Statistical model for fragility estimates of offshore wind turbines subjected to aero-hydro dynamic loads

• Published in: Renewable energy 2021-01, Vol.163, p.1495-1507
• Main Authors: Pokhrel, Jharna, Seo, Junwon
• Format: Article
• Language: English
• Subjects: FAST; Fragility; Offshore wind turbine; Regression model; Wave; Wind
• ISSN: 0960-1481; 1879-0682
• DOI: 10.1016/j.renene.2020.10.015

This paper aims to develop statistical regression-based models to estimate responses of monopile foundation 5 MW Offshore Wind Turbines (OWTs) subjected to multi-wind-and-wave loads and to assess its fragilities. Establishing the regression models began with the use of the Latin Hypercube Sampling (LHS) method incorporating 5 MW OWT input parameters related to structural and loading conditions along with material properties. With the LHS-based input parameters, 120 OWT computational models were created through Fatigue, Aerodynamic, Structures, and Turbulence (FAST) tools developed by the National Renewable Energy Laboratory (NREL). Critical responses, such as tower-top deflection, corresponding to each of the models were determined by performing its FAST aero-hydro dynamic simulations. The regression models involved a series of developed explanatory functions based on a matrix of the FAST responses and LHS parameters under a Stepwise Multiple Linear Regression (SMLR) approach. Multi-wind-and-wave fragilities were estimated for each of the critical responses of the 120 OWT models. Key findings showed that the wind-sensitive blade tip deflection resulted in the fragility of 99% at a critical wind speed of 75 m/s and a wave height of 20m, while the mudline flexural moment resulting from the same wind speed and wave height caused the fragility up to 98%. •Computational models were developed to predict responses of 5 MW offshore wind turbines.•Regression models were created based on the computational responses.•The regression models were integrated with Monte Carlo Simulations in a probabilistic fashion.•The integrated approach was capable of estimating wind and wave offshore wind turbine vulnerability efficiently.