Database for validation of design load extrapolation techniques

Authors


Abstract

Extrapolation techniques used for predicting long-term wind turbine loads have produced highly variable loading estimates dependent on the individual designer implementation. To reduce such variability, more precise definition and validation of these techniques are necessary. As part of a wider effort to test loads extrapolation techniques used in wind turbine design, two data sets were created. The first data set was created as an example typical of what designers use to extrapolate loads according to wind turbine design standards. The second data set was a series of year-long simulations that could be used to quantify the accuracy of extrapolation methods. Due to the large number of simulations, care was taken not to reproduce random samplings in individual time series. Additionally, a grid computing architecture was used to run the simulations in a reasonable time frame. The wind speeds where loads were highest were identified, which varied with loading type. The identification of dominant wind speeds is important to ensure an adequate number of simulations at such speeds, which often influence extrapolated loads. In-plane loads and deflections tended to be dominated by high wind speeds near cut-out wind speed. Out-of-plane loads and deflections were most influenced by winds that were near the rated wind speed. Some loads were influenced by a range of wind speeds. The loads dominated by high wind speeds had greater variability in the extreme values, which could be a reflection of the greatly varying wind and also greater sensitivity to higher energy content of the wind at such speeds. Copyright © 2008 John Wiley & Sons, Ltd.

Ancillary