere vindhastighed sænker elproduktionen for de bagvedliggende møller, og den højere turbulensintensitet (styrken af hastighedssvingninger) giver svingende belastninger, som kan føre til træthedsskader. For at forbedre opstilling og drift er det derfor vigtigt at forstå vågen grundigt. I dette arbejde undersøges fem vågemodeller, der forudsiger, hvordan vindhastigheden i vågen aftager (hastighedsunderskud) og hvordan vågen breder sig, samt tre modeller, der beskriver turbulensintensiteten i vågen. Modellerne valideres med målinger fra vindtunnelforsøg. Resultaterne viser, at modellerne overordnet kan beskrive vågens egenskaber rimeligt godt. Med den bedste analytiske model afviger forudsigelserne af hastighedsunderskud og vågebredde kun omkring 3% fra målingerne. Forudsigelserne af turbulensintensitet varierer mere, men der kan stadig opnås præcise resultater. Samlet set tyder arbejdet på, at analytiske vågemodeller er velegnede til at estimere hastighedsunderskud, breddeudvikling og turbulensintensitet i den fjerne våge (området længere nedstrøms bag møllen) af en vindmølle.
When wind turbines are grouped in a wind farm, they influence each other through the airflow. Turbines at the front generate a wake of slower, more turbulent air behind them. This lower-speed air reduces the power of downstream turbines, and the higher turbulence intensity (the strength of speed fluctuations) creates varying loads that can lead to fatigue damage. To improve turbine layouts and operation, a solid understanding of the wake is needed. In this study, five wake models are examined to predict how wind speed in the wake decays (velocity deficit) and how the wake widens, and three models are studied to describe turbulence intensity in the wake. The models are validated using measurements from wind tunnel experiments. The results show that the models can generally capture wake behavior fairly well. With the best analytical model, predictions of the velocity deficit and the wake width differ by only about 3% from the measurements. Predictions of turbulence intensity are more scattered, but accurate results are still achievable. Overall, the study concludes that analytical wake models are well suited to estimate velocity deficit, wake growth, and turbulence intensity in the far wake (the region farther downstream behind the turbine) of a wind turbine.
[This abstract was generated with the help of AI]
Keywords
Documents