Effet des racines mortes sur les propriétés hydrogéotechniques des résidus utilisés comme matériaux de recouvrement sur le site minier Manitou

Dans la chaine de valorisation des ressources minières, de gros volumes de résidus miniers sans valeur ajoutée sont produits et entreposés en surface dans des parcs à résidus miniers ou dans des haldes à stériles. Ces résidus miniers, appelés encore rejets de concentrateur, sont un mélange de solides (roches fines broyées) et de liquide dont la valeur économique a déjà été extraite. S’ils sont générateurs de drainage minier acide (DMA), la gestion efficace de ces gros volumes de résidus miniers constitue une importante problématique environnementale et demeure un grand défi pour l’industrie minière. Le DMA est le résultat de l’oxydation des minéraux sulfureux contenus dans les rejets miniers exposés à l’air libre. L’apport en O2 dans ces résidus sulfureux est l’élément déclencheur de la réaction d’oxydation des sulfures, qui est à la base de la production de DMA. Il existe plusieurs techniques de prévention et de contrôle du DMA visant à réduire l’apport en O2 dans les résidus miniers. Par exemple, la technique de recouvrement avec nappe phréatique surélevée (NPS) combinée à une couverture monocouche est généralement utilisée comme barrière à l’oxygène. Le principe de cette technique consiste à maintenir les résidus miniers réactifs saturés, en rehaussant le niveau de la nappe phréatique jusqu’à la surface des résidus réactifs (cas de résidus déjà oxydés) et à recouvrir ceux-ci d’une monocouche de matériaux favorisant un bilan hydrique positif et le maintien du niveau élevé de la nappe. Cette monocouche peut aussi agir à titre de barrière à l’oxygène si elle est faite de matériaux fins gardant un haut degré de saturation. L’efficacité de cette technique repose donc en partie sur les propriétés hydrogéologiques des matériaux de cette monocouche. Or, la végétation est capable de s’établir naturellement sur les matériaux de recouvrement ou par végétalisation assistée. Cette végétation affecte directement le bilan hydrique des matériaux de recouvrement minier (pompage), mais aussi indirectement en modifiant les propriétés des matériaux via la colonisation racinaire. Les racines des plantes sont capables de modifier les propriétés hydrogéotechniques des matériaux en créant des bio- macropores (d≥100

