Abstract
Selective seam weld corrosion (SSWC) occurs when a susceptible long seam, such as low-frequency electric resistance welding (LF-ERW) manufactured prior to 1970, is subjected to an active corrosion environment. When this occurs, the seam weld corrodes more aggressively than the pipe body, resulting in a deep V-shaped groove aligned with the seam axis. An SSWC anomaly poses a greater threat to pipeline integrity, when compared to a similar volume of general corrosion crossing the long seam (CCLS), due to the aggressive corrosion depth growth rate and orientation to the primary stress.
The pipeline industry has struggled to consistently distinguish between SSWC and CCLS anomalies. In an effort to overcome this gap, T. D. Williamson (TDW) participated in a DOT project that resulted in the development of a SSWC classifier. The classifier leverages the data collected by the Multiple Dataset (MDS) platform. MDS incorporates 5 primary technologies (high-field axial magnetic flux leakage, high-field spiral/helical magnetic flux leakage, low-field axial magnetic flux leakage, high resolution geometry, high resolution mapping), overcoming gaps in each individual technology, providing a comprehensive integrity assessment. The classifier has since been used with various operators to characterize long seam corrosion anomalies.
In 2022, TDW published an industry first specification that characterized the performance of the SSWC classifier. The paper will discuss the statistical backing of the published specification, demonstrating the field data basis and validation. Examples of SSWC and non-SSWC feature classifications will be discussed.
Keywords: In-line Inspection, Corrosion, Corrosion Crossing the Long Seam, Data Science, Multiple Dataset, Magnetic Flux Leakage, Geometry, Deformation, Interacting Features, Precision & Recall, Selective Seam Weld Corrosion, CCLS, DEF, GEO, ILI, MDS, MFL, SMFL, SSWC
Available after the event



