ILI Validation: what are we trying to prove?
Inline Inspection is one of the foundational components of most pipeline integrity programs. ILI provides the condition monitoring data that allows for an inspect, assess and then remediate cycle that allows for the pipeline anomalies to be managed successfully.
As the ILI will normally identify many more anomalies than will be field investigated each anomaly identified is assessed to estimate its severity. Those results determine if and when an anomaly will be investigated. Repair decisions are made based upon the results of the site investigation. So, the primary purpose of the ILI is to populate the integrity assessment with data that can be assessed effectively. In this context, effective means that the data is reliable and accurate enough to perform a conservative assessment.
But the assessment methods to assess metal loss and cracks for both leak and rupture can be applied with varying levels of conservatism. For example burst pressure calculations can be performed with or without tool accuracy tolerances and different confidence levels are included in most tool specifications that can be utilized. The way the assessment is applied should be closely tied with desired levels of conservatism. The more conservative the assessment the more digs that will be done.
The ILI validation should be trying to determine if the quality of the data met the goals of the assessment not simply whether the ILI tool meet its specifications. In other words it should be the integrity assessment itself and not the ILI tool specifications that are being validated.
The accuracy of an ILI tool predications is defined as the difference between the actual measurements and predicted. The validation process needs to compare the accuracy used in the assessment with the actual accuracy. If the tool meets the predicated accuracy then the assessment results are validated. If not then the assessment needs to be updated with the measured accuracy.
If a tool did not meet the accuracy used in the original assessment, usually based upon the vendor ILI specifications, and the assessment is re-run with a larger tool tolerance extra digs may be identified. Now the decision as to the usefulness of the ILI tool run is based upon a cost benefit of an ILI rerun versus the extra digs to accommodate the decreased accuracy of the tool.
Probability of Detection
Anomaly re-inspection intervals need to be checked to ensure that anomalies won’t grow to critical size between inspections. Typically corrosion growth analyses, SCC growth rates, and crack pressure cycle fatigue analyses are used to determine critical anomaly sizes based upon future growth and compare to inspection intervals.
The reliability of the inspection results need to be analyzed to ensure that anomalies of critical size are being reliably detected. This can be done through review of false negatives, anomalies found in the field not identified by the ILI, or undercalls, anomalies significantly underpredicted by ILI. The goal of the review is to determine the largest anomaly not being reliably detected by the ILI and comparing it to the critical anomaly sizes.
Similar to accuracy the goal of validation is to review and adjust the inspection frequency accordingly not simply to compare largest remaining anomaly size with the tool specification.
Inspection Program Validation Questions
To ensure the effectiveness of an ILI based pipeline inspection program the ILI Validation process focus on two questions:
- Was the tool performance accuracy sufficient for the assessment? If not reassess the results with the updated accuracy and re-evaluate.
- Were critical anomalies reliably detected. If not reassess the inspection interval.