Environmental Engineering Reference
In-Depth Information
- The treatment which has been applied to the foundation e.g. shotcrete or other con-
crete, or filters between the core and the foundation;
- Hydraulic gradients and stresses in the core material in the cutoff trench. High
hydraulic gradients and narrow, deep cutoffs in which the core material may arch, giv-
ing low vertical stresses, are more likely to experience initiation.
10.7.3.2
Continuation of erosion
Continuation of erosion will be dependent on the “filtering” capability of the open joints,
or coarse foundation soils, compared to the core material. This can be assessed as
described in Chapter 9, considering no erosion, excessive erosion and continuing erosion
criteria. The opening of joints can be considered as equivalent to the equivalent opening
size, so therefore equal to D
15F
/9. From the consideration of the penetration of cement
grout into jointed rock, it seems unlikely that erosion will continue where the crack open-
ing size is less than 3D
100B
of the core material. (See Section 12.2.4.4). These two
approaches could be compared, to assess whether they give consistent outcomes, and a
judgement made. For new dams, a conservative approach should be taken.
10.7.3.3
Progression to form a pipe and breach
The factors controlling progression to form a pipe and breach will depend on the likely
failure path and whether it is through the embankment or foundation. The factors
described above would then apply. It is noteable that in the case of Teton (Independent
Pane, 1976) and Fontenelle (Bellport, 1967) dams, relatively small leakages of only
300-400 litres/sec, eroded and initiated unravelling and slope instability of the down-
stream slope. This was related to the fine material, with low discharge capacity, in the
downstream zones of these dams.
Figure 10.48
shows the particle size distributions.
10.7.4
How to apply this information
The authors have found the information in Section 10.7.1 to 10.7.3 useful in the assess-
ment of existing dams, using an enhanced traditional engineering assessment approach
e.g. Bell et al. (2001) and in a quantitative risk analysis framework e.g. Foster et al.
(2002).
The rigour of breaking the internal erosion and piping process to initiation, continua-
tion, progression and breach, for each of the areas of potential weakness of a dam (e.g.
adjacent conduits, walls, over irregularities in the foundation, in the upper part of the
dam where filters may have been omitted) leads to a better understanding of the dam. The
tables allow a qualitative assessment of likelihoods - e.g. if all factors affecting initiation
of erosion are “more likely”, then the likelihood of initiation is high.
This can be converted to numbers (as in Foster et al., 2002) but that is not always nec-
essary.
It is also useful to use this approach when considering new dams, where compromises
on design or construction are being assessed and in the design of remedial works.
It should be noted that research at UNSW in 2002-2003 has developed better methods
for quantifying the rate of erosion of soils, and the initial hydraulic shear stress which will
case erosion to begin. This is described in Wan and Fell (2002, 2004a and b), this research
is on-going.
