Sometimes innovation starts with a customer question that does not have an obvious answer. For NDT Global, that question was whether peaking - a subtle manufacturing feature along a pipeline's longitudinal seam could be identified and measured reliably enough to help operators understand where it may contribute to integrity risk. The challenge was not simply detecting geometric irregularities. It was finding a reliable way to characterize the feature and, importantly, understand what it could mean for crack detection in the same location.
What followed was a multi-year development journey that went beyond solving the technical challenge of measuring peaking. Alongside advances in inspection technology and analysis, the team had to integrate a new process into Data Science operations within six months, establishing the resources, training, quality control procedures and reporting workflows needed to deliver the service consistently. This required close collaboration across data science, technology, software, analysis and operations teams, as well as a fundamental rethink of how peaking could be measured and assessed.
Why Peaking Matters
Peaking is a manufacturing-related deviation from a pipeline's ideal circular shape, typically occurring along the longitudinal seam.
While the feature may be relatively small, its presence can become significant when combined with repeated pressure cycling. As internal pressure changes, the pipe responds to those changes in geometry and stress. In some circumstances, this can contribute to fatigue cracking developing at or around the seam.
For pipeline operators, knowing where peaking exists and understanding its characteristics can therefore provide important context when assessing integrity. But measuring it accurately presents a fundamental challenge.
At first, measuring peaking might seem relatively simple: determine how far the pipe deviates from its ideal circular shape and report the height of the feature. In practice, however, a pipeline is not perfectly round.
This creates a difficult question: Where exactly is the center of a pipe that is not completely circular?
The answer can influence the measured height of the peaking feature. Sensor alignment and movement can also affect the measurement. The team therefore began looking at the problem differently.
Rather than relying solely on peaking height, researchers explored the use of the angle formed around the longitudinal seam to characterize the feature. By comparing the geometry on either side of the weld, the methodology could determine the angle associated with the peaking. This approach reduced the dependence on establishing an assumed center for the pipe and provided a more robust way to characterize the feature.
From Research to a Customer Solution
Existing inspection data could provide valuable information about peaking, but applying the approach across an entire pipeline presented additional challenges. Early inspection technologies, using ultrasonic wall thickness measurements data (UM) had limitations around circumferential resolution and the way the inspection tool interacted mechanically with the longitudinal seam. In some circumstances, the tool could roll onto the weld, changing the inspection angles and making it difficult to consistently assess peaking.
This meant that while the methodology could provide valuable information, there were areas where coverage could not be achieved with the level of confidence wanted. The customer requirement provided the motivation to take the concept further. Instead of treating those limitations as a reason the problem could not be solved, the team began looking at whether an existing technology platform could provide the necessary ingredients for a different approach.
How PROTON™ Changed the Approach
With the introduction of PROTON™, NDT Global's phased-array technology delivering accurate detection and sizing of combined pipeline defects brought several advantages to the challenge. The mechanical design was better suited to maintaining the required inspection geometry around the longitudinal seam, while its higher circumferential resolution provided greater accuracy for characterizing the pipe.
The team identified circumferential resolution of around 2 mm or less with PROTON, compared with approximately 4 mm for the earlier approach discussed during the development.
But moving the methodology to PROTON was not simply a matter of applying an existing algorithm to a new dataset. The volume and complexity of the data required new analysis methods, software development, and workflows. The team also needed to understand how the inspection should be adapted to the geometry being assessed.
This led to an important development in the approach. Rather than inspecting against an assumed perfectly round pipe, the inspection parameters could be adapted to the actual geometry of the pipe at the peaking location.
In practical terms, the team could first characterize the peaking and then adjust the inspection approach to account for it when assessing cracks.
This created the foundation for a more complete assessment: identifying and characterizing peaking while also improving the ability to detect cracks associated with those locations.
Looking Beyond the Geometry
For the team, this was an important distinction. Operators are not necessarily interested in peaking simply because a pipe is not perfectly round. The more important question is what geometry could mean for pipeline integrity.
Peaking can affect the conditions under which crack inspection is performed because changes in geometry can influence inspection angles. With PROTON, the team developed an approach that accounts for the local pipe geometry when configuring the inspection.
The objective is to maintain the required probability of detection for cracks at peaking locations rather than treating those areas in the same way as a perfectly round section of pipe. Testing has also included artificial peaking and notches on customer pipe samples, providing additional information on how the inspection responds to these features.
This work is helping the team understand not just where peaking occurs, but how the presence of peaking can be incorporated into a broader assessment of cracking risk.
Developing the technical concept was only part of the challenge. Turning a research approach into a service capable of analyzing hundreds of kilometers of pipeline and tens of thousands of pipe sections required a significant effort across the organization.
Data scientists developed and refined the analysis methodology, Technology Services worked on inspection configuration and tool development, Software specialists built the functionality needed to process the data, while the Data Service Team focused on the analysis and the practical requirements for deploying the service on a large scale.
The development also highlighted an important reality of innovation: solutions evolve as they are applied in the field.
Early inspection data revealed requirements that had not been anticipated during the initial development. Rather than treating these as a reason to stop, the team adapted the software and analysis approach to work with the available data and continued refining the methodology.
That iterative process has been an important part of moving the capability from research into practical applications.
Today, the peaking capability represents more than a new measurement technique. It demonstrates how customer challenges can drive innovation across technology, data science, and operations teams. By combining advanced inspection technology with new analytical approaches, NDT Global has created a practical way to characterize peaking and incorporate it into a broader assessment of pipeline integrity. As the methodology continues to evolve through field application, it offers operators deeper insight into the relationship between pipeline geometry, crack detection, and long-term asset reliability.
