Combining High Resolution In-Line Geometry Tools and Finite Element Analysis to Improve Dent Assessments

NOTE: This article was originally presented at the 26th Pipeline Pigging & Integrity Management Conference in Houston, TX, February 12-13, 2014.

Introduction

Historically, regulations regarding dent severity have been governed by one of two metrics: dent depth or strain. In the case of the former, plain dents with a depth up to 6% of the nominal diameter are permitted in both gas and liquid pipelines [1, 2]. However, many operators typically set stricter limits on dent depth targeting those above a depth of 2% for evaluation. Dent depth provides a straightforward means for assessing dents, as the concept relies on the generally accepted principle that, all things being equal, deeper dents should be more severe than shallower dents.

Strain-based calculations provide another means to assess dents and have become more common as geometry tools have improved and the data necessary to run the calculations has become readily available. In the case of gas pipelines, plain dents of any depth are considered acceptable provided the strains do not exceed 6% [1]. The strain-based approach calculates the strain in the hoop and axial planes of the dent based on the radii of curvature in each plane and the extensional strain based on the length of the dent. An approach is outlined in Appendix R of B31.8 [1]. While the methodology is based on first principles and the approach accounts for the shape of the dent, the application is less straightforward. Estimates of the radii of curvature can be sensitive to undulations in ILI data; typically requiring some form of smoothing or filtering in order to be successfully used. Furthermore, in many cases, the radii of curvature in any plane may vary considerably depending on whether local point-to-point curvature is calculated or if the global shape is considered.

Both the strain-based and dent depth approaches have similar shortcomings. First, neither approach is adequate for complex dents or in cases where interacting dents may be present. In the case of depth, the shape of a dent is completely neglected. A long, deep dent is not distinguished from a shorter, steeper dent. While strain-based approaches improve on this shortcoming and can be useful for well-behaved dents, applying the methodology where varying curvatures may exist in a complex dent becomes significantly more difficult. Furthermore, neither approach directly addresses fatigue, which is a significant concern in dents with pressure cycles.

Finite element analysis (FEA) provides a more adequate means for analyzing dents. When combined with today’s advanced geometry in-line inspection (ILI) tools and advances in computing, FEA can readily be used to assess dents. FEA does not suffer from the shortcomings of the aforementioned methods. Complex dents and well-behaved dents are both suitable for FEA, and the results are not sensitive to small undulations in data. The severity is calculated directly based on the response of the dent to the applied loading, regardless of shape or size.  

This article describes the tools that allow for the most advanced dent assessments and provides a comparison of anomalies analyzed according to depth, strain, and stress concentration factor (i.e., FEA). Finally, a case study is presented illustrating the effectiveness of the method.