With the age of the original Panhandle Eastern Pipeline (PEPL) Company pipelines, it’s not a matter of if anomalies will be found when an ILI tool is run, it’s a matter of how many and how severe. When a final report is received from an ILI vendor, burst pressures are typically calculated using Modified B31G, 0.85dL. The results can seem unmanageable, but success has been had doing further assessments on some anomalies without excavating them all. This assessment has been developed and performed by PEPL on three sets of Tuboscope ILI data and one set of Baker Hughes CPIG data. The method to be discussed was first employed in 2002. It provides a more accurate characterization of the defect and provides the company the ability to more effectively allocate resources. Efforts have been made to review the color scan of a vendor’s raw High Resolution Magnetic Flux Leakage (HRMFL) data, and perform an assessment using Effective Area Analysis without excavating hundreds of anomalies that prove no threat to the pipeline. This assessment is done by hand on the computer and in many cases returns a burst pressure higher than that calculated using Modified B31G, 0.85dL. The following is a case study that shows how multiple defects have been assessed prior to excavation in an attempt to more accurately characterize the defect, and allow for a better allocation of resources. Digs have been performed to validate the process, and the results will be discussed.
Magnetic Flux Leakage (MFL) based Defect Characterization of Steam Generator Tubes using Artificial Neural Networks
