Résumé - Abstract
Tailing Storage Facilities (TSF) are large structures used to store uneconomic mineral. They are generally reclaimed after mine operations with engineered covers that limit the flow of water or oxygen to prevent tailings oxidation. One of the most promising reclamation approaches is to build multilayer Covers with Capillary Barrier Effects (CCBEs), which act as oxygen barriers when a retention layer remains above 85% of saturation, reducing oxygen migration by a factor of 10 000.
Long-term monitoring programs traditionally use moisture content sensors in the layers of interest to evaluate the cover performance over time. However, since they have a limited investigation volume, many sensors may be needed to: (1) recover high spatial resolution of the moisture content distribution, (2) assess cover performance under different conditions (e.g. slope, vegetation, water table), (3) study the impact of heterogeneity of some materials used in the construction on the cover performance.
In this context, Time-Lapse Electrical Resistivity Tomography (TL-ERT) is a promising tool to spatially extend point data, and then reduce the number of sensors needed and instrumentation costs. Indeed, large-scale conductivity images obtained from ERT can be converted into distributions of moisture content with experimental petrophysical relationships. TL-ERT has been successfully applied to monitor spatiotemporal changes in moisture content for different contexts. However, there is still a lack of knowledge regarding the accuracy of TL-ERT for moisture content estimations in mining wastes, which is critical to evaluate CCBEs performance since small changes in saturation (typically around 85%) must be precisely detected.
This poster presents preliminary results from the hydro-geophysical monitoring carried out in large-scale experimental CCBEs and monolayer covers at the TSF of Canadian Malartic Mine, an active open-pit gold mine in Quebec. Four 30m x 10m x 2m experimental covers have been constructed and instrumented with hydrogeological sensors and geophysical electrodes in 2019 and 2020. Recently, the PRIME system - an autonomous TL-ERT monitoring system developed by the British Geological Survey (BGS) - has been installed to recover remotely daily distribution of electrical conductivity within the covers using 256 electrodes. These daily images of conductivity are then combined with point data to provide spatially extended estimations of moisture content.
We propose a methodology to assess the accuracy of moisture content estimations using both field measurements and numerical modelling over a 2-year monitoring period. We compare (1) point measurements made by capacitance sensors (moisture content θ, temperature T and electrical conductivity σ), (2) distributions of θ, T and σ obtained from thermo-hydrogeological modelling with PFLOTRAN and (3) moisture content distributions derived from TL-ERT. This approach allows to combine different datasets to obtain spatially extended estimations of moisture content for large-scale reclamation covers.