49 CFR §192.607 defines unique populations for pipeline material verification. These are defined by wall thickness, pipe grade, manufacturing process, manufacturing date, and construction date. Using ILI to identify pipe attributes has the potential to significantly improve the efficiency of satisfying the material verification requirements outlined in 49 CFR §192.607 for delineating unique populations along a pipeline. However, some of the ILI technologies used in this process are relatively new, and ILI vendors do not yet provide a performance specification for their performance in identifying populations.
This analysis explores quantifying ILI tool performance in defining pipe populations by similar heuristic rules. It incorporates them into a probabilistic model specification for ILI population identification for three attributes: wall thickness, pipe seam detection, and yield strength. The models were fit using populations identified by the ILI vendor and validated using records review and field verification by an operator. Models adequate for characterizing the ILI tools, assumptions regarding pipe populations, and heuristics that can be used as proxies for the ILI vendor’s interpretation of measurement data are discussed.
The methodology outlined in this study aims to quantify ILI’s ability to classify transmission pipeline segments into populations with unique material properties to comply with PHMSA’s material verification requirements. The framework uses statistical methods similar to those outlined in API 1163 for ILI measurement validation, which the industry has been using for about twenty years. These models are designed to estimate the accuracy of future population identifications and identify populations that are more likely to be misclassified, thereby informing an operator’s decision-making to prioritize population validation efforts.
Available after the event



