This paper and accompanying talk were presented at the 37th International Pipeline Pigging and Integrity Management (PPIM 2025).
Introduction
The global demand for hydrocarbons continues to grow, placing increased pressure on pipeline infrastructure. Pipelines serve as the primary mode of transport for oil and gas, spanning continents and oceans to deliver energy to consumers. Despite their efficiency, pipelines are susceptible to failures, which can have far-reaching consequences. A notable example occurred in 2016 in North America, where a pipeline failure resulted in the release of 2,000 metric tons of hydrocarbons, causing environmental damage, economic losses, and public safety concerns [1].
The integrity of pipelines has been a subject of ongoing research and technological development. Traditional inspection tools, such as magnetic flux leakage (MFL) and ultrasonic devices, have been the mainstay of pipeline integrity management since the 1980s. These tools have proven effective in detecting corrosion, cracks, and other forms of damage. However, they face limitations when inspecting pipelines with complex geometries, such as tight bends, diameter changes, or non-circular valves [1,2]. These challenges are particularly acute for “unpiggable” pipelines—those that cannot accommodate conventional inspection tools. In the United States alone, approximately 70% of gas lines constructed before modern inline inspection (ILI) technologies were developed are considered unpiggable, representing up to 40% of pipelines in service as of 2012 [3,4].
Addressing the challenges of unpiggable pipelines requires innovative solutions. Advances in free-floating inline inspection tools since the early 2000s have opened new avenues for pipeline monitoring. These tools, which operate independently of pipeline geometry, are capable of detecting leaks [5,6], reconstructing pipeline paths [7,8], and assessing wall conditions [9,10]. Among their many applications, the use of remnant magnetometry for weld inspection is a promising area of research. This paper focuses on the application of principal component analysis (PCA) to remnant magnetization data, demonstrating its potential to identify weld anomalies and improve pipeline integrity management.
The role of free-floating inline inspection tools
Free-floating inline inspection tools have emerged as a versatile solution for monitoring unpiggable pipelines. Unlike traditional pigging devices that require continuous contact with the pipeline interior, free-floating tools navigate pipelines autonomously, collecting data as they move through the system. This capability allows them to operate in pipelines with complex geometries, including those with sharp bends, varying diameters, and non-standard features.
Material heterogeneity and the challenges of weld inspection
Welds are critical components of pipeline systems, serving as the junctions between pipe segments. However, they are also the most common locations for failures. The welding process introduces material heterogeneity, particularly in the heat-affected zone (HAZ) adjacent to the weld. This zone experiences significant thermal and mechanical changes, including grain coarsening, phase transformations, and residual stresses.
The HAZ is particularly susceptible to stress-induced cracking and corrosion, making it a common failure point [11]. Additionally, weld defects such as porosity, incomplete fusion, and slag inclusions act as stress concentrators, further increasing the risk of failure [12]. Detecting these defects is a key objective of pipeline integrity management.
While traditional MFL devices have made significant strides in weld inspection, they face limitations in handling pipelines with complex geometries. Remnant magnetometry offers an alternative approach by analyzing the residual magnetic signatures left by welding and other stress-inducing processes. These signatures provide valuable insights into the structural integrity of pipeline joints, enabling the identification of potential defects.
Principal component analysis: a framework for dimensional reduction
The analysis of remnant magnetization data generates large datasets, particularly for long pipelines. For example, magnetometers on free-floating tools sample at frequencies of up to 1000 Hz, generating hundreds to thousands of data points per weld signature. Managing and interpreting such data requires effective dimensional reduction techniques.
Principal component analysis (PCA) is a statistical method that transforms high-dimensional data into a lower-dimensional space while preserving as much variance as possible. By performing a singular value decomposition on mean-centered data, PCA identifies a set of orthogonal eigenvectors, or “principal components,” that capture the most significant patterns in the dataset.
