Keynote Lecture – Alternative Numerical Approaches for Modelling Runout from Submarine Landslides and Implications for assessing Impact Forces on Subsea Infrastructure

Keynote Lecture – Alternative Numerical Approaches for Modelling Runout from Submarine Landslides and Implications for assessing Impact Forces on Subsea Infrastructure

Assessment of geohazard risks to offshore infrastructure such as pipelines and subsea foundations relies on modelling debris flow runout behaviour. This is a complex process that involves rapidly changing geometry, rheological properties and velocity distribution within the flow material, and also effects of entraining water and shallow seabed material into the debris flow. Ultimately, should the flow path intersect infrastructure on the seabed, estimates need to be made of the resulting forces and potential for damage. The objectives of the paper are to (a) compare three different computational approaches to model the debris flow runout; and (b) propose a method to estimate the steady-state kinematic forces that a given debris flow might impose on a partially buried pipeline. The different computational approaches cover (a) two different depth-averaged techniques, one based on a traditional Eulerian framework using a finite volume scheme, and one based on smoothed particle hydrodynamics and (b) large deformation finite element approaches including the material point method. While the depth-averaged approaches are much more efficient computationally, and hence can be applied to complex three-dimensional topography, the LDFE approaches capture different types of flow such as overtopping and separation into discrete blocks. For this study, the seabed topography is simplified into either planar, or varying in two dimensions (the latter being taken from a real case history), while the debris flow is modelled as a softening Herschel-Bulkley material. The different cases considered demonstrate how changing the material rheology affects the form of the runout and the distribution of internal velocities, both of which are also affected by the computational approach adopted. These are important issues to be resolved in order to obtain sensible estimates of impact forces.