Pipeline integrity KPIs: which metrics actually indicate program effectiveness

The metrics a pipeline integrity program chooses to track shape what the program actually optimises for - and many of the most commonly reported metrics measure activity rather than effectiveness, which can create a misleading picture of program health even when reported numbers look strong.

The activity-metric trap#

Number of inspections completed, total kilometres surveyed, and count of findings closed are all commonly reported integrity program metrics, and all three share the same structural weakness: they measure how much work was performed, not whether that work found and addressed the risk that actually matters. A program can report a high inspection count while concentrating that effort on easily accessible, low-risk segments and systematically under-inspecting harder-to-reach or genuinely higher-risk locations - the activity metric looks strong throughout, while real risk exposure remains unaddressed and effectively invisible to anyone reading only the summary numbers.

Detection-to-repair time#

A more meaningful indicator is the interval between when a condition becomes detectable by whatever methods are available and when it is actually remediated - combining both how promptly it was found after becoming detectable, and how promptly it was addressed once found. This metric connects directly to the root-cause pattern explored in our piece on why pipeline failures still happen: many documented incidents trace back to a condition that existed and was, in principle, detectable for a meaningful period before it was actually found or acted upon. Shortening detection-to-repair time is a direct lever on exactly that gap, and it is a genuinely outcome-oriented metric in a way that "number of inspections completed" is not.

Data currency relative to risk, not just data existence#

Rather than simply tracking whether a segment has ever been inspected, a more useful metric maps, for each segment, when it was last assessed by each relevant method and compares that against how fast the specific threats most relevant to that segment can plausibly progress. A segment last surveyed five years ago should be flagged very differently depending on whether the dominant risk there is slow general corrosion or fast-developing encroachment risk near active construction - a single "last inspected" date, without that context, understates risk in some cases and overstates it in others.

False positive and false negative rate as core KPIs#

For any inspection method or system that flags findings for engineering review - which increasingly includes fusion-based, partially automated systems - two numbers matter more than an aggregate accuracy figure: false positive rate, which determines how much of a reviewing engineer's limited time is consumed by findings that turn out not to be genuine issues, and false negative rate, which determines how much real risk is going undetected despite a clean report. A system with an impressive-sounding overall accuracy figure can still have a false positive rate high enough to make it a net time cost rather than a net time saver for the engineers using it - see our related piece on false positive rate in inspection software for why this specific number deserves more scrutiny than aggregate accuracy claims typically receive.

Choosing metrics that reflect the program's actual job#

The throughline across all of these more meaningful metrics is that they measure outcomes tied to risk reduction and response speed, rather than volume of activity performed. A pipeline integrity program genuinely improving its risk posture over time should be able to show, and should be asked to show, that detection-to-repair intervals are shrinking, that data currency is improving specifically where risk is highest, and that the false positive and false negative rates of whatever detection methods it relies on are known, tracked, and trending in the right direction - not simply that more inspections were performed this year than last.

Related reading#

This connects directly to the outcome-focused framing in risk-based inspection and to the audit-ready data standards required to actually measure these KPIs credibly rather than just assert them.