This report documents the results of the NTC part of a
MEND-coordinated, multi-participant research project. The
project was designed to evaluate the effectiveness of shallow
water covers in the prevention of acid mine drainage from
reactive sulphidic tailings, using Louvicourt Mine as the
experimental site. Laboratory tests conducted on Louvicourt
tailings were used to derive the intrinsic oxidation rate necessary
for perform mathematical modelling of submerged tailings
oxidation.
In this study, various Louvicourt tailings samples were
characterized for grain size distribution, quantitative mineralogy,
geochemical whole-rock composition, and extended acid-base
accounting (ABA). Flow-through cell leach tests were used to
investigate the influences of four parameters on metal releases
by tailings under simulated submergence. Eight humidity cells
containing duplicates of four samples were tested for eighty
weeks to determine the rates of sulphide oxidation and acid
neutralization. Pre- and post-humidity cell analyses were
performed to complete geochemical and mineralogical mass
balances and to validate the humidity cell data interpretation.
Data generated from these laboratory tests were used to predict
field acid generation for a hypothetical field exposure.
Mathematical modelling was used to evaluate the effects of four
oxygen transport mechanisms on the degree of subaqueous
sulphide oxidation.
ABA results indicate that the tailings are potentially net
acid-generating. A four-month in-plant monitoring campaign
conducted in 1994-1995 showed a variation of sulphide content
from 11 to 49%. The sulphides in the tailings are dominated by
pyrite, with minor or trace pyrrhotite, sphalerite, and
chalcopyrite. Carbonate mineral contents in the samples varied
from nearly nil to as high as 24%. The main carbonate minerals
are ankerite and siderite, both containing varying amounts of
magnesium and manganese. The main silicate neutralising
mineral is clinochlore.
Flow-through cell leach experiments with different leachant
solutions using the Taguchi design approach suggest the
following influence on metal releases: leachant Fe2+
concentration (strong) > leachant DO level (strong) > leachant
pH (medium) > hydraulic gradient (weak). Presence of Fe2+ in
the inflow increases metal releases likely through a one-time ion
exchange process. High DO in the inflow promotes Zn releases
through oxidation of sulphides whereas low DO facilitates the
release of Mn. Lower pH favours metal releases probably
because of higher solubility of hydroxides and carbonates of
most metals. The tailings were found to contain sufficient
buffering capacity to maintain the pore water pH nearly neutral.
Mechanisms controlling metal releases include solubility control
and dissolution rate control. Overall metal releases are low
throughout the experiments except during the initial flush-out of
accumulated soluble constituents.
The humidity cell results show that the Louvicourt tailings have
relatively high oxidative reactivity. The oxidation rate of the
eight tests ranged 864-2143 (average 1449) mg CaCO3
eq/kg/week, and the NP consumption rate ranged 955-2238
(average 1500) mg CaCO3 eq/kg/week. All samples are
potentially net acid-generating, with predicted humidity cell lag
times ranging 0.56-2.5 (average 1.2) years. Predictions based on
a hypothetical field exposure of the tailings indicate that, for a
typical tailings, the lag time before acid generation is 4.5 years.
For a worse-than-average case the lag time reduces to 2.6 years.
Sphalerite oxidation appeared to be accelerated by galvanic
effects after the leachate pH dropped below about 3.0. Ankerite
seemed to contribute fully to the total available NP. Siderite and
clinochlore were less reactive and contributed less to the total
available NP. Siderite dissolution seemed to be accelerated after
onset of acid generation whereas clinochlore dissolution was
relatively unaffected by acidification.
A new technique was employed to calculate the dissolution rates
of individual neutralizing minerals and sulphide minerals from
weekly leachate volume and chemical data. The validity of this
technique appears to be acceptable judging from the
independently measured mineralogical mass balances.
Due to the “non-ideality” of the humidity cell tests, not all
particles placed in the cells were accessible for oxidation and
neutralization reactions. This was probably attributable to the
formation of impermeable particle aggregates as a result of
cementation and coating. Methods for correcting for the
non-ideality were proposed and demonstrated. It was found that,
without agitation, on average only about 37% of the sample
mass in the humidity cells was available for oxidation and
neutralization reactions.
Four basic cases that may occur after reactive tailings are
disposed of under a shallow, 0.3-m water cover were
mathematically modelled based on typical tailings properties
and other site conditions found at the Louvicourt Mine. The four
cases are stagnant water cover, fully oxygenated and mixed
water cover, fully oxygenated and mixed water cover with
downward infiltration, and tailings resuspension. The stagnant
water cover through which oxygen must diffuse across
transports the least amount of oxygen to the submerged tailings,
with the flux being on the order of 3 g O2/m2 interface/year.
Although this is the most desirable condition, to date field data
collected in other studies indicated that it is highly questionable
that this condition exists in reality, since winds that are almost
always present in the field naturally cause mixing, circulation,
wave action, and aeration in shallow water bodies. The three
other cases are more realistic scenarios, which increase the
oxygen flux into the submerged tailings significantly. Modelling
results suggest that, compared with the base case of stagnant
water cover, mixing/oxygenation of the water cover and tailings
resuspension each is capable of increasing the oxygen flux by
one order of magnitude, whereas downward infiltration of
fully-aerated water cover can enhance the oxygen flux by a
factor of three. The range of oxygen fluxes seen in the modelling
results suggest that for most sites, a simple, well-maintained
water alone without additional measures is sufficient to suppress
oxidation of sulphides in reactive tailings while maintaining the
discharge from the water cover during wet seasons in
compliance. Nevertheless, for exceptional circumstances where
this is not achievable, supplemental measures, such as physical,
chemical, and biological barriers/oxygen interceptors, are
available to further reduce the oxygen flux and enhance the
effectiveness of the water cover.
It is recommended that the findings of this laboratory/modelling
study be compared with the results from the field experimental
cell study (by INRS-Eau) and the laboratory column study (by
Canmet) for consistency and corroboration. Any discrepancies
among the three in basic findings should be addressed and
ultimately resolved.
Future research opportunities on water covers should be taken
advantage of to study factors controlling water cover
aeration/mixing and factors controlling resuspension of tailings.
The goal of such researches should be to establish the capability
of quantitatively predicting the degree of aeration and
resuspension in shallow water covers from basic information
such as tailings properties, meteorological data, and
physiography of the site.
