Fatigue analysis of offshore well conductors: Part II – Development of new approaches for conductor fatigue analysis in clays and sands

• Published in: Applied ocean research 2016-04, Vol.57, p.96-113
• Main Authors: Zakeri, Arash, Clukey, Edward C., Kebadze, Elizbar Buba, Jeanjean, Philippe
• Format: Article
• Language: English
• Subjects: Cyclic soil response; Fatigue analysis; Marine; p–y curve; Soil–conductor interaction; Well conductor
• ISSN: 0141-1187; 1879-1549
• DOI: 10.1016/j.apor.2016.03.001

•Describes offshore well conductor fatigue analysis and demonstrates the knowledge gap and the need for soil models.•Presents a novel physical testing programme that takes a fundamental approach to investigate conductor–soil interaction under cyclic loading.•Outlines a framework for soil model development applicable specifically to conductor fatigue analysis and demonstrates its suitability for the purpose. This paper, Part II, presents two soil modelling approaches developed specifically for fatigue analysis of well conductors. The first approach uses Winkler springs and can account for soil damping. The second approach is based on continuum soil mechanics and uses the kinematic hardening principles. They focus on cyclic soil behaviour at the steady-state condition. The paper demonstrates appropriateness of the approaches in predicting fatigue damage through comparisons with the centrifuge fatigue lives measured from the Series 1 tests in NC to lightly OC kaolin clay. It also presents and discusses the analyzed data and the fatigue lives measured from the Series 2 (medium-dense sand), Series 3 (NC to lightly OC GoM clay), and Series 4 (OC natural clay) tests. Finally, soil models (based on the first approach) are presented for fatigue analysis of well conductors installed in NC to lightly OC clays, over-consolidated (OC) clays and medium-dense sands, and their ability to predict conductor fatigue damage under wide range of loading conditions is demonstrated. Overall, the accuracy of fatigue life predictions using these novel soil models is very high – generally within about a few percentages of the measured values.