Covers made of various particulate media (often called «dry covers») sometimes
constitute the only applicable technology to control the production of acid waters
from sulfide tailings. Selecting such type of cover is based on their capacity to
reduce the infiltration of oxygen and/or water, as is shown in this project.
However, because covers are usually made of different components, they may
become very costly. Hence, cover design should be optimized, from a practical and
economical point of view. In this regard, the use of non reactive (clean) tailings
may help reduce significantly the costs involved. The laboratory tests performed
during this research project have served to characterize such milling wastes and to
evaluate the efficiency of layered cover systems.

The initial phase of the project, which started in 1991, has allowed a thorough
investigation of the hydro-geotechnical properties of various tailings, including the
hydraulic conductivity, the water retention curve and the oxygen diffusion
coefficient. The analysis of the hydraulic behavior of layered systems, in which the
capillary layer is made of clean tailings, was done using drainage column tests. The
efficiency of these systems to limit the production of AMD has also been evaluated
with control columns where reactive tailings were covered by a layered system
(see MEND 2.22.2a).

The results obtained show that the effective oxygen diffusion coefficient Dc of
partly saturated tailings becomes practically equal to that of water when the
saturation ratio Sr reaches about 90%. At this value of Sr, the gaseous phase is
discontinuous, and oxygen transport is controlled by diffusion (with solubilisation)
through the water filled pores. In theory, the efficiency of a «dry cover» then
becomes comparable to that of a water cover of the same thickness, the later being
considered as the most efficient method to control AMD.

The bulk of the column tests results and of the numerical simulations from this first
phase of the project (presented in chapter 3, with some new -never publishedresults)
fully support the working assumptions: layered covers can be used to limit
the production of acidic waters. Furthermore, the column tests have helped validate
the numerical calculations on water distribution and on the development of
capillary barrier effects in layered covers. It has been shown that, following a 2year
test period, the tailings layer, placed between two sand layers, stayed close to
saturation at all time, even during dry periods of up to 60 days.

Covered reactive tailings have also shown a leachate with a well controlled pH.
For instance, a pH below 2 was observed after 200 days (between 7 and 8 wetting
cycles) for uncovered reactive tailings, while that of the same tailings remained
close to neutral when located under covers with capillary barrier effects (CCBE).
Other geochemical parameters (Eh, electrical conductivity, sulfates and metal
concentrations) have also confirmed the efficiency of the CCBE. The main results
obtained during this first phase of the project (between 1991 and 1996) are
presented and analyzed in this report (MEND 2.22.2b).

The second phase of the project, which started in 1995 with field work (following
MEND 2.22.2a), aims to confirm and to refine the results obtained in phase I.
Phase II started with the construction of experimental cells in situ to evaluate,
under more realistic conditions, the performance of various CCBE (see MEND
2.22.2c). In parallel, a laboratory was conducted. It served to characterize the
material properties and to evaluate, under well controlled conditions, the behavior
of the same cover systems as constructed in the field. The effect of cover geometry
(including the slope length and the slope angle) is also being investigated further.

Control columns were set up using the same configurations as used in the field.
These will help isolate the effects related to the systems themselves and those that
depend on the exposure conditions. In this report (MEND 2.22.2b), the material
properties and the content of the various columns are presented. Results obtained
after 6 wetting cycles are also shown. The hydraulic behavior of the various
layered systems is similar to the one observed in phase I. As for the efficiency to
control production of AMD, the results are still incomplete. Nevertheless, the field
data obtained so far and the numerical calculations results show that the various
systems being investigated should function according to the basis assumptions.
Laboratory and field work will continue in 1998 and will appear in MEND Report
2.22.2c in 1999.