Reliability Analysis and Design Procedures for Offshore Wind Turbines
by Puneet Agarwal, (Dept. of Civil, Architectural, and Environmental Engineering, The University of Texas, Austin, 1 Universty Station, C1748, TX 78712 E-mail: pagarwal@mail.utexas.edu) and Lance Manuel, (Dept. of Civil, Architectural, and Environmental Engineering, The University of Texas, Austin, 1 Universty Station, C1748, TX 78712 E-mail: lmanuel@mail.utexas.edu)
pp. 1-9, (doi: http://dx.doi.org/10.1061/41016(314)194)
Access full text
Purchase Subscription
Permissions for Reuse 
| Document type: |
Conference Proceeding Paper |
| Part of: |
Structures Congress 2008: Crossing Borders |
| Abstract: |
Design standards for wind turbines such as the guidelines from the International Electrotechnical Commission (IEC) require, for at least one design load case, statistical extrapolation of loads from limited simulation analyses. Extrapolation refers to prediction of a rare load fractile. Several statistical extrapolation techniques are possible; one includes the use of the peak-over-threshold method in which load maxima between successive upcrossings of a selected threshold are defined as the extremes for random variables. With such "extremes" models, it is common to fit parametric distributions to the loads data and to then attempt to predict load levels associated with rare occurrence rates (in ten minutes), generally on the order of one in 1 million or even rarer. Structural reliability principles can be employed to improve the efficiency in such computations. In this study, ultimate limit states will be evaluated for a 5MW offshore wind turbine by first studying short-term load statistics (conditional on wind speed and wave height); then, upon integration of these short-term distributions with site-specific environmental climate descriptions, long-term turbine loads will be derived. |
|
