293

Caroline Scheevel1, Jim Hart2, Bailey Theriault3, Casey Dowling4, USA.2, USA.3, USA.4

Caroline Scheevel1, Jim Hart2, Bailey Theriault3, Casey Dowling4
1BGC Engineering, Minneapolis, USA. 2SSD, Inc., Reno, USA. 3Geosyntec Consultants, Inc., Bedford, USA. 4BGC Engineering, Golden, USA

Operators who integrate inertial measurement unit (IMU) survey data into their geohazard management programs often find themselves holding large quantities of data and facing uncertainty about what to do next. Previous attempts to establish triage criteria for vendor-reported strain features have examined the correlation between the presence of mapped landslides and the magnitudes of reported horizontal and vertical bending strains (Theriault et al, 2020; McKenzie-Johnson et al., 2024). However, the relative magnitude of each bending strain component is strongly related to the pipeline’s three-dimensional geometry and orientation relative to the landslide movement vector—landslides acting transverse to the pipeline centerline axis will generate notable horizontal strain signatures, whereas pipe-parallel landslides will tend to generate vertically dominant bending strain signatures.

This paper leverages a database of more than 14k bending strain interpretations across North America to build on the previous efforts. Bending strain features have been classified both by orientation to anticipated ground movement (slope-parallel, slope-oblique, and slope-transverse) and by interpretation (ground-movement impact, uncertain signature, or not ground-movement related). The strains have been evaluated for metadata trends that correlate with ground movement impact, providing an updated triage framework. The framework accounts for the peak bending stain magnitudes and their of the horizontal and vertical strain components and whether their relative magnitudes agrees with the anticipated direction of relative ground movement. This work also evaluated whether strain features with similar metadata are more likely to indicate ground movement impact if they are reported in different geographic regions—for example, is a horizontal strain in the Midwest more likely to be landslide-induced than a horizontal strain in Appalachia? These results, leveraging anonymized data from 50 operators, will help operators prioritize strain features for further assessment regarding geohazard impact, including at sites that may have been deformed by not-yet-identified landslides.

Read/Download PaperGradeAvailable after the event