Abstract

In the mining value chain, large volumes of non-value-added tailings are produced and stored on the surface in tailings facilities or waste rock piles. These tailings, also known as mill tailings, are a mixture of solids (crushed fine rock) and liquids from which the economic value has already been extracted. If they generate acid mine drainage (AMD), the effective management of these large volumes of tailings is a major environmental issue and remains the biggest challenge for the mining industry. AMD is the result of the oxidation of sulphide minerals contained in mine tailings exposed to the open air. The supply of O2 to these sulphide tailings is the trigger for the sulphide oxidation reaction, which is the basis for the production of AMD. There are several techniques for the prevention and control of AMD to reduce the O2 input to tailings. For example, the elevated water table (EWT) technique combined with a single layer cover is commonly used as an oxygen barrier. The principle of this technique is to keep the reactive tailings saturated by raising the water table to the surface of the reactive tailings (in the case of already oxidized tailings) and to cover them with a monolayer of material that promotes a positive water balance and maintenance of the high- ater table. The effectiveness of this technique is therefore partly based on the hydrogeological properties of the monolayer materials. Vegetation can be established naturally on the overburden or by assisted revegetation. This vegetation directly affects the water balance of the mine overburden (pumping), but also indirectly by modifying the properties of the material via root colonisation. Plant roots are able to modify the hydrogeotechnical properties of the materials by creating bio-macropores (d≥100 μm) through the decomposition of dead roots, thus leading to a reorganisation of the pore distribution in the materials. The presence of macropores may lead to an increase in the saturated hydraulic conductivity as well as a decrease in the water retention capacity of the materials. This could alter the performance of the mine cover system. Thus, it is important to study the impact that dead roots can have on the hydrogeotechnical properties of the materials used in this type of structure. At present, there are no studies that have focused on understanding the effect of dead root decomposition on the hydrogeotechnical properties of materials in a monolayer combined with an elevated water table. The objective of this study is to evaluate the effect that dead roots may have on the hydrogeotechnical properties of materials in a monolayer cover constructed from recycled non-acid generating Goldex tailings and combined with an elevated water table. The work was conducted on a reclaimed and vegetated mine site in 2009, located in a boreal forest context, the Manitou site. In situ work and laboratory analyses were carried out over two measurement campaigns (summer 2021 and summer 2022). During the first measurement campaign (summer 2021), ten stations combining seeded herbaceous vegetation and spontaneously colonizing trees (Salix sp.) were identified in Park 2, as well as five control stations with no vegetation. Vegetation surveys were carried out on the ten vegetated stations in order to inventory the species present and their cover. The above-ground biomass of the willows was also characterized. Samples were taken at different depths in the profile to characterize the basic geotechnical properties of the Goldex tailings (particle size distribution, porosity and relative density of grains). Finally, undisturbed and root-colonised samples were taken for saturated hydraulic conductivity (ksat) and water retention curve measurements in the laboratory. Then, half of the willow roots of each station were isolated (root death) from the main trunk of the shrub to a depth of 1 m and a width of 1 m by vertical insertion of a metal plate of 1 m2 surface area. In 2022, one year after killing the willow roots, new undisturbed samples were collected for new hydrogeological analyses. The ksat and CRE results of the two measurement campaigns were compared. Also, root colonization in tailings samples collected in summer 2021 was visualized (VGstudio max software) and bioporosity quantified (Fiji/J image software) on 6 subsamples by X-ray tomography and validated by the standard analysis method WinRhizo Pro 2017a. Then, simulations of total porosity evolution following four dead root decomposition (DRD) scenarios were made from the data of these six samples to investigate the potential effect of dead roots on hydrogeological properties (DRD 0 %, DRD 25 %, DRD 50 % , DRD 75 % and DRD 100 %). Thus, the CRE and ksat corresponding to each TDR scenario were predicted for each of these six samples. The CRE of the coarsest and finest material in each profile were also predicted in order to compare the effect of dead root decomposition and grain size heterogeneity on the water holding capacity of Goldex residues. The results of the vegetation characterisation of the Manitou site's Park 2 stations show a vegetation cover with a total cover of 100 %. Eight different herbaceous plants are present in addition to the willows, and it is the seeded Poaceae (grasses) that still dominate the vegetation (>80 % cover) 12 years after revegetation. However, the willows present on Park 2 demonstrate the ability of the surrounding vegetation to naturally recolonise this mine site. The saturated hydraulic conductivity (ksat) of samples taken before and after the willow roots were killed and decomposed in situ varies between 1.5E-05 and 2.91E-04 cm/s for all stations. This variation is non-significant for most stations with a difference generally less than ± ½ order of magnitude (measurement accuracy) between the two measurement campaigns, except for station SH14 which shows a variation of about one order of magnitude, between summer 2022 (ksat = 1.63E-04 cm/s) and summer 2021 (ksat = 1.98E-05cm/s). The air entry pressures (AEV) of the water retention curves (WRC) measured on the samples taken during the two in situ measurement campaigns vary from 12 to 40 kPa for most samples, which is typical of AEVs of coarse to fine silt, except for station SH20H where the AEV is 94 kPa, typical of finer and compacted material. At this stage of vegetation development, root colonisation does not appear to influence the hydrogeological properties of the Goldex tailings. Following the three-dimensional visualisation of the roots by X-ray tomography, the roots seem to colonise more the upper part (first 10 cm) of the Goldex tailings monolayer than the deeper parts of 25 cm and more. The average root length density (diameter > 53.3 μm) (RLD) obtained by tomographic analyses in samples taken at 5-10cm is higher (12.5 cm/cm3) than in samples taken at 25-30 cm (0.41 cm/cm3). For the surface samples (50-10 cm) of the tailings, the calculated bioporosity (eb, i.e. connected tubular pores of diameter > 53 μm occupied by roots) varies from 17 to 27 % and represents 26 to 56 % of the initial void index of the tailings, whereas the two deeper samples (25-30 cm) have bioporosities of 0.3-0.4 %, or 0.5-0.6 % of the initial void index of the residue, which is similar to the control sample TS21 (eb=0.28 % or 0.3 % of the initial void index of the residue). The 3D visualisations of the roots also show that the surface samples are colonised by coarser roots, with a diameter of up to 5.11 mm (station SH17H), while the deeper samples are colonised by only finer roots, with a maximum diameter of 1.03 mm (station SH12B). With 100% root mortality and decomposition, there could be a development of tubular and connected biopores with dimensions between 53 μm and 5.11 mm, which could represent up to 21 % of the tailings volume for station SH17H for example. The CRE prediction results taking into account this root biopore development show a reduction in the water holding capacity of the Goldex tailings. For example, for station SH17H, the AEV would decrease by 17 kPa due to the development of root biopores in addition to the pre-existing pores. The simulations also revealed that the particle size heterogeneity of the Goldex tailings would have a more significant impact on the desaturation of the tailings than a total decomposition of the roots creating biopores. For example, for station SH17H, the predicted AEV of the Goldex residues due to the granulometric heterogeneity of the profile was about 3 times greater than that associated with the root decomposition scenarios (0% versus 100% decomposed). Regarding the evolution of ksat, we note that the development of root biopores could lead to an increase in the predicted ksat. However, this increase in predicted ksat is not significant for all stations, as the increase is less than half an order of magnitude. These results suggest that at this stage of vegetation development on Manitou Park 2, the creation of biopores via root decomposition, even if total, would have little effect on ksat and CRE of the residues. However, the simulations do not take into account the shape, connectivity and size of the biopores created by the roots. The quantification of root parameters RLD, RSD, RDV and root diameter by X-ray Tomography was validated by the standard method of washing, sorting and digitising roots with WinRHIZO Pro 2017a software. A spatial resolution of 53.3 μm of the root biopores was achieved on 74 cm3 volum samples (50 mm diameter) and the method allowed visualization of the spatial arrangement of the root biopores. The comparison of root volume estimation by X-ray Tomography with the standard root survey method (WinRhizo) showed that root survey by X-ray Tomography is an accurate and less tedious non-destructive method on tailings samples with 12-13 years old vegetation.