With the frequent occurrence of extreme weather conditions, the safe operation of offshore facilities has been seriously challenged. Previous research attempted to simulate hydrodynamic performance or structure dynamic analysis to single environmental load. The present thesis proposes two methodologies to assess the operational risk quantitatively with combined wind and wave loads in a harsh environment considering the dependence structure between the real-time environmental parameters. The first developed model calculates the environmental loads using wind load response modeling, the Morison model and the ultimate limit states method. Then physical reliability models and joint probability functions derive the probabilities of failure at the level of structural components corresponding to combined loads. BN integrates the root probabilities according to the unit configuration to calculate the failure probability of the Semi-submersible Mobile Unit (SMU). The model is examined with a case study of the Ocean Ranger capsizing accident on Feb 15, 1982. The model uses the prevailing weather conditions and calculates a very high probability of failure 0.7812, which proves the robustness and effectiveness of the proposed model. The second proposed model is the copula-based bivariate operational failure assessment function, which assesses the dependencies among the real-time environmental parameters. Dependence function is described by the parameter δ from the wave data and concurrent meteorological observation data which are obtained from the Department of Fisheries and Oceans Canada (DFO). Then, the true model is selected with the help of Akaike’s information criterion (AIC) differences and Akaike weight. Comparing the results from the proposed approach and the traditional approach, it is shown that operational failure probabilities considering dependence are noteworthy higher and deserve attention. In other words, the traditional approach underestimates the operational risk of offshore facilities, especially in harsh environments.
The full report can be found here.
Adapted from https://research.library.mun.ca/14398/