Probabilistic prediction of external corrosion fatigue life in buried steel pipes

Abstract

This research investigates a methodology for probabilistic life prediction of buried steel pipelines subjected to external corrosion. A unified methodology is developed considering multiple stages of degradation related to external corrosion (due to soil) and fatigue. These stages include corrosion pit nucleation, pit growth, transition from pit to short crack, short crack growth, transition from short to long crack, stable long crack growth, and unstable fracture. The methodology is useful in obtaining stage-specific forecasts for the fatigue life of buried steel pipelines subjected to external pitting corrosion fatigue. State-of-the-art computational models are used to predict damage initiation and evolution at each stage. The variability in environmental, material, and loading parameters is propagated through these models to obtain a probabilistic estimate of the remaining service life (RSL) of the pipe. Insights from probabilistic RSL prediction highlight the influence of soil type and pipe coating material on corrosion fatigue life. Global sensitivity analysis is then employed to quantify the relative importance of environmental factors (pH, pipe/soil potential, and chloride concentration), material properties (threshold stress intensity factor), and the range of cyclic stress experienced by the pipe.