The safety and stability of waste dump are vital influencing factors to the mine sustainability and mine employees. Based on a real mine project in a certain open-pit mine waste dump in Tibet, the in situ test on waste rocks from waste dump, including measurements of density, water content, rock size, and natural repose angle, was conducted. Afterwards, these sample waste rocks, of which grain size is less than 5cm, were selected for indoor large-scale shear test under natural and saturated conditions. By using some engineering methods, the physical and mechanical parameters of waste rocks layer were then determined accordingly. MIDAS-GTS/NX has the advantage of pre-processing modeling. FLAC(3D) has good computational and analytical capabilities. The process of dump accumulation is simulated numerically. According to the calculation results of FLAC(3D), the distribution of stress, displacement and plastic zone in the dump is obtained. FOS (factor of safety) for each analytical step in this model was then calculated through the strength reduction method. The limit equilibrium method is used for waste dump stability analysis considering three states: only applied gravity, applied gravity and rainfall, and applied gravity and underground water. The results from this analysis show that the waste dump is stable. The potential failure modes of waste dump mainly consist of the combined sliding mode which has circular sliding in upper side and broken line sliding that cuts through gravel-soil layer into heavily weathered layer in the bottom. This paper documents some of the procedures and approaches utilized for waste dump life-of-mine design analysis. It provides reference for further waste dump optimization.

A new neural network and grey-relational model for mining subsidence is established on the basis of typical data of ground and strata movement in Chinese observation station. The sensitivity theory of the subsidence factors is proposed and the computation formula is derived as the first-order partial derivatives of the output variable with respect to the input variable, as well as the grey relation grade among the subsidence coefficient and each influence factors is analyzed using the grey relational analysis method. The results show that the order of neural network sensitivity and grey relation grade is: loose bed thickness, deep-thickness ratio and dip angle, which suggests the loose bed thickness is the critical factor. These results have reference value for the analysis and control to the mining subsidence.