Competing risks survival analysis of ruptured gas pipelines: A nonparametric predictive approach

Tee, Kong Fah; Pesinis, Konstantinos; Coolen-Maturi, Tahani

Abstract

Risk analysis based on historical failure data can form an integral part of the integrity management of oil and gas pipelines. The scarcity and lack of consistency in the information provided by major incident databases leads to non-specific results of the risk status of pipes under consideration. In order to evaluate pipeline failure rates, the rate of occurrence of failures is commonly adopted. This study aims to derive inductive inferences from the 179 reported ruptures of a set of onshore gas transmission pipelines, reported in the PHMSA database for the period from 2002 to 2014. Failure causes are grouped in an integrated manner and the impact of each group in the probability of rupture is examined. Towards this, nonparametric predictive inference (NPI) is employed for competing risks survival analysis. This method provides interval probabilities, also known as imprecise reliability, in that probabilities and survival functions are quantified via upper and lower bounds. The focus is on a future pipe component (segment) that ruptures due to a specific failure cause among a range of competing risks. The results can be used to examine and implement optimal maintenance strategies based on relative risk prioritization.

Citation

Tee, K. F., Pesinis, K., & Coolen-Maturi, T. (2019). Competing risks survival analysis of ruptured gas pipelines: A nonparametric predictive approach. International Journal of Pressure Vessels and Piping, 175, Article 103919. https://doi.org/10.1016/j.ijpvp.2019.06.001

• Journal Article Type: Article
• Acceptance Date: Jun 5, 2019
• Online Publication Date: Jun 12, 2019
• Publication Date: Aug 31, 2019
• Deposit Date: Jun 13, 2019
• Publicly Available Date: Jun 12, 2020
• Journal: International Journal of Pressure Vessels and Piping
• Print ISSN: 0308-0161
• Electronic ISSN: 1879-3541
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 175
• Article Number: 103919
• DOI: https://doi.org/10.1016/j.ijpvp.2019.06.001
• Public URL: https://durham-repository.worktribe.com/output/1294643