| Popis: |
In shallow waters, subsea pipelines can suffer fatigue damage from Vortex Induced Vibrations (VIV) by the combined effects of waves and currents. A full characterization of the joint probability distribution of waves and currents involves at least 5 variables, 2 for currents (magnitude and direction), and 3 for waves (significant wave height, mean wave direction, and wave period at the peak of the spectrum). In lieu of sufficient data to adequately characterize the associated joint probability density function, DNV GL in their “Recommended Practice” propose an approximation: Assume that the direction of current and wave effects is always the same, but for a given direction, waves and currents are assumed statistically independent. In this paper 28 years of hindcast data are used to test the accuracy of this colinearity approximation in the Sea of Okhotsk. Rather than attempting to estimate a joint probability density function in 5 variables, the span fatigue assessment is simply performed for the entire 28 years of the hindcast database to obtain an average rate of fatigue damage. It is found that this history-based approach can lead to fatigue damage rates that are much higher than those derived from the colinearity assumption. This non-conservatism of the colinearity assumption, arises for pipeline orientations for which both waves and currents can contribute strongly to the VIV response without being exactly colinear. It is concluded that caution is needed in using the colinearity assumption, but an update of the span assessment procedure should also address issues for which current assessment procedures are conservative, such as seabed proximity and trench effects, nonlinearity in the pipe-soil interaction, and the change in axial forces due to transverse displacements of the pipe, which are expected under extreme combined current and wave conditions, as envisioned on on-bottom stability design criteria. |
