Precise Stochastic Fatigue Life Assessment of Umbilicals and Power Cables

Abstract

Abstract Increasing activity within offshore wind energy and related installations has raised the focus on the integrity and fatigue life of power cables and umbilicals. A limiting factor for assessing fatigue life has been the lack of efficient and precise computational tools that are not overly conservative. This paper presents features of an advanced computational method implemented in a new tool for accurately assessing the integrity and predicting the fatigue service life of dynamic power cables and umbilicals. Internal friction between components of power cables and umbilicals creates hysteretic stress responses that global finite element analysis (FEA) does not capture. It is widespread practice to assess irregular wave global responses through time-domain FEA and local cross-sectional analysis through specialized FEA tools. The global responses are combined with results from local FEA tools like UFLEX (developed by Sintef Ocean) to obtain accurate stresses capturing local effects, including friction induced hysteresis. Kongsberg Maritime introduced an innovative method in Hoen-Sorteberg et. al 2013 (OTC-24328-MS), employing tension-dependent hysteresis for flexible risers. This methodology has now been expanded to support power cables and Umbilicals. Fatigue life assessment of unbonded structures like power cables and umbilicals involves analyzing various elements in the cross-section at multiple locations and orientations to accurately identify critical hotspots, considering load sharing and local friction between components like steel tubes, armor wires, and conductor strands exhibiting complex stress hysteresis behavior. Local FEA provides essential data to account for these complex hysteresis effects and linear stress factors like tension and component bending. The improved methodology and its implementation permit comprehensive stochastic fatigue assessment of power cables and umbilicals without imposing an undue computational burden. The outcome leads to more precise and often substantially prolonged fatigue life predictions by mitigating the inherent conservatism found in methods based on regular waves. This paper demonstrates the practicality of the methodology and robustness through real-world implementation. The local-global approach yields precise stress levels comparable to direct local model analysis but with markedly enhanced computational efficiency. Previous similar methods suffered from limitations in both stress accuracy and computational speed, limiting their practical utility. A significant advantage is that this computational tool operates independently without needing specific FEA software integration, allowing users to utilize data from various sources.