0
Research Papers

Empirical Wind Turbine Load Distributions Using Field Data

[+] Author and Article Information
Puneet Agarwal

Department of Civil, Architectural, and Environmental Engineering,  University of Texas, Austin, TX 78712pagarwal@mail.utexas.edu

Lance Manuel1

Department of Civil, Architectural, and Environmental Engineering,  University of Texas, Austin, TX 78712lmanuel@mail.utexas.edu

1

Corresponding author.

J. Offshore Mech. Arct. Eng 130(1), 011006 (Jan 29, 2008) (6 pages) doi:10.1115/1.2827937 History: Received July 08, 2007; Revised October 21, 2007; Published January 29, 2008

Abstract

In the design of land-based or offshore wind turbines for ultimate limit states, long-term loads associated with return periods on the order of the service life ($20years$, usually) must be estimated. This requires statistical extrapolation from turbine load data that may be obtained by simulation or by field tests. The present study illustrates such extrapolation that uses field data from the Blyth offshore wind farm in the United Kingdom, where a $2MW$ wind turbine was instrumented, and environment and load data were recorded. From this measurement campaign, the load data available are in two different formats: as $10min$ statistics (referred to as “summary” data) or as full time series (referred to as “campaign” data). The characteristics of the site and environment and, hence, that of the turbine response are strikingly different for winds from the sea and winds from the shore. The load data (here, only the mudline bending moment is studied) at the Blyth site are hence separated depending on wind regime. By integrating load distributions conditional on the environment with the relative likelihood of the different environmental conditions, long-term loads associated with specified return periods can be derived. This is achieved here using the peak-over-threshold method based on campaign data but long-term loads are compared with similar estimates based on the summary data. Winds from the shore are seen to govern the long-term loads at the site. Though the influence of wave heights on turbine long-term loads is smaller than that of wind speed, there is possible resonance of tower dynamics induced by the waves; still, to first order, it is largely the wind speed and turbulence intensity that control design loads. Predicted design loads based on the campaign data are close to those based on the summary data discussed in a separate study.

<>

Figures

Figure 2

(a) Scatter diagram showing mean wind speed versus significant wave heights; (b) wind rose

Figure 3

Distribution of usable campaign data sets by mean wind speed and significant wave height bins for (a) winds from the sea and (b) winds from the shore

Figure 4

Time series and power spectra of the wind speed at nacelle and met. mast, sea surface elevation, and mudline bending moment for the following 10min time series: (a) ShoreV18H0N1, (b) ShoreV18H0N9, and (c) SeaV12H3N41. Only a 200s portion is shown for each time series where the maximum load was recorded.

Figure 5

Probability of load exceedance curves for winds from the sea, winds from the shore, and winds from all directions

Figure 6

90% confidence intervals on probability of load exceedance curves for winds from the sea, winds from the shore, and winds from all directions based on bootstrapping

Figure 1

Location of the turbines and an onshore meteorological mast at the Blyth site (from Camp (5))

Errata

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections