Search

Tailings impoundments are often prone to failure due to various causes related to complex loading conditions. For instance, tailings impoundments can fail during or after a seismic event, mainly because of the liquefaction of loose, cohesionless tailings. To improve the seismic stability of impoundments, the density of the tailings can be increased or waste rock inclusions (WRI) can be added in the impoundment. This paper presents the results of numerical analyses of the impoundments loaded with a variety of ground motions with different frequencies, scaled at two different levels of intensity representative of two mining regions of Eastern Canada. The results show that the conventional impoundments are subjected to medium to high displacements at the end of shaking; these show a high probability of rupture during the earthquake or in the post-seismic phase. Tailings densification can decrease the volume of liquefied tailings and displacements during the seismic phase but does not significantly reduce the probability of post-seismic failure. WRI can slightly reduce the volume of liquefied tailings, depending on the intensity of the motions, but the reinforcement reduces significantly the lateral displacements; impoundments with WRI show even better performances during the post-seismic phase. The use of WRI is deemed a good approach to ensure the seismic stability of tailings impoundments and reduce the risk of seismically induced failure. [ABSTRACT FROM AUTHOR]

A comprehensive campaign of laboratory and field tests has been conducted to characterize the behaviour of tailings from a hard-rock mine. Cyclic triaxial, direct simple shear, and triaxial simple shear tests have shown that the contractive tailings are susceptible to liquefaction and that their behaviour is similar to loose sand despite their higher fines' content. To model the behaviour of tailings' impoundments in dynamic simulations, the PM4Sand constitutive model is calibrated based on the dataset presented here. Some of the model parameters are defined based on consolidation and compression tests, while the critical state line (CSL) is defined based on the results of monotonic triaxial and direct simple shear tests. The CSL is lower for tailings than for most natural sands, which is consistent with previous studies. The calibrated model can reproduce reasonably well the monotonic behaviour and provides an excellent fit of the cyclic strength curves measured in the lab at different confining pressures and density index. The results presented in this paper indicate that the PM4Sand model can be used to simulate the cyclic behaviour of low-plasticity, hard-rock tailings. The proposed procedure also provides general guidelines regarding model calibration for other types of tailings. [ABSTRACT FROM AUTHOR]