Évaluation de la performance des géocomposites bentonitiques comme barrière aux fluides dans un contexte de recouvrement minier

The two main types of mine waste generated by the mining industry are waste rocks and tailings. Both types of wastes may contain sulphide minerals and when they are stored on the surface and exposed to oxygen and water, they have the potential to generate acid mine drainage (AMD). The latter is characterized by a pH <6, elevated level of dissolved heavy metals and high sulfate contents (Aubertin et al., 2002; Blowes et al., 2014). A proven reclamation method to mitigate AMD production is the construction of engineered multilayered covers over the mine wastes to prevent the infiltration of water and / or oxygen (Aubertin et al., 2002; INAP, 2012). Presently, geosynthetic clay liners (GCLs) are increasingly being considered as low saturated hydraulic conductivity materials as part of engineered covers. A GCL is manufactured in thin sheets (5-10 mm) that consist of a layer of bentonite interlaid between two geotextiles. Although literature on the properties of GCLs in a mining context are more frequent, research gaps still exist, particularly with respect to the application of GCLs in mine waste covers in cold climates (exposed to freeze-thaw cycles). The main objective of this research projet is to assess in the laboratory the performance of a GCL as a fluid barrier in the context of an engineered mine cover installed over acid-generating mine wastes. As part of this study, tests were performed on a conventional GCL (geosynthetic clay liner with non-woven and scrim-reinforced geotextiles). Laboratory experiments were divided into two parts: the laboratory characterization of the GCL used and more particularly, the bentonite it contains, then the determination of the hydraulic and diffusive properties of GCB samples hydrated with deionized water (DIW) or with a synthetic acid mine drainage (SAMD). The hydraulic and diffusive properties evaluated were the saturated hydraulic conductivity before and after freeze-thaw cycles, the water retention curve (WRC) and the effective diffusion coefficient (De) at high saturation value (Sr), both for permeation with DIW and SAMD. The evolution of the quality of the percolating SAMD through the GCL samples was also monitored during the permeability tests, as well as mineralogical composition of the bentonite after permeation.The results indicated that under low-pressure conditions, similar to those encountered in a mine cover, the saturated hydraulic conductivity of samples permeated with SAMD was approximately 10-8 cm/s, which is an order of magnitude higher than for samples permeated with DIW. An additional order of magnitude of difference was observed (10-7 cm/s) for a sample subjected to freeze-thaw cycles (FTCs), while an increase of two orders of magnitude (10-6 cm/s) was measured for the sample subjected occasional desaturation after a series of FTCs. This decrease in performance was attributed to the formation of cracks during the frost or the drying phase that did not self-healed during the subsequent permeation, due to the loss of swelling capacity induced by SAMD. The results of the chemical, electrochemical and mineralogical analyses also indicated that the SAMD followed the typical pH-buffering sequence associated with geochemical reactions observed during the neutralization of AMD with low pH and high Fe and SO42- concentrations. The permeability tests were ended at pH of 3,2 et 3,5, before chemical equilibrium could be reached.Similar WRCs were obtained for GCL hydrated with DIW and DMAS. In both cases, the WRC indicates that large suction must be applied (0.1 to 0.45 MPa) to desaturate an initially well-hydrated GCL. The De values did not vary according to the permeating liquid and led to values of 2.1 x 10-9 m²/s to 1.6 x 10-8 m²/s, for Sr of 88 at 94%. These values correspond to those estimated by the Aachib et al. (2004) model. However, since GCLs permeated with SAMD were all thinner than those permeated with DIW, it is expected that the diffusive flux through the GCL would be higher in the case a permeation with SAMD, for a same De value. Based on these results, it was concluded that under controlled laboratory conditions, and for GCL and SMAD used in this study, permeation with SAMD under low-pressure conditions negatively affected the performance of the GCL as a barrier to water and oxygen.Based on this research, it is recommended to conduct additional laboratory tests to determine the long-term saturated hydraulic conductivity (upon reaching a chemical equilibrium) of the GCL under study and further document the formation and repair of cracks that form during freezing and drying. It is also suggested to carry out long-term permeability tests on polymer-modified GCL (treated to enhance the chemical resistance of the bentonite) in order to compare the obtained results with those of conventional GCL. Finally, it is recommended to construct field experimental cells to validate the degree of saturation that can be achieved in the field and the influence of the soil layers adjacent to GCL on the quality of the hydration.