The fatigue reliability assessment of deepwater risers plays an important role in the safety of oil and gas development. Physical-based models are widely used in riser fatigue reliability analyses. However, these models present some disadvantages in riser fatigue reliability analyses, such as low computational efficiency and the inability to introduce inspection data. An improved fatigue reliability analysis method was proposed to conduct the fatigue reliability assessment of deepwater risers. The data-driven models were established based on response surface methods to replace the original physical-based models. They are more efficient than the physics-based model, because a large number of complex numerical and iterative solutions are avoided in fatigue reliability analysis. The annual crack growth model of the riser based on fracture mechanics was established by considering the crack inspection data as a factor, and the crack growth dynamic Bayesian network was established to evaluate and update the fatigue reliability of the riser. The performance of the proposed method was demonstrated by applying the method to a case. Results showed that the data-driven models could be used to analyze riser fatigue accurately, and the crack growth model could be performed to analyze riser fatigue reliability efficiently. The crack inspection results update the random parameters distribution and the fatigue reliability of deepwater risers by Bayesian inference. The accuracy and efficiency of fatigue analysis of deepwater risers can be improved using the proposed method.

No abstract

No abstract

No abstract