PR-185-163609-R01 Model Error Assessment for Pipeline Metal-loss Defects Phase II

Abstract

This project included seven, full-scale burst tests conducted on a single grade and size of pipe to evaluate a metal-loss defect assessment model developed under EWI Project No. 54922CSP, "Assessment of Corrosion Model Error for Metal Loss Defects in Pipelines" (PRCI Project EC-2-6). The goal of the modeling and full-scale test efforts was to minimize error and reduce the conservatism of Level-1 corrosion assessment models, with the eventual benefit of potentially reducing unwarranted maintenance efforts without increasing operator risk. Seven tests were performed on a recent vintage of API 5L X70M (L485M) grade pipe having an outside diameter of 24 in. (610 mm) and nominal wall thickness of 0.50 in. (12.7 mm). The seven test pipes contained machined flat-bottomed defects for which the Shape Factor is 1, to focus on the role of the planar defect size and its depth and avoid the complexities of river-bottom shaped defects that are beyond the current scope. Defect lengths ranged between 3 and le; L/(Dt)0.5 and le; 10, at bounding values of width taken at arc angles of 20 and 60 degrees, and depth at d/t of 30% and 80%. The simulated corrosion was centered in pups approximately 12 ft (~3.7 m) in length, which is sufficient to prevent any interaction between the end caps with the defects. The pipes were then end-capped and pressurized with water to failure. The actual failure pressures were evaluated relative to predictions based on the model developed in EC-2-7 coupled with the reference stress developed in PRCI Project EC-2-6.