Environmental Engineering Reference
In-Depth Information
The larger movements which have been recorded are most likely related to poor compaction
and/or use of low modulus, or thick rockfill under the face slab near the perimetric joint.
Large displacements were measured on the perimetric joint of Golillas Dam, predomi-
nantly as joint opening. Despite the use of a triple waterstop, large leakage rates were
measured during first filling and continue to be measured. There were no apparent low
modulus rockfill zones around the perimetric joint and comparatively low strains were
measured in the rockfill during and following first filling, as illustrated in
Table 15.8
. It is
likely that the large perimetric joint displacements measured at Golillas are related to the
face slab geometry. The dam is located in a steep-sided narrow valley with a crest length to
height ratio of 0.81. Similarly large opening and settlement of the perimetric joint were
observed at Alto Anchicaya Dam which was constructed in a steep-sided valley. Amaya and
Marulanda (2000) observe that “in both dams the plinth is essentially a vertical wall
anchored to the rock, with the rock surface under the plinth having the same inclination.”
Both Foz Do Areia and Xingo dams have exhibited similar magnitudes of displacement
on perimetric joints. They have very similar upslope and downslope face angles and
approximately the same crest length to height ratios. Reported face slab movements
(
Table 15.10
)
for both these cases were well in excess of typically measured deformations
on perimetric joints and face slabs, which is likely to be related to the nature of zoning
within the rockfill and variations in modulus between zones. Both dams had upstream
zones of well compacted angular rockfill and large downstream zones of moderately com-
pacted rockfill. Hunter and Fell (2002, 2003c) reported large variations in average secant
modulus during construction between these two zones in both dams. Perimetric joint
movement at Xingo may also have been related to irregular left abutment and foundation
geometry. Variations in rockfill thickness between the face slab and foundation resulted in
differential settlements behind the face slab and eventually cracking in the face slab.
15.6.4
Cracks in CFRD dams
The progressive development of CFRD embankments has resulted in improved design of
the rockfill, plinth and face slab with associated reduction in leakage. However as higher
and higher dams are built design peculiarities have been associated with leakage incidents.
These incidents appear to have resulted from either foundation features or differential
deformation of the rockfill (Marulanda and Pinto, 2000).
Mori (1999) identified three general types of face slab cracking:
(i) Type A - shrinking cracking associated with concrete curing. These cracks are typically
horizontal, very small width (
1 mm) and limited in extent to the width of a slab;
(ii) Type B - associated with bulging of the lower part of the rockfill embankment. Mori
observed that, in cases where upstream earthfill is in place; this type of cracking is
usually restricted to the middle third of the dam height;
(iii) Type C - cracks associated with settlement of the rockfill embankment, particularly
differential settlements related to staged construction or zoning of materials with con-
trasting moduli.
Early CFRD dams, such as Courtright, Salt Springs and Wishon, suffered substantial face
slab cracking prior to and during first filling. This damage is attributable to the large settle-
ments (typically in excess of 1% vertical strain) that occurred in the dumped and sluiced
rockfill embankments.
Regan (1997) describes remedial works used in some of these early dams, to reduce
leakage rates through the face slab.
Changes in construction methods in about the 1960s, from these early dumped and sluiced
embankments to well zoned and compacted embankments, have brought about substantial
