Application of Markov modelling and Monte Carlo simulation technique in failure probability estimation - A consideration of corrosion defects of internally corroded pipelines

Abstract

For effective integrity management, the reliabilities at times of exposure of pipelines to corrosive environment need to be understood. This paper described the procedure for using Markov modelling and Monte Carlo simulation to determine the reliabilities of internally corroded pipelines. The corrosion wastage of the pipeline was classified with Pipeline Corrosivity Indexes (PCIs), which were expressed as functions of retained pipe-wall thickness at exposure times. The model was tested on X52 grade pipeline that was monitored by Magnetic Flux Leakage (MFL) In-Line-Inspection (ILI) technique and the failure probabilities were determined for different failure scenarios such as small leakage, bursting and rupture. It was observed that, as the time of exposure of the pipeline to corrosive condition increased, there was a slight variation of the reliability of the pipeline that failed by bursting and rupture. The result also indicated a very high likelihood of small leakage of the pipeline than bursting and rupture. Since the failure probability of the corroded pipeline increases with increased time of exposure, it is expected that this model will be viability for the integrity management of internally corroded pipelines.