The 4th International Symposium on Mine Safety Science and Engineering (ISMSSE2018)

351 / 2018-05-29 23:34:19

3D numerical modelling of tailings dam failure run-out routings with smoothed particle hydrodynamics

Smoothed particle hydrodynamics; tailings dam failure; numerical modelling; overhead tailings ponds; emergency response plan.

Mine Dynamics Disaster > tailings pond

Tailings dam failure accidents with limited emergency response time and enormous threats often lead to heavy casualties, severe financial losses and irrecoverable environmental impacts. Recent examples are the 2015 Fundão accident and the 2014 Mount Polley accident. China, which has become one of the largest developing economics with huge amounts of both mineral resources exploitation and consumption, is now facing huge challenges from potential tailings dam failures. Currently, there are 8869 tailings ponds in China, of which 1425 were designated ‘overhead tailings ponds’, i.e. high-risk tailings ponds located upstream and within 1 km of residential areas and important facilities. Heavy casualties, serious property losses and unrecoverable environment damage would be inevitable if any of these tailings dams collapse. In order to understand how the tailings dam failure routing paths are affected and how to minimize its foreseeable losses, an interdisciplinary research funded by the National Natural Science Foundation of China (NSFC) on tailings dam failure routings simulation was carried out.
Tailings run-out routings are considered to be difficult to analyze and predict due to their large-scale deformation and complex downstream terrains. Grid methods such as the finite difference method (FDM) and the finite element method (FEM) with mesh sometimes lead to inevitable numerical difficulties including mesh winding, twisting and distortion in solving problems with large deformation and free surfaces. Smoothed Particle Hydrodynamics (SPH) is a fully mesh-free, particle-based, pure Lagrangian method. In this study, SPH was applied into the tailings slurry routings modelling, to avoid severe mesh distortions caused by large-scale deformation. Factors including tailings pond capacity, dam slope, sediment spatial distribution laws, hydraulic parameters, downstream topographic and land cover were considered. High-resolution and real-scale 3D models extracted from remote-sensing and unmanned aerial vehicle (UAV) photogrammetry were adopted. A case study of the dam material routing to the downstream community (Bento Rodrigues) during the known 2015 Fundão accident was carried out for validation. And further application will be introduced into potentially dangerous tailings ponds in China, in order to improve the Emergency Response Plan (ERP) such as emergency evacuate route, rescue plan and protection measures of important facilities.