An electromagnetic acoustic transducer (EMAT) is an ultrasonic sensor that, unlike conventional piezoelectric ultrasonics, does not require a liquid couplant to inspect and characterize crack-like defects in steel. Instead, ultrasonic waves are induced in the steel using an electro-magnetic field. Not requiring a liquid couplant makes EMAT In-Line Inspection (ILI) tools suitable for inspecting gas transmission pipelines, and evaluating natural gas pipelines subject to the Stress Corrosion Cracking (SCC) threat. In this project, a large high-quality data set of EMAT and field-measured SCC in the...
An electromagnetic acoustic transducer (EMAT) is an ultrasonic sensor that, unlike conventional piezoelectric ultrasonics, does not require a liquid couplant to inspect and characterize crack-like defects in steel. Instead, ultrasonic waves are induced in the steel using an electro-magnetic field. Not requiring a liquid couplant makes EMAT In-Line Inspection (ILI) tools suitable for inspecting gas transmission pipelines, and evaluating natural gas pipelines subject to the Stress Corrosion Cracking (SCC) threat. In this project, a large high-quality data set of EMAT and field-measured SCC in the pipe body was collected from various operators. The sizing performance of the EMAT technology, both from the depth and length perspective, was characterized using the Level 3 methodology described in American Petroleum Institute (API) Recommended Practice (RP) 1163 (1). However, one of the known challenges with API RP 1163 is that it does not capture variability in field measurement between different techniques.
To address this challenge, the data set was evaluated with a partial pooling model to estimate the underlying ILI EMAT performance. The model characterizes performance by partially sharing the underlying ILI EMAT performance across the data subsets, with each subset containing confounders known to influence measurement error, as well as confounders to capture the random error not necessarily captured in the available data. Crack-like anomalies are a complex phenomenon, with non-uniform profiles, axial positions, morphologies, and interacting elements. It’s quite reasonable that two processes accurately measure different parts of the same anomaly resulting in different, but accurate measurements of the same phenomena. As a result, this project came up with a list of recommended best practices for field data collection, to ensure that the data used to validate tool performance is collected consistently and accurately, and that the variability between different operators is reduced.
This document aims to help integrity and Non-destructive Examination (NDE) specialists incorporate these guidelines into their procedures, processes, or specifications to improve in-ditch data collection for the purpose of validating EMAT technology performance.