Environmental Engineering Reference
In-Depth Information
19.4.5
The prediction of desiccation rates
If one is to rely on the increased dry density and strength which can be obtained from
desiccation of the tailings, it is necessary to predict the rate of drying. The rate is dependent on:
- The properties of the tailings, in particular the settled water content, the permeability
and the suction pressure which develops on drying;
- The climate - evaporation and rainfall;
- The deposition cycle - depth of and time between each cycle of deposition.
In Swarbrick and Fell (1990, 1991) the results of a research program to develop a
method for predicting desiccation rates is described. Based on laboratory and field drying
experiments, it has been shown that desiccation occurs as:
(a) Settling until the rate of water release equals the potential evaporation;
(b) Stage 1 drying, which occurs at a linear rate with time, generally at the same rate as
from a free water surface;
(c) Stage 2 drying, which occurs at a decreasing rate. This decreasing rate has been shown
to satisfy the sorptivity equation i.e.
E
b
t
(19.17)
cum
where E
cum
cumulative evaporation after the linear stage t
time after the linear
sorptivity coefficient (mm days
0.5
).
Swarbrick and Fell (1992) give details of the method for predicting drying rates using
laboratory drying tests on samples 600 mm
stage (days); b
300 mm.
It is very important to note that the drying of tailings is dramatically affected by the
presence of salts in the water/liquor. This is a particular problem in alumina red mud
where the salts are NaOH and NaCO
3
, but is also a problem in other tailings such as from
coal washeries where natural groundwater is highly saline, or if saline water is used for
processing or neutralizing the tailings. The salt concentrates on the surface, and the
osmotic suction slows and eventually stops evaporation from the tailings surface. To keep
desiccation proceeding it is necessary to break up the salt on the surface using special
equipment such as that shown in
Figure 19.15
or using swamp dozers once the tailings are
strong enough for them to work.
600 mm
19.4.6
Drained and undrained shear strength
19.4.6.1
Drained shear strength
The effective strength parameters c
for tailings can be obtained from triaxial or direct
shear tests on samples of the tailings. It is particularly important that sandy tailings are
tested at the correct relative density and stress range, because of the strong dependence of
,
on the relative density and the curved nature of the Mohr envelope.
Table 19.7
shows some typical values of effective friction angle. It will be noted that most
are relatively high, reflecting the absence of any structure and the grinding of rock to angu-
lar particles. Unless the tailings become cemented on deposition, e.g. by gypsum, the effec-
tive cohesion is c
0.
As discussed below, it will in many cases be wrong to rely on the drained shear strength
of the tailings because if they are loose and saturated, they will be contractive on shearing
and develop positive pore pressures. The undrained strength is a more reliable way of
assessing their strength. The beginning of contractive behaviour leads to a sudden increase
