Experiences with qualification of pipeline girth weld AUT

Speaker:
Hasting, Håkon; DNV GL AS; Norway

Authors:
Hasting, H.S.; DNV GL AS,; Norway

ID: ECNDT-0172-2018
Download: PDF
Session: NDT Reliability 3
Room: J1
Date: 2018-06-15
Time: 10:50 - 11:10

Systems for automated ultrasonic testing (AUT) has been applied for inspection of pipeline girth welds in the offshore pipeline industry for more than two decades. It is the most widely used NDT method for this application due to factors like efficiency, safety and good integration with risk based pipeline design. Use of AUT has been required in the DNV-OS-F101 standard for offshore pipelines since the 2000 edition. This requirement has been accompanied by a requirement to qualify each applied AUT procedure with regards to Probability of Detection (POD) and sizing accuracy. A guideline to provide consistent qualification practice was formalised with DNV-RP-F118 in 2010. DNV GL has gathered experience through numerous AUT qualification programs using this recommended practice over the recent years.

It has been experienced as challenging to derive conclusions of truly general validity out of these AUT qualification programs. The main challenges are considered to be the complexity in these AUT systems, with redundancy in detection and inspection coverage by several channels at a time, and that there are multiple essential variables that has significant impacts on the inspection performance at the same time. The design of the AUT qualification study can often be identified to impact on the result of the study as well. Efforts to reduce the effect of sub-optimal design of the study are often restricted due to practical arrangements, cost and schedule. Similarly, this can limit the possibilities to precisely evaluate the reliability of the inspection through statistical POD assessments of the data collected through the studies. Accurate knowledge about POD for AUT procedures is considered to have increasing value when there is a need to better utilise the materials. This paper discuss how maximum information about the AUT performance can be drawn out of AUT qualification results within DNV-RP-F118 guidelines, based on experience over the recent years.