26
Geotechnical News • September 2016
GEOSYNTHETICS
yield erosion of filter sand on the steep
sidewalls and result in construction
delays.
Specification of the geotextile
filter
The original spillway lining was
founded directly on a sequence of
sand and silts of the Fort Langley
Formation. Drilling in the period
1979 to 1990 established the geologic
sequence to consist of very dense out-
wash sands overlain by inter-layered
fine sands and silts which, in turn, are
overlain by stiff over-consolidated
clayey silts. No bedrock was encoun-
tered within 100 m of ground level.
Additional drilling in 1991 included
two mud-rotary drill-holes on the
right bank, three auger-holes through
the spillway forebay and ogee areas,
and seven large diameter churn-holes
(wells) that were drilled around the
plunge pool area. The strength of the
silts and inter-layered sands and silts
along the alignment of the spillway
varies from firm to very stiff.
The primary function of the geotextile
is filtration. It must protect against any
unacceptable movement of base soil
through it and into the void space of
the drain, without adversely imped-
ing groundwater seepage across it
and into the drain. Grain size analysis
on samples retrieved from boreholes
indicated the base soil to be sandy
silt, with a mean grain size 0.01 ≤ d
50
≤ 0.05 mm, a value 0.04 ≤ d
85
≤ 0.6
mm and a coefficient of uniformity C
U
≈ 20. The results of Atterberg limits
testing classified the base soil as low
plasticity silt (ML).
The under-drain below the sidewall
of the spillway is required to handle
inflow from two separate sources:
• groundwater seepage from the base
soil; and,
• leakage through joints or cracks in
the concrete liner during spillway
operation.
For purposes of design, it was
assumed the geotextile filter is subject
to steady unidirectional flow from
the base soil into the drain. Three
candidate geotextiles were selected
in order to evaluate filtration compat-
ibility with the base soil. They were all
needle-punched nonwoven geotextiles.
The preliminary selection was made
on the basis of values for pore-size
opening, cross-plane permeability and
strength reported in technical literature
by the manufacturers.
“A significant
finding of the
filtration
compatibility
testing that was
common to the
HCR results, and
also to the GR
results, was the
absence of any
continuous or
significant piping...’’
At the request of BC Hydro, the char-
acteristic value of pore size opening
was independently verified for the
candidate geotextiles with reference
to the Filtration Opening Size (FOS),
a hydrodynamic sieve test. The index
test was performed by a commercial
laboratory in Canada, on samples of
candidate geotextile provided to BC
Hydro by the respective manufactur-
ers. A similar value of O
95
≈ 0.07 mm
was determined for each of the three
geotextiles. Filtration compatibility of
three candidate geotextiles was then
evaluated through program of Gradi-
ent Ratio (GR) testing (after ASTM
D 5101) and Hydraulic Conductivity
Ratio (HCR) testing (after Williams
and Abouzakhm, 1988; now standard-
ized as ASTM D 5567). The testing
was performed by a commercial labo-
ratory in the USA, using samples of
the base soil taken from site in combi-
nation with the candidate geotextiles.
The base soil against which the
geotextile is placed is non-plastic silt.
Given the characteristic grain size of
the base soil (0.04 ≤ d
85
≤ 0.6 mm),
and given the index pore size open-
ing of the candidate geotextiles (O
95
≈
0.07 mm), the capacity for soil reten-
tion was identified as very important
to a confident evaluation of soil-geo-
textile compatibility. Analysis of the
filtration compatibility test data placed
considerable emphasis on the relation
between hydraulic conductivity of the
soil-geotextile composite zone over
the duration of the test. More specifi-
cally, interpretation of compatibility
was based on the variation of hydrau-
lic conductivity with elapsed time, and
also its variation with pore volume
exchange across the geotextile. A
significant finding of the filtration
compatibility testing that was com-
mon to the HCR results, and also to
the GR results, was the absence of any
continuous or significant piping of soil
from the reconstituted test specimen
through the geotextile. Furthermore,
all three candidate geotextiles yielded
a very similar performance. The
original development of the HCR test
was intended for soil with a hydraulic
conductivity less than or equal to 5 x
10
-2
cm/s. The hydraulic conductivity
of the non-plastic silt at the Alouette
Dam was found to be considerably
lower than this value. Accordingly,
greater emphasis was placed on the
HCR test results, when evaluating the
compatibility of the nonwoven geotex-
tile and silt for purposes of design.
Results of the laboratory testing were
used to inform the selection of mate-
rial properties reported in the speci-
fication documents of the contract
for the project. The geotextile was
required to have an opening size 40
μm
≤ O
95
≤ 75
μm
. Requirements for
material strength, permittivity and
UV degradation were established with
reference to routine design guid-
ance at that time. From comparison
to current practice, they would be in
