Environmental Engineering Reference
In-Depth Information
much smaller contents of sulphides in rocks used commonly in dam construction may pro-
duce undesirable effects.
2.9.4.1
Sulphide oxidation effects in rockfill dams - some examples
Corin Dam (Australian Capital Territory).
Sandstone and quartzite which formed most of
the rockfill and filters in this 75 m high earth and rockfill dam contained about 1% of pyrite
(Haldane et al., 1971). The reservoir started to fill in 1968, with water of pH about 6.5.
During and after filling, seepage from the downstream rockfill was found to be highly acidic
(pH around 4). Both acidity and conductivity were highest after heavy rainfall. It was con-
cluded that oxidation of pyrite in the fill was occurring due to the percolation of rainfall as
well as underseepage from the dam. Haldane et al. (1971) reported that the acid waters
(base flow rate about 25 l/s) were diluted sufficiently by other waters to prevent any prob-
lems downstream.
Testing of the reservoir water in the 1980's and 1990's showed it to be slightly alkaline
and to have an average sulphate content of 0.6 mg/l. Sampling of the seepage showed that
there had been a gradual decrease in its acidity, pH being around 5 by March 1999
(Tabatabaei, 1999, and Tabatabaei, Pers.comm.).
Kangaroo Creek Dam (South Australia).
Under Section 2.4.4 Decomposition of sul-
phide minerals, Fell et al. (1992) reported on observed oxidation of metallic sulphides and
associated fretting in schist exposed in excavations at the site of this concrete-faced rock-
fill dam. During the early site studies the schist was judged to contain less than 1 percent of
sulphides. Based on the absence of deterioration in 40-year old schist rockfill near the site,
the schist was adopted as the main source of rockfill, with some dolomite in the uppermost
quarter of the bank (Trudinger, 1973). It was assumed that pyrite oxidation would not
occur in the constant moisture environment at depth in the dam.
As at Corin Dam, the underseepage includes rainwater which enters the rockfill. The
seepage rate fluctuates widely with rainfall and storage level, but is usually between 0.5 l/s
and 2 l/s. Chemical analyses of the seepage have been made only since 1999 (29 years after
first storing of water). The seepage has a pH of around 7 (slightly lower than that of the
reservoir) but much higher concentrations of calcium and magnesium sulphates than the
reservoir water. These concentrations have been decreasing since 1999 and it can be
inferred that they were higher during the early years of storage. It is considered that
oxidation of the pyrite, and acid attack on the dolomite, have been occurring since
completion of the dam in 1970. However, the embankment is performing satisfactorily,
and the effects of the seepage on the quality of the water downstream have not been
significant.
Carsington Dam (Derbyshire, UK).
This dam failed by upstream sliding during con-
struction, in 1984. It was a 38 m high zoned rockfill structure (Cripps et al., 1993, and
Banyard et al., 1992). Its clay core (extremely weathered mudstone) contained 0.8%
pyrite and 0.25% calcite. The shoulders, of moderately weathered mudstone, had 5%
pyrite and 7% calcite. Drainage blankets were of limestone. The dam was reconstructed
between 1989 and 1991. During the reconstruction it was revealed that sulphuric acid
generated from the pyritic fills had attacked the limestone and resulted in blockage of the
drains by precipitates of gypsum and iron hydroxide. Also, carbon dioxide had accumu-
lated in drains, manholes and excavations. Seepage from the dam had high concentrations
of sulphates, iron and other metals. While these features had not been responsible for the
failure, they were undesirable, and the new design included the following features aimed
at reducing or counteracting the effects of pyrite oxidation:
-
Allowance for degradation, in the embankment shear strength parameters;
-
Non-calcareous fill used for the drainage blankets and filters;
-
Concrete structures protected by coatings of bitumen;
