Development of behaviour models for pitting corrosion of mild steel pipeline welds in marine environment

Natural deterioration of steel structures exposed to the marine environment compromises the long-term integrity, serviceability and safety of new and existing infrastructure and increases their risk of failure. Mitigation techniques such as cathodic protection or surface protective coatings are commonly applied, however if these are ineffective or cease to exist the evolution of corrosion with time is of interest. Experience has shown that generally welded structures are prone to even higher risks as a result of adverse effects of weld-heated areas. As quantitative data are scarce, particularly for longer-term exposures, probabilistic risk assessment is a sound framework to quantify the risk of structural failure arising from localized corrosion phenomena such as pits and crevices. Herein observations are reported and analysed using extreme value theory of the pitting corrosion of API X56 Spec 5 L grade pipeline steel exposed continuously to natural Pacific Ocean seawater for 3.5 years, and extended to much longer exposure periods by comparison to similar steels exposed for up to 33 years. Relationships between pit depth and material metallurgy and corrosion properties were investigated by means of standard macro-etching, rest potential and zero resistance ammetry techniques. It is considered that the observed results are the result of the lack of homogeneity at the corrosion interface caused by differences in grain size, grain structure and the potential for pitting to occur preferentially along boundaries. The statistics obtained for long term localized depth data and their suitability for long-term prediction is considered. It is shown that maximum pit depth and pit depth variability are not simple linear functions of exposure time as often assumed in practice. An example is given for the reliability analysis of a welded pipeline subjected to localized corrosion. The effect of random variable uncertainty is assessed using a sensitivity study. They show the considerable influence on the probability of failure of pit diameter and the parameters describing the pitting corrosion model. The results provide a means to estimate the long-term reliability of pipelines.