Advancement of Probabilistic Analysis of Seam Weld Cracking Integrity Management

Probabilistic analysis has been increasingly used in the pipeline industry to provide insight on reliability as input to quantitative risk assessments. This paper details an overview of recent advancements in its application to seam weld cracking threat. Three major aspects are covered. The first aspect is material property. There are limited fracture toughness data outside of Charpy V-Notch (CVN) tests available in the industry for vintage electric resistance welded (ERW) seams. A large number of testing has been conducted to better understand pre-1980 ERW pipe fracture behavior. The database includes conventional toughness tests such as 467 compact tension (CT), 20 single edge notched bend (SENB) and 106 pin-loaded single edge notched tension (SENT) specially designed to simulate the constraint condition of full scale pipe. The distributions presented in the paper characterize the fracture toughness of pre-1980 ERW pipe and provide sufficient information to represent this class of material for probabilistic analysis. The second aspect is the use of validation spools for in-line inspection (ILI) performance evaluation. Nondestructive evaluation (NDE) data collected during excavations are a traditional source to validate the tool run. Recently, validation spools are manufactured to quantify the ILI performance by designing configurations with a variety of “synthetic” flaws having accurately known sizing information. The use of these spools enables the rapid and comprehensive assessment of ILI performance by applying API STD 1163 Level 3 analysis. It also provides the opportunity to determine probability of detection (POD) as a function of flaw size, which is critical to interpret the residual failure probability after running ILI and completing repairs. The third aspect is to leverage the accumulated industry flaw data to infer a postulated population to characterize the condition of a pipeline without historical ILI. This approach is beneficial in a few scenarios, for instance, to prioritize assets for ILI runs, to evaluate difficult-to-pig pipelines, and to estimate the probability of failure (POF) of pipelines only subject to hydrotest mitigation. The paper provides some case study to illustrate the implementation of the process and debriefs the ongoing industry progress.