| This study focuses on the evaluation of the efficiency of several reclamation methods applied to two partially oxidized tailings impoundments. At the Manitou mine site, located near Val-d'Or, Québec, the adopted reclamation method consists of adding a layer of slightly alkaline (non acid generating) tailings onto the surface, together with an elevated water table. For the Selbaie mine site, located north of La Sarre, Quebec, the impoundment has already been covered by a 1 m to 2 m thick monolayer cover made of till. The reactive tailings have also been partially oxidized and the discharge water is acidic and contaminated, and has to be lime-treated.;Reactive tailings and cover material (till) samples were collected in situ in 2007 from both disposal sites. Non acid-generating tailings were provided by the Goldex mine. These materials have been characterized in the laboratory to define their hydrogeological and geochemical properties (water retention curves, saturated hydraulic conductivity, oxygen diffusion and consumption parameters, mineralogical compositions, total C and S contents, etc.). Instrumented columns (230 cm in height, and 15 cm internal diameter) were set up in the laboratory to study the hydrogeological and geochemical behaviour of the tailings-cover systems.;They were filled with 170 cm of reactive tailings, covered with 30 to 40 cm of cover material. Smaller columns (50 cm in height, 10 cm in diameter) were also included with the reactive tailings only to investigate their geochemical behaviour when left uncovered. Monthly wetting and drainage (drying) cycles were repeated to simulate, in a simplified (but controlled) manner, the site climatic conditions. Up to 19 cycles were applied to the columns. Volumetric water content, pressure head, dissolved and gaseous oxygen concentrations and climatic conditions were monitored. The experimental program also included the measurement of the pH, the electrical conductivity, and the sulphate, iron, and other metal concentrations in the leachate.;Vadose/W 2007 (GeoSlope Inc.) was used in this study to simulate water flow and oxygen diffusion in the columns. Hydro-geochemical simulations were also conducted using the Min3P code (Mayer et al., 2002). The first step of the modelling work was to compare and calibrate the hydrogeological and chemical parameters using the column test results. Once the column models had been calibrated to reproduce experimental measurements, similar material parameters were used to simulate field conditions. Different types of covers were also investigated, comparing both oxygen fluxes and water quality.;At the Manitou site, the simulations tend to indicate that a monolayer cover is usually inefficient at preventing AMD generation under the test conditions. Raising the water table to 2 m or 1 m below the tailings surface decreases the oxygen flux through the cover, but the tailings continued to oxidize nevertheless. Moreover, during dry years (without significant infiltration in July and August), the cover efficiency drastically decreases. A two-layer cover (water retention layer on a capillary break) could be an effective solution, but evaporation would have to be controlled. Increasing the thickness of the water retention layer could be an option, but costs could significantly increase. Finally, the simulations show that a three-layer CCBE (cover with capillary barrier effects) could be a cost-effective method to prevent further oxidation and AMD generation (considering both the oxygen flux decrease and water quality improvement). For the Manitou site, a CCBE cover should be made of a 1 m retention layer (Goldex tailings), sandwiched between two sand, gravel or waste rock layers (50 cm to 100 cm thick). Concerning the Selbaie site, none of the simulated configurations seemed very effective (except the CCBE with very coarse waste rock). The available till does not seem to be a suitable material for constructing a retention layer in a monolayer or two-layer cover. (Abstract shortened by UMI.). |
