ROSEN’s newly developed Automated Signal Comparison And Normalization (AutoSCAN) software package played an essential role in the timely and successful completion of a project for a client in the Middle East.

 

AutoSCAN is a new system that enables repeat ROSEN MFL A ILI run data to be compared on a signal to signal basis.  The objective is to accurately identify and quantify changes for all metal loss features between inspections as part of a corrosion growth assessment (CGA) service.  This process offers far greater accuracy than traditional feature (‘box’) matching and provides greater confidence in the amount and rate of corrosion growth.

In this particular case, the contract comprised an MFL A inspection as well as a fitness for purpose (FFP) assessment of a large diameter crude oil pipeline. The FFP study was to include a detailed assessment to determine if the line was safe for continued operation as well as a schedule for reinspection based on accurate corrosion growth rates. Given the criticality of the pipeline, the client needed the FFP study to be delivered within a short period of time.

In order to predict the time at which the detected corrosion features would exceed tolerable dimensions, growth rates were calculated using depth changes between inspections measured by AutoSCAN. To achieve this, the two sets of inspection data were aligned, calibrated, and normalized to eliminate bias errors between the runs. The signal response from all metal loss features present in the signal data from the first ILI was matched using pattern recognition algorithms to the corresponding signal response in the new signal data. A single sizing model was applied to both sets of normalized signal data in order to calculate the amount of depth change between inspections for all metal loss features. Hence the past corrosion growth rate could be calculated accurately for each feature. This allowed experienced corrosion and integrity engineers to select justifiable representative growth rates for planning future inspections.  The pattern matching algorithms and calculated depth changes were manually validated for a large sample of features.

The study found significant levels of active corrosion and provided sufficient justification for the client to plan for an additional repeat inspection. Furthermore, a review of the distribution of corrosion features, growth rates, and operational data enabled a diagnosis of likely corrosion mechanisms that had led to the observed damage. Satisfied with the overall service, the client requested that the inspection be followed by a further combined AutoSCAN and FFP study.

Applying AutoSCAN to this third set of ILI data will ensure the required accuracy when re-assessing the growth of the detected corrosion features over the short inspection interval. This repeat corrosion growth monitoring is essential for the determination whether the revised corrosion management strategy was effective in reducing the high corrosion rates that were previously calculated.  These high rates currently threaten the safe operational life of the pipeline.

In conclusion, using AutoSCAN as part of a detailed integrity study will allow the client to better understand and manage the integrity of their critical pipeline.