Case Study

Beyond the Detection Threshold

Catching Bottomside Pipeline Corrosion Before it Becomes a Release

82% average ranking performance

cross 9 confirmed release locations spanning 8 pipelines and 3 operators, with a perfect ranking score on two pipelines.

Validated on unseen data

Correctly identified one of two releases on a new pipeline's first pass, then both after retuning.

No new data collection required

Built entirely on ILI anomaly data and elevation profiles already available to operators.

Manages an extensive network of gas and liquid transmission pipelines across North America.

Already using Irth's Asset Integrity for Pipelines (AIP) platform for inline inspection (ILI) data management and analysis

Wanted to turn existing ILI data into a proactive, predictive view of release risk, rather than reacting only after the next inspection or incident.

Challenges

Bottom-side internal corrosion forms small, deep pinhole pits that often fall below ILI tools' reliable detection thresholds
No standardized way to combine relevant risk signals, pinhole concentration, depth outliers, active corrosion, valley location, and orientation across a large network
Full theoretical chemical and microbiological corrosion models require operational data most operators don't have
Maintenance prioritization tended to be reactive, following known anomalies instead of anticipating where the next release was most likely

Solution

Development of Irth's data-driven Internal Corrosion (IC) Susceptibility Model

Identifies the pipeline locations with the highest probability of a bottom-side corrosion release, using ILI anomaly data and elevation profiles the operator already had