Étude du comportement hydrogéologique de couvertures avec effets de barrière capillaire sur une halde à stériles de grande dimension

Reclamation of large waste rock piles is a great challenge for the mining industry. Covers with capillary barrier effect (CCBE) are mainly used to control acid mine drainage (AMD) on mine tailing ponds and more rarely on waste rock piles. This method involves the installation of a layer of a fine-grained material on a layer of a coarser grained material. The top layer's saturation is maintained by the capillary barrier effect (Aubertin et al. 1995, Bussière et al. 1999) . The saturated layer limits the oxygen diffusion through the mining waste, which limits the AMD. The suction in the water retention layer must not exceed the air entry value of the fine-grained material for this method to be effective, because it could lead to a desaturation. Those conditions could theoretically occur on large waste rock piles. Therefore, the objective of this project is to determine the efficiency of a CCBE as a method to reclaim large waste rock piles, with physical models at intermediary scale on the field and numerical simulations to validate field measurements. To do this, two instrumented field cells (CS2 and CS3) have been built on the Canadian Malartic mine site (CMM). The CS2 cell is a typical CCBE which is made of a 0,5m thick capillary break layer, a 1 m thick water retention layer and a 0,3 m thick protection layer. The CS3 cell is a CCBE for which the waste rock pile serves as a capillary break. The cell is made of a 1,8 m thick water retention layer and a 0,3 m thick protection layer. Tailings from CMM were used to build the water retention layers and waste rocks from CMM were used to build the capillary break and the protection layers. The physical, hydrogeological, chemical and mineralogical properties of those materials have been characterized in the laboratory. To determine if the covers are effective to limit oxygen diffusion, hydrogeological data (suction and volumetric water content) was collected in the field cells and compared with numerical simulations solutions. Oxygen migration was also studied using in situ oxygen diffusion/consumption test. The hydrogeological data measured in both cells shows relatively high volumetric water content values in the water retention layer and lower values in the capillary break layer and in the top protection layer. The average suctions are under 5 kPa in the capillary break layer and in the waste rock under the cells. In the lower water retention layer, the average suction values are 6 (CS2) and 7 kPa (CS3) and they increase up to values around 20 kPa in the upper water retention layer. The behavior shown by the volumetric water content and suction profiles is what is expected of a CCBE; the capillary break is present in both cells. The degrees of saturation calculated from the maximum volumetric water content values measured and for an average porosity of 0.44 are 70 and 100 % in the CS2 cell. In the CS3 cell, the degrees of saturation in the water retention layer, calculated from the maximum volumetric water content values measured with an average porosity of 0.48, are between 50 and 70 %, depending on the depths. According to the water retention curve (WRC) obtained in the laboratory, the air entry value is about 30 kPa. The degree of saturation should be higher to match the suction values obtained. Since the volumetric water content increases with time, the system is wetting. The suctions would better fit on a wetting water retention curve. Volumetric water content values could increase with time. Oxygen diffusion/consumption tests were used to determine the effective diffusion coefficients (De) by adjusting the data obtained during the tests with the curves obtained with the Pollute V07 (Pollute) software (Rowe et Booker, 2004). The reaction consumption rate (Kr) was fixed at 1x10-3hr-1 for all the simulations in Pollute. This value was obtained from the laboratory in diffusion cell (two reservoirs cell). For the CS2 cell, the values of De are between 3x10-8 and 1x10-10 m2/s and they are between 3x10-9 and 1x10-10 m2/s for the CS3 cell. Those tests seem to be controlled by sections of the water retention layer which are probably more saturated than what was measured with the sensors. Oxygen fluxes through the base of the covers were calculated with the analytic equation presented in Mbonimpa et al. (2003) for a reactive material under steady state conditions with the Kr and De obtained from the oxygen diffusion/consumption tests. It represents a case where the waste rock pile would be completely covered. The calculated fluxes are between 10-7 and 10-9 mol/m2/yr for the MOC tests performed in 2017 in the CS2 cell. For the CS3 cell, the fluxes calculated are close to 10 - 9 mol /m2 / yr for all the tests except for a single test where the flux is 10- 4 mol/ m2 / year. These results are all below the thresholds for an effective CCBE presented in Dagenais et al. (2012), which are between 0.6 and 1.25 mol /m2 / yr. Numerical simulations have been performed to represent a system equivalent to field conditions. Results of those simulations show that suctions values are close from the ones obtain in the field measurements. They are approximately 4 kPa at the bottom of the water retention layer and in the capillary break layer just below. The suctions values increase with the rise in the water retention layer. The same results are obtained with different water table depths simulated under the cell, i.e. -1m, -5m and -15m. According to these simulations, the depth of the water table has no effect on the volumetric water content, nor on the suctions in the water retention layer. This also shows that the suctions are not higher when the water table under the cells in deeper, as it was observed in the field. The simulations predict a water retention layer saturated at over 85 %, which is not what we observed in the field cells. Another simulation was performed with volumetric water content values close from the field cells initial state. The results show that the water retention layer may take several years to fully saturate. This could partly explain the lower values obtained in the field cells. Based on the results obtained from numerical simulations and field tests and measurements, the CCBEs tested could be effective in decreasing the oxygen diffusion in the long term and thus control the formation of AMD from the underlying waste rock on the horizontal parts of a waste rock pile.