Design, specification and construction of filters - Geotechnical Engineering of Dams - page 370

Table 9.8. Criteria for the excessive and continuing erosion boundaries (Foster, 1999; Foster and

Fell, 1999a, 2001).

Proposed criteria for Proposed criteria for

Base soil excessive erosion boundary continuing erosion boundary

Soils with D 95B 0.3 mm D 15F 9 D 95B

Soils with 0.3 D 95B 2mm D 15F 9 D 90B

Soils with D 95B 2 mm and Average D 15F D 15F which gives For all soils: D 15F 9D 95B

an erosion loss of 0.25 g/cm 2 in
the CEF test (0.25 g/cm 2 contour
line in Figure 9.23)
or
Coarse limit D 15F

fines content 35%

D 15F which
gives an erosion loss of 1.0 g/cm 2
in the CEF test (1.0 g/cm 2 contour
line in Figure 9.23)

Soils with D 95B 2 mm D 15F 9 D 85B

and fines content 15%

Soils with D 95B 2 mm D 15F 2.5 D 15F design, where D 15F

and fines content 15-35% design is given by: D 15F design

(35-pp% 0.075 mm)(4D 85B -0.7)/
20 0.7

Table 9.8 shows the criteria for the excessive and continuing erosion boundaries.
The criteria for the excessive erosion boundary are selected from the case studies and the
laboratory testing, and dams which experience erosion to this limit may have large piping
flow discharges - up to say 1 to 2 m 3 /sec. Whether a dam can withstand such flows with-
out breaching, depends on the discharge capacity of the downstream zone and whether
unravelling or slope instability may occur. It is also likely that, if a dam experiences a pip-
ing event which does eventually seal, it will experience another later, as the erosion
process moves laterally.
The criteria listed in Table 9.8 should be used with caution and, for final decision mak-
ing, should be supported by laboratory tests using the filter/transition and core materials
from the dams. The results should be tempered with sound dam engineering judgement.
The procedure has been used in a limited number of dams and has proven to be a valu-
able aid to decision making. Bell et al. (2001) describe the assessment of internal erosion
and piping for Eucumbene Dam. Figure 9.24 shows the particle size distributions of the
core, the particle size required for a filter designed according to Sherard and Dunnigan
(1989) and filter/transition materials tested in the continuing erosion tests. These distribu-
tions are truncated at the coarse end to facilitate laboratory testing.
Tests on these materials all eventually sealed, although considerable erosion occurred.
There are some broader implications to these studies in that it is apparent that, for
most soils, there is a considerable margin between design no-erosion and continuing ero-
sion criteria. This assists in explaining why, despite the statistical variability in the parti-
cle size distributions of the base soil and filters, (which might on first consideration imply
that there was a significant potential for piping to occur), this is unlikely because the no-
erosion and continuous erosion boundary criteria are not exceeded. Figure 9.25(a) shows
this conceptually.
However it is important to recognise that for fine ground base soils, the separation
between the design limit, no-erosion and continuing erosion boundaries is much less -
Figure 9.25(b) and Tables 9.2 and 9.8. Hence if the filter design is close to the limiting
D 15F /D 85B , very close control on construction will be necessary.