PR-244-193800-R01 Quantification of ILI Sizing for Severe Anomalies

PRCI research project, EC-4-6 Improving Corrosion ILI Sizing Models for Long Complex Corrosion Anomalies (Koduru 2018), performed by C-FER Technologies (1999) Inc. ("C-FER"), involved collecting in-ditch high-resolution laser scan records, identifying complex corrosion anomalies, and conducting a burst pressure assessment considering inline inspection (ILI) sizing errors and the effect of interaction rules. A methodology for classifying complex corrosion using high-resolution laser scans was developed in that project. In addition, a secure database was developed to store the anonymized high-resolution laser scan data collected during the project. The project described in this report expanded the database with more than 100 high-resolution laser scans and associated ILI data. Severity criteria based on the corrosion cluster attributes that represent the corrosion morphology were developed using the expanded dataset. Exploratory data analysis was conducted to identify the relationship between ILI sizing errors and corrosion cluster attributes. Sizing errors were defined as a function of cluster dimensions, and sizing correction models were developed to estimate the actual maximum depth and length from the ILI maximum depth and length. The correction models were validated against matched pairs of laser scan and ILI severe corrosion clusters. Applying sizing corrections to severe clusters is shown to reduce the error in the estimated burst pressure capacity.