Environmental Engineering Reference
In-Depth Information
However, not all dams have instruments sufficient to monitor the performance of all of
their filters. Also the critical filters (Figure 9.2) may be seriously tested only in an emer-
gency, e.g. when a crack develops in a dam core or the foundation beneath either of the
dam shoulders is disrupted.
(a)
Change of grading due to dissolution
. The critical filter zones in most large dams are
often less than 2.8m wide. According to
Table 3.6
, if pure water was to flow at a rate
of 3 3 1024m/s through such a zone formed by 50 mm diameter calcite particles, it
would lie entirely within the “solution zone”. That is, the water would never become
saturated with carbonate, and the particles would be progressively reduced in size, by
dissolution. With much smaller particles, as in a fine filter zone, with acidic water and
an aggressive reaction the size reduction might become significant during the life of
a dam.
(b)
Partial dissolution and recementation
. There have been many accounts of recementa-
tion of crushed carbonate materials, due to acids formed by the oxidation of sulphide
minerals (See Section 2.9.4). Several examples have been in South Australia where
cemented zones have occurred in pavements, embankments and stockpiles formed by
crushed rocks of Cambrian and Precambrian Ages. Tests have shown the cementing
materials to be calcium sulphate (gypsum) in crushed marble and magnesium sulphate
in crushed dolomites. Both the marble and dolomite have performed well as base
courses in pavements.
The minus 30 mm crushed marble is well graded and contains 5% to 10% fines (all
calcite) with plasticity index of 2. Despite Los Angeles Abrasion Losses of 60% or
more, its performance in pavements is remarkably good. Falling Weight Deflectometer
tests on a sealed pavement built in 1998 showed its initial modulus to be 700 MPa.
The same test after 4.6 years of service showed 1700 MPa. There was no rutting, per-
manent deflection or change in density (Andrews 2003). The authors consider that the
formation and deposition of the gypsum cement, and possibly some associated inter-
locking of the marble particles due to pressure solution (
Figure 3.33
)
have been the
main causes of this increase in stiffness with time.
(c)
Interlocking of grains due to pressure solution
. Both gypsum and magnesium sulphate
are highly soluble in water. This raises the following question. Assume that a chimney
zone filter has become cemented by either of these during a period of very low and
zero seepage. If now increased flow occurs through a crack in the core and cemented
filter, how would this filter behave as its cement becomes redissolved? The results of
testing by Hazell (2003) suggest that the filter might retain the crack, or “hang up”.
The test used was developed to assess the effects of chemical stabilising agents in base
materials for unsealed roads. Samples are prepared by removal of the plus 2.36 mm
fractions and compacting the remainder at OMC to 98% MMDD in 100 mm long by
100 mm diameter moulds. After removal from the mould samples are bench dried for
7 days. Each is then loaded by an annular weight of 1560 g, through which water
drips on to it at a rate of about 400 ml/h. Hazell (2003) advises that during this test
all other local non-cementing road base materials (most are crushed quartzites) collapsed
within 24 hours, but the crushed marble remained intact, although completely saturated
and soft to touch, for at least 460 hours. Assuming that by this stage the gypsum cement
would have been completely dissolved, then the sample must have gained strength from
some other process. The authors suspect that particle interlock may have developed due
to pressure-solution at the grain contacts.
The process and effects of pressure solution in quartz sands in geological time frames
are described by Dusseault and Morgenstern (1979), Palmer and Barton (1987) and
Schmertmann (1991). Harwood (1983) discusses the role of this process in the reduction
