Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Wind speed predictions: Are we at the limit of our knowledge or can we improve?' taking place on Wednesday, 12 March 2014 at 11:15-12:45. The meet-the-authors will take place in the poster area.

Julia Bethke anemos GmbH, Germany
Co-authors:

(1) anemos GmbH, Reppenstedt, Germany

Printer friendly version: printer.gif Print

Abstract

Verification and performance of wind climatologies in wind energy applications

Introduction

Wind climatologies such as reanalysis data or wind field simulations are widely used for applications in wind energy, e.g. for long-term correlation of short-term measurements or initial site search. Site selection is addressed within the EU FP7 project SOPCAWIND (Software for the Optimal Place CAlculation for WIND farms, Grant 296164), where wind climatologies are used as input data for the developed software.
Within this project for the first time different wind climatologies are verified in-depth with numerous certified high-quality mast measurements covering a broad range of heights up to 200 m. Their advantages and limitations are discussed.

Approach

We investigate the performance of commonly used reanalysis data, e.g. MERRA, ERA-Interim, and the anemos wind atlas for Europe (EU-20) in wind energy applications.
EU-20 is a wind field simulation derived by the mesoscale model MM5 using the MERRA reanalysis as driving data. Its spatial and temporal resolution is 20 km and 10 min, respectively.
Certified high-quality measurements of 45 met masts with 160 anemometers covering a range of complexity types, measurement heights between 30 m and 200 m and the time period from mid-2010 to mid-2013 are compared to EU-20 and reanalysis data. Hourly averages are analysed.

Main body of abstract

The ability of the wind climatologies to reproduce annual and diurnal cycles of wind speed, as well as long-term averages and frequency distributions of wind speed and wind direction are investigated. Furthermore, atmospheric stability and vertical wind shear are examined.
The correlation with hourly measurements of wind speeds is very good for all data sets. EU-20 and MERRA yield highest correlation coefficients. Correlation increases with decreasing terrain complexity.
For low complexity terrain the frequency distribution of wind speed is usually met quite well by EU-20 for a broad range of velocities, however underestimating higher velocities. MERRA generally strongly overestimates wind speed. Diurnal and annual cycles are reproduced most accurately by EU-20. Additionally, in contrast to the reanalysis data, EU-20 is able to describe the change of atmospheric stability during the day to a certain extent. Wind directions are met very well by all data sets.
For areas with higher complexity and increased roughness (e.g. forests) the performance of wind climatologies weakens. For offshore sites an underestimation of the level of wind speed is observed.

Conclusion

Correlations with wind speed and wind direction are high, indicating that all data sets are useful for long-term correlation. However, EU-20 yields an overall better performance than the reanalysis data when it comes to detailed analysis. This suggests that wind field simulations downscaled from reanalysis data are more appropriate for applications where the absolute value is important, e.g. the initial estimation of the wind potential and site search.
Performance increases with decreasing complexity, except for offshore sites. Therefore, as a next step a measure of uncertainty depending on terrain complexity and roughness will be derived.


Learning objectives
The presentation will give a hint on the advantages and limits of commonly used reanalysis data sets and wind field simulations downscaled from reanalysis data in wind energy applications.