Environmental Engineering Reference
In-Depth Information
Crest wall joint
Zone 2B
Zone 2E
Zone 2D
d
Concrete face
slab with joints
b, c
Zone 2F
Gabions or other
meshed protection
Zone 3B
Zone 3A
Perimet
ri
c joint
Plinth
a
e
Progression
Initiation of a concentrated leak
Opening of perimetric joint
Opening of vertical joints
Cracking of the face slab
Opening of the crest wall joints
Internal erosion and piping of
the foundation under the plinth
Breach
Continuation
(a)
(b)
(c)
(d)
(e)
Will zones 2D, 2E
2F and 3A act as a
filter system according
to filter design criteria?
Will the leak progressively
erode fines (or form a pipe),
leading to progressively
larger leakage?
Will overall instability of the
downstream slope, or
unravelling occur, leading to
loss of freeboard?
Figure 15.41.
Failure paths for concrete face rockfill dam.
large leakage, leading to a breach mechanism by unravelling or instability of the downstream
slope may result. At what stage this may occur is dependent on the zoning and the nature
of the rockfill; e.g. gravel fill, and low strength, weathered rockfill, is likely to have lower
permeability and be more likely to unravel or reach an unstable condition than other rockfill.
The one failure of a modern CFRD (Gouhou dam, Chen, 1993) occurred at apparently
relatively small leakage, but the zoning of that dam was poor, with the fill being silty
sandy gravel without high permeability drainage zones.
Hydro Tasmania, who owns and operates a number of CFRD, recently have been devel-
oping a framework for quantifying the likelihood of failure of CFRD. In a “workshop” held
in October 2001 Richard Herweynan and Paul Southcott from Hydro Tasmania, and Robin
Fell (Southcott et al., 2003) concluded that the failure path for a CFRD could be likened to
an internal erosion and piping failure for earth and rockfill dams, with the face slab and the
joint system being equivalent to the earth core, Zones 2D and 2E the filters. Hence the failure
path is as shown in Figure 15.41. The method is therefore strongly based on Foster (1999),
Foster and Fell (1999b, 2000) and Fell et al. (2000).
The likelihood of failure can be assessed qualitatively or quantitatively in this
approach. In either case, the failure paths should be considered for first filling, under nor-
mal operating conditions, flood (up to dam crest level) and under seismic loads. The prob-
ability that the reservoir level will be exceeded in any year needs to be incorporated into
any quantification of the normal operating, flood and seismic conditions.
For an overall risk assessment, the likelihood of failure by overtopping under flood;
scour/erosion of the dam due to overtopping of spillway chute walls and spillway approach
walls, and from waves caused by landslides into the reservoir should be considered.
This discussion concentrates only on those failure paths covered by Figure 15.41.
Clearly, what is important to any assessment of the likelihood of failure is the ability of
the owner of the dam to respond to the leakage incident, by lowering the water in the
reservoir to below the cracked zone, or by relieving the head driving the seepage and/or to
“seal” the crack or opening by, for example, placing sand on to the leak and have it wash into
the crack, and be prevented from eroding further by filter action on Zone 2E, or Zone 2F.
15.8.2
Assessment of likelihood of initiation of a concentrated leak
Table 15.14
summarizes the factors to be considered for assessing the likelihood of initi-
ation of leakage under normal operating and flood (to dam crest level) conditions. Similar
