Finite Element Modelling for Refined Assessment of Subsea Slope Stability

Abstract This paper presents work undertaken at the project front-end stage of reducing to ALARP the risk of subsea flowline rupture from impact of debris flow upon collapse of a slope due to a seismic event. The scope involved soil sample testing, novel finite-element analysis techniques and development of site-specific probabilistic seismic hazard assessment (PSHA). The subsea flowline traverses a route which has steep local slopes in close vicinity. Preliminary seismic slope stability assessment indicated the slopes have marginal factors of safety and would fail during a seismic event, causing mass gravity flow. Seabed topography will channel the mass gravity flow to impact the flowline. Landslide runout modelling of the slope failure quantified the mass gravity flow impact forces onto the flowline, and subsequent calculation demonstrated the flowline would instantaneously rupture under these impact forces. Finite-element modelling was used to refine the preliminary slope stability assessment, to remove the conservatism assumed in initial work and obtain a more representative actual stability of the slope. ABAQUS software was used to build a model of the slope. Site specific geotechnical input parameters were required for the model. As such, a suite of specialized tests on borehole samples were performed to generate these parameters and used in the model. A site response dynamic analysis was run using a purpose-built non-linear program to analyze the response of the slope under seismic acceleration forces. Initial input motions for the dynamic analysis were derived from the PSHA of an adjacent site. Concurrent work was performed to develop site-specific PSHA for later comparison. The dynamic slope stability demonstrated that the slope remains stable under the most onerous motions of a 4000-year return period Abnormal Level Earthquake (ALE), with significant margin. Additional analyses were performed to assess the severity of an event that might be required to cause slope failures by factoring the acceleration time histories by factors of up to 3.6. The slopes still remained stable throughout the intensified accelerations. The site-specific PSHA acceleration time histories were compared to those used in the analysis. It was found that input motions used in the analysis exceeded the site-specific 10,000-year return period seismic event motions, and in the case of the analysis using the factor of 3.6, by a very large margin. This provided project confidence that the steep slopes within the vicinity of the flowline are indeed stable under a seismic event. The flowline is demonstrated to be free from the threat of a mass gravity flow impact.