Probabilistic Fatigue Reliability Assessment of Critical Tubular Joints in Gulf of Guinea Steel Jacket Structures

Abstract

Deterministic fatigue calculations systematically overstate residual life because they neglect uncertainty in stress concentration factors, environmental loading, material properties, and corrosion rates. This paper develops a probabilistic reliability framework for the six critical tubular joints of a representative Gulf of Guinea steel jacket. Uncertainty is propagated by the First-Order Reliability Method (FORM) and Monte Carlo simulation using limit-state functions based on both S–N cumulative damage and fracture-mechanics crack growth. At the original design life of 25 years, the reliability index of the governing lower-bay K-joint (J5) falls to ( ), well below the conventional target of 3.1. Deterministic estimates overstate residual life by 28–34 %. FORM sensitivity analysis identifies the SCF bias factor as the dominant contributor to failure probability, followed by hot-spot stress range and the Paris-law coefficient. The resulting time-dependent component reliability indices provide the essential input for system-level reliability assessment and risk-based inspection planning of ageing Gulf of Guinea jackets.