Applications of statistics and probability to enhance the reliability And maintenance strategies for corroded pipelines

Abstract

Pipeline load carrying capacity and safety are often reduced by corrosion and associated damage. The prediction of future defects and the pipeline's remaining lifetime are obtained by using consistent assessments of corrosion rates. However, its modelling often involves simplifications and assumptions to compensate a lack of data, imprecision and vagueness, which cannot be justified completely and may, thus lead to biased results. To overcome these issues, an imprecise probabilities approach is proposed for reliability analysis, decision-making, risk-based design and maintenance. It is shown how this approach can improve the practise using B31G, Modified B31G, DNV-101 and Shell-92 failure pressure models. In addition, a robust and efficient probabilistic framework for optimal inspection and maintenance schedule selection for corroded pipelines is proposed. Optimal solution is obtained through only one reliability assessment removing huge computational cost of reliability-base optimization and generalised probabilistic methods and in turn, making the analysis of industrial size problem feasible.