Characterization of the bottom turbid layer in a pit lake and its response to convective cooling

Base Mine Lake (BML) is a pit lake used to store the Fluid Fine Tailings (FFT) generated by the oil sands industry, in the Athabasca region (northern Alberta, Canada). It is well studied to try understanding its evolution over time and to determine if an effective separation of the FFT from the cap water will eventually be reached. The mixing mechanisms taking place in the lake and affecting the suspended solids are of particular interest. Turbidity increases at the bottom of the water column suggesting the presence of a turbid layer which might be subject to mixing. The solids concentration exceeds the maximum limit of the sensors and deeper values cannot be recorded which prevents to fully understand the vertical structure of the near FFT interface region. To estimate the suspended solids concentration (CV ) over the entire water column, a laboratory calibration of the turbidity sensors is performed and the linear relationship CV [mg=L] = 0.84 x Turb [NTU] is obtained. This corresponds to concentrations of approximately 200mg=L for the region located just above the peak of turbidity, at the bottom of BML. The point of zero turbidity corresponding to the upper limit of the sensors is estimated at approximately 40 g=L. This value must be considered as a minimum solids concentration for the region located below the peak of turbidity. A small optical device is built using a waterproof camera in order to validate the previous results by applying another method than turbidity. The values obtained are comparable and this visually-based estimation can be useful to better understand turbidity variations. Convective cooling, which is one of the processes possibly affecting the bottom turbid layer during fall turnover, is studied in the laboratory on a kaolinite suspension. A simplified experiment using the melting of an ice cube to generate surface cooling shows the rapid rise of the turbid layer as cold water penetrates across its interface. This results in a decrease of bottom turbidity and an increase of mid-depth turbidity. The convective process is then roughly modelled using plume theory and conservation equations. The results obtained are in the same order of magnitude than the values observed. A more realistic experiment with a more uniform and gradual cooling of the water surface suggests that convective cooling might affect the bottom turbid layer by eroding its interface and slowly entraining particles to the water above.