Environmental Engineering Reference
In-Depth Information
an excessive amount of the site investigation effort and analysis, to the seepage which will
flow through and beneath the main embankment.
From the examples shown in
Figures 19.36
,
19.37
and
19.32
, it will be apparent that
the assessment of seepage flow rates will involve:
- Knowledge of the permeability of the tailings, as these are commonly part of the seep-
age path. In many cases they may control the seepage rates;
- Knowledge of the permeability of the soil and rock underlying the storage and sur-
rounding the storage. In Figure 19.32 it would be necessary to be able to model the
whole of the area between the streams, necessitating knowledge of rock permeabilities
well beyond the storage area;
- Modelling of the seepage, usually by finite element methods, which may involve several
section models and/or a plan model. This modelling should account for the development
of flow as shown in Figure 19.37 and not just model an assumed steady state coupled
flow situation, i.e. the storage and groundwater coupled as in Stage 3, Figure 19.37.
19.6.3
Some common errors in seepage analysis
Finite element seepage models are readily available and commonly used to estimate seep-
age from tailings storages. It is the authors' experience that the users of such programs
sometimes fail to understand the actual boundary conditions which will apply or ignore
details which are important to estimates. Some examples of errors follow:
The examples assume that the tailings have lower permeability than the underlying rock:
(a) Assuming saturated coupled flow, as shown in
Figure 19.38b
. This results in an
underestimation of the rate of seepage compared to the correct conditions shown in
Figure 19.38(a);
(b) Assuming saturated flow as shown in Figure 19.38b, with the water table at ground
surface downstream, when flow is insufficient to result in such a high water table
downstream (i.e. Figure 19.38c is more correct). This results in an underestimation of
seepage rates but, more importantly, seepage will not emerge at the downstream toe
of the dam, so will not be able to be collected or monitored at the surface;
(c) Failing to model and account for seepage through hills surrounding the dam, e.g.
ignoring seepage to the north and east in Figure 19.32. This can result in significant
underestimation of seepage rates;
(d) Assigning incorrect boundary conditions to the model as shown in
Figure 19.39b
.
This results in underestimation of seepage rates, and may lead to an overestimation of
the amount of seepage which may emerge at the toe of the dam;
(e) Failure to model permeable zones in contact with surface water, e.g. rock rip-rap zones
on the upstream slope as in
Figure 19.40
and high permeability alluvial, colluvial or lat-
eritised soil zone on the surface as shown in
Figure 19.41a
. In both cases, failure to model
the high permeability zone can lead to significant underestimation of the seepage rate;
(f ) Use of incorrect permeability in the foundation e.g.:
- Using the results of flow-in type permeability tests in soil and weathered rock, where
smearing and/or blocking of fissures and joints yields lower than actual permeabilities;
- Use of permeability values obtained by methods which are unable to detect major
leakage paths, e.g. from boreholes in deeply lateritised areas where drainage fea-
tures are often localized and near vertical;
- Failing to allow for the variability of tailings permeability - as discussed above.
(g) Incorrect estimation of travel times for contaminant fronts, e.g.:
- Use of porosity instead of storage coefficient in estimating flow velocities;
- Use of unrealistic values (too high) of storage coefficients in jointed rocks;
