In 1996, the Geotechnical Research Centre at The University of Western Ontario initiated a project to review water cover research and applications. The objectives of the project were to document the design and performance of water covers used to prevent acid generation in sulphide-bearing mine waste and compile results from water cover research. Water cover sites in Canada, Norway and Sweden were reviewed. Research programs reviewed were from Canada, Norway and the United States. They include MEND-sponsored work at the CANMET Elliot Lake Laboratory and Noranda Technology Centre, Pointe-Claire, Québec, and research undertaken by former United States Bureau of Mines, Norwegian Water Research Institute and Norwegian Hydrotechnical Laboratory.
The Province of Ontario, Canada Centre for Mineral and Energy Technology (CANMET) and Brunswick Mining and Smelting Corporation provided funding for the project.
The survey of water cover sites showed that the selection of the minimum depth of water is based on maintaining saturation of tailings in the event of a drought and preventing tailings resuspension. Predictions of resuspension are based on empirical correlations with bed shear stress, bed water velocity, or the wave height to water depth ratio. These correlations were generally obtained from experiments on particles other than tailings. Estimates of bed shear stress, velocity, or wave height are calculated from empirical equations relating these parameters to wind speed and pond fetch length, using linear wave theory. There was no conclusive verification of resuspension prediction at any of the sites we reviewed.
Methods presently used to predict the quality of pond and effluent waters do not consider many biological, geochemical and physical phenomena that may significantly affect water cover performance. In general, not enough long-term field data are available to verify water quality predictions, however, at one site field data have already shown the predictions to be unconservative.
The character of the reviewed sites, in terms of site hydrology, the initial composition of the tailings, the degree of oxidation of the tailings prior to flooding, the physical layout of the site, and the treatment of the tailings before or after flooding, varies considerably. Due to the dissimilarity of the sites examined, no correlation between water depth and design performance could be made. Observations common to many sites include a dramatic increase in pH during the summer, possibly due to microfauna activity, CO2 degassing and increase in alkalinity from surface runoff, and the formation of a thin coating of orange iron oxyhydroxides and organic matter at the tailings-water interface.
In column and lysimeter experiments for both oxidized and unoxidized tailings, flooding produces a thin surface coating of brownish-orange iron oxyhydroxides which in previously unoxidized tailings marks the extent of vertical oxidation into the tailings. The finite thickness of the oxidation zone apparently occurs because of the attainment of steady-state conditions between pyrite oxidation and sulphate reduction to sulphide. With time, the growth of organic deposits at the surface occurs, which further impedes upward metal flux to the water cover. Column experiments on oxidized tailings show that flooding causes the dissolution of oxidation products and the subsequent release of metals into the water cover, often resulting in metal concentrations higher than permissible levels. The dissolution of some oxidation products, or other minerals, may be slow and contribute to high metal concentrations in the water cover for many years. The use of a sand or a peat protective cover at the tailings-water interface reduces metal fluxes into the water cover to negligible levels. However, metal flux into the water cover may resume when the absorptive capacity of the protective cover is reached. The use of a peat layer creates highly anoxic and reducing conditions in the tailings. Peat also produces acidity that can significantly increase the mobility of some metals in the tailings pore water.
The addition of lime to previously unoxidized tailings prior to flooding inhibits the acitivity of iron-oxidizing bacteria by keeping pH high. Lime addition to oxidized tailings increases pH and limits the mobility of most metals in the water cover and pore waters. Mixing the top portion of tailings with lime was found to be more effective than simply adding the lime to the surface. Traditional methods used to estimate the required amount of neutralizing material, such as acid-base accounting, were inaccurate when applied to long-term (37-week) experiments on flooded tailings.
Laboratory test results show that coarser tailings under a water cover initially generate higher amounts of acid drainage than finer tailings. This could be due to the initially faster rate of vertical oxygen transport in coarser tailings.
The effectiveness of a Biologically Supported Water Cover (BSWC) has been demonstrated in the laboratory. A BSWC involves colonizing flooded tailings with plants to facilitate organic buildup for limiting resuspension and metal flux through biological adsorption. Preliminary results suggest that its effectiveness in the field is dependent on the depth of water and wind conditions.
It is apparent from laboratory and field studies that flooding tailings is the most successful method presently known for preventing and controlling ARD. However, in many cases the water cover discharge will require treatment to meet regulatory standards, and it is not currently possible to accurately predict the required amount and length of treatment. State-of-the-art predictive contaminant models ignore many important phenomena in water covers, including secondary mineral dissolution and metal release independent of oxidation and resuspension. Some of the work reviewed hypothesized that flooding will eventually establish a diagenetic environment for some tailings minerals with accumulation of organic matter. It is, however, uncertain how long this will take to develop and whether it will occur under all conditions. Furthermore, some metals may continue to be released to the water cover in such an environment. Although water covers are a promising technology, fundamental understanding of many phenomena influencing their performance is unknown, and the minimum depth of water required cannot presently be confidently determined. The present prediction tools do not consider other important tailings properties such as thixotropy and cohesion.
We recommend that additional research and field work be initiated to address the uncertainties and issues raised in this review. The key areas for such work may include verification of resuspension predictions at a few selected sites, fundamental understanding of tailings properties and behaviour, the nature and rate of accumulation of the organic matter-iron hydroxide sediment at the tailings-water interface, and the contribution of resuspension to tailings oxidation and long-term water quality.
