A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Assessing some statistical and physical modelling uncertainties of extreme responses for monopile-based offshore wind turbines, using metocean contours
https://orcid.org/0000-0002-1137-3764
Abstract This study examines the influence of probabilistic models for wave parameters in the joint environmental model and hydrodynamic/soil models on extreme mudline bending moments for monopile-based wind turbines at representative wind speeds, using the environmental contour method. For significant wave height, the 3-parameter Weibull model using the method of moments (MoM) provides the best fit to hindcast data across different wind classes, for the statistical models and data considered in the study. The hybrid Log-normal-Weibull (LonoWe) model also provides a reasonable fit but is sensitive to the transition point between distributions. Both models yield the largest extreme responses, with differences of approximately 0.5–3.5%. The 3-parameter Weibull model with maximum likelihood estimation (MLE) and the 2-parameter Weibull model result in less conservative contours, leading to up to 13% lower extreme responses, compared to LonoWe and Weibull (MoM). Regarding peak period, both the Log-normal and 3-parameter Weibull models provide reasonable fits, with the latter being more accurate near the steepness (breaking) limit. The stochastic variation among maxima due to seed variability and the uncertainty in quantile estimates as a function of number of samples was found to be crucial, particularly for severe sea states at the cut-out speed. Soil modelling is particularly important when the turbine is parked and encounters peak wave periods close to the turbine’s natural periods, while the effect of soil modelling on the extremes during turbine operation is negligible. Additionally, the impact of diffraction becomes relatively important for short wave periods. However, it is worth noting that the choice of load models has less impact on extreme responses compared to variations in the contours caused by different statistical models or seed variability.
Highlights Extreme responses of monopile-based offshore wind turbines using contour method. Metocean contours strongly depend on the selection of Hs-Tp conditional distributions. Probabilistic models and stochastic variation from seed variability dominate extremes. Foundation modelling affects extreme responses primarily for the parked states. Extreme responses in rated speed become increasingly important for larger turbines.
Assessing some statistical and physical modelling uncertainties of extreme responses for monopile-based offshore wind turbines, using metocean contours
https://orcid.org/0000-0002-1137-3764
Abstract This study examines the influence of probabilistic models for wave parameters in the joint environmental model and hydrodynamic/soil models on extreme mudline bending moments for monopile-based wind turbines at representative wind speeds, using the environmental contour method. For significant wave height, the 3-parameter Weibull model using the method of moments (MoM) provides the best fit to hindcast data across different wind classes, for the statistical models and data considered in the study. The hybrid Log-normal-Weibull (LonoWe) model also provides a reasonable fit but is sensitive to the transition point between distributions. Both models yield the largest extreme responses, with differences of approximately 0.5–3.5%. The 3-parameter Weibull model with maximum likelihood estimation (MLE) and the 2-parameter Weibull model result in less conservative contours, leading to up to 13% lower extreme responses, compared to LonoWe and Weibull (MoM). Regarding peak period, both the Log-normal and 3-parameter Weibull models provide reasonable fits, with the latter being more accurate near the steepness (breaking) limit. The stochastic variation among maxima due to seed variability and the uncertainty in quantile estimates as a function of number of samples was found to be crucial, particularly for severe sea states at the cut-out speed. Soil modelling is particularly important when the turbine is parked and encounters peak wave periods close to the turbine’s natural periods, while the effect of soil modelling on the extremes during turbine operation is negligible. Additionally, the impact of diffraction becomes relatively important for short wave periods. However, it is worth noting that the choice of load models has less impact on extreme responses compared to variations in the contours caused by different statistical models or seed variability.
Highlights Extreme responses of monopile-based offshore wind turbines using contour method. Metocean contours strongly depend on the selection of Hs-Tp conditional distributions. Probabilistic models and stochastic variation from seed variability dominate extremes. Foundation modelling affects extreme responses primarily for the parked states. Extreme responses in rated speed become increasingly important for larger turbines.
Assessing some statistical and physical modelling uncertainties of extreme responses for monopile-based offshore wind turbines, using metocean contours
https://orcid.org/0000-0002-1137-3764
Abstract This study examines the influence of probabilistic models for wave parameters in the joint environmental model and hydrodynamic/soil models on extreme mudline bending moments for monopile-based wind turbines at representative wind speeds, using the environmental contour method. For significant wave height, the 3-parameter Weibull model using the method of moments (MoM) provides the best fit to hindcast data across different wind classes, for the statistical models and data considered in the study. The hybrid Log-normal-Weibull (LonoWe) model also provides a reasonable fit but is sensitive to the transition point between distributions. Both models yield the largest extreme responses, with differences of approximately 0.5–3.5%. The 3-parameter Weibull model with maximum likelihood estimation (MLE) and the 2-parameter Weibull model result in less conservative contours, leading to up to 13% lower extreme responses, compared to LonoWe and Weibull (MoM). Regarding peak period, both the Log-normal and 3-parameter Weibull models provide reasonable fits, with the latter being more accurate near the steepness (breaking) limit. The stochastic variation among maxima due to seed variability and the uncertainty in quantile estimates as a function of number of samples was found to be crucial, particularly for severe sea states at the cut-out speed. Soil modelling is particularly important when the turbine is parked and encounters peak wave periods close to the turbine’s natural periods, while the effect of soil modelling on the extremes during turbine operation is negligible. Additionally, the impact of diffraction becomes relatively important for short wave periods. However, it is worth noting that the choice of load models has less impact on extreme responses compared to variations in the contours caused by different statistical models or seed variability.
Highlights Extreme responses of monopile-based offshore wind turbines using contour method. Metocean contours strongly depend on the selection of Hs-Tp conditional distributions. Probabilistic models and stochastic variation from seed variability dominate extremes. Foundation modelling affects extreme responses primarily for the parked states. Extreme responses in rated speed become increasingly important for larger turbines.
Assessing some statistical and physical modelling uncertainties of extreme responses for monopile-based offshore wind turbines, using metocean contours
Katsikogiannis, George (author) / Haver, Sverre K. (author) / Bachynski-Polić, Erin E. (author)
Applied Ocean Research ; 143
2024-01-02
Article (Journal)
Electronic Resource
English
Multivariate analysis of extreme metocean conditions for offshore wind turbines
| Online Contents | 2015
Site-specific controller design for monopile offshore wind turbines
| Elsevier | 2018