Spatial and Time-Dependent Reliability Assessment: Identification of Critical Segments

This chapter presents the results for the reliability framework proposed in Chap. 8 . The framework consists of three main stages: In the first place, new corrosion defects are generated and located on the pipeline. This approach contemplates a Homogeneous Poisson Process (HPP) to determine the number of new defects and the Non-Homogeneous Poisson Point Process (NHPPP) described in the previous chapter to simulate their location on the pipe. The second stage clusters the defects and implements a stochastic degradation process. For this purpose, three clustering criteria and two synthetic learning approaches are compared to identify the “corrosion colonies” and their equivalent dimensions. Regarding the degradation process, a stochastic approach based on Lévy process is implemented to simulate the corrosion evolution and the Mean Time to Failure (MTTF). Finally, the third stage evaluates a combined failure probability under a spatial and temporal assessment, using dynamic segmentation and a failure region.