Preliminary Considerations of a Risk-Informed Approach for Climate Change Resilient Pipeline Geohazards Management

Abstract Risk-informed management of pipeline geohazards within the bounds of historical climatic records has been adopted widely in the pipeline industry. The emergence of extreme rainfall events, and related indicators of climate change, that fall well outside the historical data pool presents a challenge in terms of reliability of predictions of significant geohazards capable of damaging pipelines. A review is undertaken of several major precipitation events outlying from their respective recorded historical trends that have impacted pipeline assets globally. Several sites are selected to estimate magnitude and frequency of future extreme weather events characterized in the context of different extreme-value statistical distributions. A general approach is evaluated based on the back analysis characterization of this sample of events to assess if and how these events may be prudently treated as part of a risk-informed pipeline geohazard management program. Several aspects of implementing this approach are discussed including its application based on geological provinces and microclimatic zonation, and the integration of rainfall ground-movement models and the associated effect on pipeline strain accumulation. An elaboration of a fuller evaluation and implementation of this approach as part of a machine learning supported pipeline geohazard management workflow to address climate change resiliency of pipelines is discussed. The findings of this paper support focused implementation of an extension of the practice of risk-informed geohazard management accounting for events that may deviate significantly from historical records for select critical high consequence areas of interest. For the purposes of illustration, this paper focuses on landslide hazards.