The goal of this project is to develop inspection and analysis methodologies using circumferential magnetic flux leakage (MFL) that improves the accuracy of detection and sizing of pipeline cracks. The project will combine axial and circumferential MFL in a novel fashion in order to develop algorithms for reducing the effects of magnetic noise during signal processing. The project will also combine high and low magnetization technology, like that done for mechanical damage, in order to develop algorithms for detecting stress patterns around cracks. Finally, the project will combine these algorithms with data analysis methods to develop more reliable detection methodologies for cracks in pipelines.

Battelle Columbus Laboratories – project management and research product

Location:
Columbus, Ohio 43201

Magnetic flux leakage (MFL) tools are the most common inspection tools used by natural gas and liquid pipeline companies. However, conventional MFL tools have significant limitations in their ability to reliably detect and size axially aligned cracks. This inability to detect long narrow defects is not so much a result of MFL technology as of the standard method of implementing the technology. Circumferential MFL is a new approach that has the potential to detect and quantify axially oriented defects such as cracks, seam weld defects, mechanical damage, and groove corrosion.

Circumferential MFL works by orienting the magnetic field around the pipe (circumferentially) rather than along the axis. This allows axial defects that might otherwise seem magnetically transparent to disrupt more of the magnetic field and be more easily detected. Initial applications of circumferential MFL have found that flux leakage from cracks is small, and the signals from cracks at the interior of the pipe are difficult to detect. In this research effort, two technology enhancements to improve detection of cracks using MFL technology were investigated: (1) combining high- and low-magnetization technology for stress detection and (2) combining axial and circumferential MFL methods.

Circumferential MFL can be used to detect corrosion, mechanical damage, and crack defects. However, the detection capabilities and sizing accuracies may not be sufficient for all pipeline threats. Inspection tools that use more sophisticated technologies for detecting and sizing defects may have better performance capabilities, but will likely be expensive to operate. Circumferential MFL will be useful in identifying locations for detailed testing. While performance enhancements may be limited, circumferential MFL inspections will be part of the inspection process for many decades.

• Collected axial and circumferential data at multiple magnetization levels using the Pipeline Simulation Facility pull rig,
• Developed analysis, noise filtering, stress sensing, and signal processing algorithms and established a single, optimized signal-processing methodology, and
• Assessed the capabilities and limitations of signal processing for crack detection using blind data.

The technology developed and tested during this project provides only limited improvements to inspection capability. Although a method combining high- and low-magnetization technology showed promise, its commercial development was not successful for two reasons. First, the stress diminishes the crack signal, while the opening of the crack increases signal. The stress-induced changes in flux leakage around cracks were small and any critical information on the severity of cracks and crack-like defects is difficult to distinguish from changes caused by the crack opening and other inspection variables. Second, it is difficult to magnetize pipe material in the circumferential direction. A relatively low, non-uniform magnetization level produced by the circumferential magnetizer makes detection changes due to stress extremely difficult to detect.

The combination of circumferential and axial MFL to improve crack detection and distinguish cracks for axially oriented volumetric defects was also examined. While successful results were obtained, circumferential MFL can only detect larger cracks. Even with the field aligned properly, circumferential MFL technology has difficulty detecting all cracks on the outside surface that have the potential to grow to failure within a normal inspection interval of five to ten years.

This project is completed and all deliverables have been received.

$184,234

$100,000

NETL – Dan Driscoll (daniel.driscoll@netl.doe.gov or 304-285-4717)
Battelle – Bruce Nestleroth (nestlero@battelle.org or 614-424-3181)

Final Report [PDF-2485KB] June 2003 

Pertinent Publications:
“Innovative Electromagnetic Sensors for Pipeline Crawlers”, J. Bruce Nestleroth and Richard J. Davis, Proceedings of the Natural Gas Technologies II Conference (Phoenix, AZ), February 2004.