Quantitative risk analysis for linear infrastructure supported by permafrost : methodology and computer program

Authors: Brooks, Heather
Advisor: Doré, GuyLocat, Ariane
Abstract: Permafrost is ubiquitous in the Arctic and Antarctic, and present in high elevation regions throughout the world. The communities and industrial development in permafrost regions require transportation infrastructures (roadways, airports, railways, etc.) and, in these regions, transportation is of vital social, economic, and political importance (Regehr, Milligan, and Alfaro 2013). However, warming climate conditions will endanger existing and future transportation infrastructure in Alaska to the tune of $282 to $550 million (2015 USD) depending on future emission scenarios (Melvin et al. 2016). Given these conditions, tools are required to aid decisionmakers in prioritizing infrastructure maintenance, replacement, and construction, and potentially justifying the use of mitigation strategies of permafrost embankments. Risk analysis methods can be used but their existing application to permafrost engineering is limited. Risk is a product of hazard, consequence and vulnerability for each of the dangers under consideration. The probability and costs of a danger’s occurrence is a hazard and the consequence, respectively, while vulnerability correlated the damage with the consequence. Since little failure data is available for permafrost infrastructure, the hazard must be determined from reliability analysis methods (First-Order Second-Moment or Monte Carlo Simulation), which aggregate the uncertainty of input parameters to determine the result’s variation. These methods require the characterization of random variable uncertainty, which can be difficult without sufficient data, often more than the current standard-of-practice. Additionally, the method requires a limit state function for the danger to be analyzed. Common dangers effecting permafrost embankment infrastructure included: settlement, cracking, sudden collapse, lateral embankment spreading, drainage and ponding water, and active layer detachment landslides. Of these dangers, only a few have existing limit state functions or have limit state functions that can be developed by the author. The dangers with limit state functions or hazard functions include: total and differential thaw settlement, particle position bridging over voids, active layer detachment landslides, and culvert gradient and structural failure. A Microsoft Excel-based program was created to calculate the risk for permafrost embankment linear infrastructure, using statistical methods applied to limit state functions to determine hazards for common permafrost dangers, estimated direct costs for the repair of a hazard’s occurrence, and scaling factors to account for the indirect costs of damage to the infrastructure’s users and connected communities. Hazard calculations are based on geotechnical property and climate uncertainty, as characterized by probability density functions, using Monte Carlo Simulation methods. A climate change fragility analysis recalculates the hazard with warming air temperatures. Repeated analyses along the infrastructure provide a risk profile of the infrastructure, now and with a warming climate. The program is used to determine hazard for the Airport Access Road in Salluit, and hazard, risk and cost/benefit assessments were conducted using this program for the Iqaluit International Airport.
Document Type: Thèse de doctorat
Issue Date: 2019
Open Access Date: 31 October 2019
Permalink: http://hdl.handle.net/20.500.11794/37155
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
Description SizeFormat 
34674.pdf22.69 MBAdobe PDFThumbnail
All documents in CorpusUL are protected by Copyright Act of Canada.