Statistical analysis of wind characteristics at candidate wind-turbine sites
Accurate prediction for a wind turbine facility performance is dependent on appropriate wind characteristics data. This report presents preliminary recommendations of appropriate data recording procedures for initial site evaluation and for large machine performance evaluation. Conclusions are based on autocorrelation, autospectrum persistence, cross-correlation, and coherence analysis of a limited sample of wind speed and turbine power data collected at the MOD-OA 200 kw wind turbine generator site at Clayton, New Mexico. For preliminary site assessment, this study shows that data should be gathered from two heights on a meteorological tower. One height should approximate the elevation of the turbine hub. A second reference height, e.g. 10 m, would be useful for comparisons among different sites. For site selection, these data should be short-term averages. For instance, 15-second average wind speeds correlate well with turbine performance, and 1-minute average wind speeds provide a comparison to other sites with only National Weather Service (NWS) data. Once a turbine is installed, it is desirable to collect additional data for machine performance evaluation. A sampling rate of once every 10 minutes is suggested, with the following five pieces of data recorded: a single 1-second average wind speed, a single 15-second average wind speed, a single 1-minute average wind speed, a full 10-minute average wind speed, and the variance of all 15-second averages within the 10-minute sampling interval. These data provide extensive information for blade stress analysis, turbine performance and comparison to NWS sites. Wind speed data from the nacelle anemometer of an operating turbine exhibit relatively high correlation with turbine power that is independent of averaging times under a minute, so shorter averages may be omitted if only power comparisons are desired.
- Research Organization:
- Johns Hopkins Univ., Baltimore, MD (USA)
- DOE Contract Number:
- AC06-76RL01830
- OSTI ID:
- 6176187
- Report Number(s):
- PNL-4231; ON: DE83013742
- Resource Relation:
- Other Information: Portions are illegible in microfiche products
- Country of Publication:
- United States
- Language:
- English
Similar Records
Coastal Ohio Wind Project
The importance of round-robin validation when assessing machine-learning-based vertical extrapolation of wind speeds