In this study, these principal components are referred to as “eigenjoints,” representing the magnetic signatures of pipeline joints. The explanatory power of each eigenjoint is quantified by its singular value, which provides a measure of the variance explained by that component. Cumulative energy plots visualize this information, illustrating how variance is distributed across the eigenjoints (Figure 1).
For the pipeline analyzed in this study, PCA revealed that the first eigenjoint captured 65% of the variance in the dataset, while the first 20 eigenjoints explained 98% of the variance. This finding highlights the effectiveness of PCA in reducing data complexity while retaining essential features.
By expressing each joint as a linear combination of eigenjoints, it becomes possible to classify welds based on their contributions to the most significant eigenjoints. Outlier joints, characterized by unusual magnetic signatures, can be identified by analyzing deviations in eigenjoint contributions (Figures 2–4).
Results
The analysis focused on a pipeline segment containing 321 joints. Using PCA, the joints were categorized based on their eigenjoint contributions. Outliers were defined as joints whose contributions to the first two eigenjoints exceeded two standard deviations from the mean.
Outlier joints exhibited distinctive baseline offsets, a feature predominantly captured by the first eigenjoint. This baseline offset accounted for 65% of the variance in the dataset, demonstrating the power of PCA to identify key features without manual inspection.
Discussion
The application of PCA to remnant magnetometry data offers several advantages for pipeline integrity management:
- Scalability: PCA is well-suited for analyzing large datasets, making it ideal for pipelines spanning hundreds or thousands of kilometers.
- Efficiency: By reducing the dimensionality of the data, PCA simplifies the classification of pipeline joints, allowing inspectors to focus on anomalies.
- Insight: PCA quantifies the variance explained by specific features, providing a deeper understanding of pipeline joint morphology.
Future work will explore the integration of PCA with machine learning algorithms to automate anomaly detection. Techniques such as clustering and supervised classification have the potential to enhance the accuracy and reliability of joint classification. Additionally, expanding the dataset to include diverse pipeline materials and configurations will improve the generalizability of this approach.
Conclusion
This study demonstrates the potential of PCA in analyzing remnant magnetic signatures of pipeline joints. By enabling efficient dimensional reduction and classification, PCA provides a powerful tool for managing unpiggable pipelines and detecting weld anomalies. As the field of remnant magnetometry continues to evolve, the integration of PCA with advanced analytics will further enhance pipeline integrity management.
References
[1] "A review on pipeline integrity management utilizing in-line inspection data," Engineering Failure Analysis, vol. 92, pp. 222—239, 2018.
[2] "An overview of major methods for inspecting and monitoring external corrosion of on-shore transportation pipelines," Corrosion Engineering, Science and Technology, vol. 50, pp. 226—235, 2015.
[3] "Report to the National Transportation Safety Board on Historical and Future Development of Advanced In-Line Inspection Platforms for Use in Gas Transmission Pipelines," 2012.
[4] "Ultimate Guide to Unpiggable Pipelines," 2013.
[5] "SmartBall: a new approach in pipeline leak detection," International Pipeline Conference, vol. 48586, pp. 117—133, 2008.
[6] "Leak detection and prevention using free-floating in-line sensors," Pipeline Pigging and Integrity Management, 2019.
[7] "The Future of In-Line Inspection: Free-Floating Smart Sensors," International Pipeline Conference, vol. 51869, pp. V001T03A021, 2018.
[8] "Miniature Sensors for Monitoring the Operational Conditions of Pipelines," Offshore Technology Conference, pp. OTC—23186, 2012.
[9] "Design of A Free Floating Ultrasonic Inner Spherical Detector for Pressure Pipeline," Journal of Physics: Conference Series, vol. 1948, pp. 012092, 2021.
[10] "Defect localization using free-floating unconventional ILI-tools without AGMs," Pipeline Pigging and Integrity Management, 2022.
[11] "Characterization of SA333 and SA53 steel pipe weld joints," AIP Conference Proceedings, vol. 2395, 2021.
[12] "A rapid method of inspecting welded joints utilising the magnetic memory of metal," Welding International, vol. 11, pp. 410—413, 1997.