Finite Element Modelling of a Series of Ground Displacement Episodes and Stress Relief Procedures

Abstract

Abstract The Wapiti River South Slope (the Slope) near Grand Prairie, Alberta, Canada, is 500 m long and consists of a steep lower slope and a shallower upper slope. Both the upper and the lower slopes are located within a landslide complex with ground movements of varying magnitudes and depths. The Alliance Pipeline (Alliance) NPS 42 Mainline (the pipeline) was installed in the winter of 2000 using conventional trenching techniques at an angle of approximately 8° to the slope fall line. Evidence of slope instability was observed in the slope since 2007. The surficial geology of the slope comprises a colluvium layer draped over bedrock formation in the lower slope, and glacial deposits in the upper slope. Available data indicated two different slide mechanisms. In the lower slope, there is a shallow translational slide within a colluvium layer, and in the upper slope there is a deep-seated translational slide within the glacial deposits. Both the upper and lower slope landslides have been confirmed to be active in the past decade. Gradual ground displacements in the order of several centimeters per year were observed in both the upper and lower slopes between 2007 and 2012. Large ground displacements in the order of several meters were observed between 2012 and 2014 in the lower slope that led to the first stress relief and subsequent slope mitigation measures in the spring and summer of 2014. Monitoring of the slope after mitigations indicated significant reduction in the rate of ground movement in the lower slope. Surveying of the pipeline before and after stress relief indicated an increase in lateral pipeline deformation in the direction of ground movement, following the stress relief. This observation raised questions regarding the effectiveness of partial stress relief to reduce stresses and strains associated with ground movements. Finite element analysis (FEA) was conducted in 2016 to aid in assessing the condition of the pipeline after being subject to ground displacements prior to 2014, stress relief in 2014, and subsequent ground displacement from July 2014 to December 2016. The results and findings of the FEA reasonably matched the observed pipeline behaviour before and after stress relief in the lower slope. The FEA results demonstrated that while the lateral displacement of the pipeline, originally caused by ground movement, increased following the removal of the soil loading during the stress relief, the maximum pipeline strain was reduced within the excavated portion. The FEA was also employed to assess the pipeline response to potential ground displacement scenarios following December 2016. For this assessment, three ground displacement scenarios that comprise different lengths of the pipeline were analyzed. An increased rate of ground displacement, with a pattern that matched one of the analyzed scenarios, was observed in the upper slope in the spring of 2017. The results of FEA were used to assess the pipeline response to the increased rate of displacement in the upper slope. Subsequently a decision was made to stress-relieve the pipeline. The second stress-relief was conducted in the summer of 2017. This stress relief was conducted locally at the toe and head of the active slide in the upper slope, where the FEA showed the greatest stress concentrations in the pipeline.