Citation

Arif M, Khan F, Ahmed S, Imtiaz S. Saf. Extrem. Environ. 2022; 4(3): 247-255.

Copyright

(Copyright © 2022, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s42797-022-00064-2

PMID

unavailable

Abstract

Probabilistic approaches under stationary conditions are used to design offshore systems. However, the frequency and impact of extremes in offshore environments have been changing. The present study investigates whether there are significant inter-period trends in extreme wind loads from the historical period to the future. This study comprises two elements. The first element considers the traditional Block Maxima based Generalized Extreme Value (GEV) approach and generates different return levels-based risk profiles (under stationary condition). In the second element, the GEV location parameters are considered as time-dependent and a dynamic risk profile of different returns for wind events is developed. Distribution parameters are computed using the well-known Maximum Likelihood Estimate (MLE) method. The proposed methodology is tested using data from the Grand Banks region of the east coast of Canada. The study observed a 100-year extreme wind speed of 33.7 m/s (occurrence probability, 9.99E-05) assuming a stationary condition, whereas for the non-stationary model, the predicted extreme wind speed for 2030 is 41.5 m/s (occurrence probability 2.82E-05) in the Grand Banks region, Canada. The study highlights that a 100- year return period is not an adequate design criterion under the current situation; one must consider the non-stationary behaviour or a higher return period such as 500 or 1000 years.

Language: en

Keywords

Offshore system extreme value; Return level; Risk-based design; Wind load