Design, specification and construction of filters - Geotechnical Engineering of Dams - page 390

9.7.2.3 Permeability requirement
There are several views on the selection of geotextile permeability including:
Giroud (1996) suggests

k i k

(9.9)

Gs s

where
k G permeability of the geotextile

i s hydraulic gradient in the soil adjacent the geotextile (which may be greater than 1)

k s permeability of the soil.

Giroud recommends a large safety factor (e.g. 10) should be used to allow for uncer-
tainties in the permeabilities.
Christopher et al. (1993) suggest

k G 10k s (9.10)

Heerten (1993) suggests the permeability will be satisfactory if other requirements are
met. Heerten (1984) suggested

k Gclogged k s (9.11)

Loudiere et al. (1982), reflecting the views of the French Committee on Geotextiles and
Geomembranes suggests k G 100k s , where k G is the uncompressed geotextile permeability.

From these data, it would seem wise to require an uncompressed geotextile permeabil-
ity of at least 10k s and preferably 100k s .
The further requirement to maintain sufficient cross geotextile flow capacity is to con-
trol clogging, which occurs when the fine soil particles either penetrate into the geotextile,
blocking off the fine (as well as the coarse) pore channels, or are deposited on the
upstream surface of the geotextile. Heerten (1993) indicates that clogging is more likely
where the base soil is cohesionless sand and silt, gap graded (internally unstable) cohe-
sionless soils and high hydraulic gradients. Clogging is more likely if the geotextile is
much finer than it needs to be to meet filtering requirements and if it is thin, with little
pore capacity.
Christoper et al. (1993) list several criteria to reduce the likelihood of clogging. These
are reproduced in Table 9.14 .
Heerten (1993) suggests the use of the second of his criteria in Table 9.13 (O 90,w D 50B ),

to control the minimum pore size in the geotextile, and a thickness of geotextile, t G ,
from 30 O 90,w

t G 50 O 90,w . He quotes several German organisations as suggesting

4.5 mm 6.0 mm. Thick needle-punched non-woven geotextiles are more likely to

meet these criteria than woven geotextiles.
Christopher et al. (1993) suggest that for critical/severe applications, filtration tests on
the soil and geotextile should be carried out.
The authors are aware of cases where limonite depositing from seepage water has
clogged geotextile filters. Palmeira and Fannin (2002) also warn of chemical and biologi-
cal clogging.

t G

9.7.2.4 Durability requirement
There are two aspects to consider:

(a) Durability during installation - geotextiles are prone to tearing as they are being
placed, or by materials being compacted on top of or adjacent to them;