Environmental Engineering Reference
In-Depth Information
Morgenstern and Kupper (1988) indicate that, in their experience, the laboratory flume
tests did not show trends consistent with field behaviour.
The concept of a master profile is, however, well established and has been noted by oth-
ers, e.g. Smith, Abt and Nelson (1986). Hence, one can predict later performance quite
well from the initial behaviour of a tailings storage.
For planning purposes in a new project one will have to rely on flume tests and rela-
tionship to other measured profiles for similar tailings.
De Groot et al. (1988) point out that the observed slopes in dredged sands below water
are much flatter than the predicted equilibrium slope (which might be expected to be
steeper than slopes above water because of reduced gravitational forces below water).
This is explained as due to flow slides where the upper part of a slope oversteepens as sed-
imentation occurs, resulting in slope instability and flow which produces flatter slopes.
They note that observed slopes below water were similar to that in the beach above water.
This is also the authors' usual observation for high clay content tailings, although in an
Alumina red mud storage the slope was steeper below water.
19.4.2
Particle sorting
In conventionally thickened tailings, because the coarser particles contained in the tailings
settle more rapidly than the finer particles, a gradation of particle size occurs on the
beach, with coarser particles depositing near the discharge point. As discussed above par-
ticle sorting also occurs within a layer, with coarse, high specific gravity particles settling
to the base of a layer. These can form a sand “parting” even in high fines content tailings.
Conversely, if the coarse particles are low specific gravity as occurring in coal washery tail-
ings, they will “float” to the surface of a layer.
The lateral variations in grading with distances from the discharge point can have an
important effect on tailings permeability - with the coarser more permeable tailings
deposited near the discharge spigots and the finer low permeability slimes deposited fur-
ther away. This is used in the upstream method of tailings dam construction to control
seepage pore pressures and, hence, to maintain stability. Blight and Bentel (1983) and
Blight (1987, 1988) discuss this feature and conclude that the particle size distribution can
be roughly predicted from:
Ae
BH/X
(19.13)
where
D td an eHdownthebeach)
Doft
(
50
A
(19.14)
(
he totaltailings)
50
B
characteristic of tailings and is a function of the discharge rate, and the specific grav-
ity of the tailings;
X
length of beach as in
Figure 19.11
.
Abadjiev (1985) presents a similar formula which can be applied to not only the D
50
size but to D
90
, D
60
, D
10
etc. He also presents data from several tailings dams showing the
lateral gradation.
The authors' experience is that the amount of sorting which occurs is dependent on the
method of deposition. Spray bars and closely spaced spigots give very good sorting provided
discharge rates are kept low. This is consistent with the observations of Blight and Abadjiev.
It should be noted that for tailings thickened to 50% to 60% solids content there is no
sedimentation phase during the deposition, little or no bleeding of water/liquor and little
