gen chloride (bleach) to the test pans. The bleach
between the field tests and the typical laboratory
tests that may have been contributing factors.
was added to the test pans on 3 June 1994 to
One key difference was the confining pressure. In
the field, a confining pressure less than 1 lb/in.2
possibility, a free swell test was performed in the
(6.9 kPa) was produced by the thin (10 in. [25 cm])
laboratory using granular bentonite, the pond
layer of cover material. In the laboratory, 1 lb/in.2
water used in these tests, pond water with bleach
of confining pressure or higher is often used. The
added, Milwaukee city tap water, and tap water
lower confining pressure allows the bentonite to
with bleach. The addition of bleach to the water in
swell further and store more water in the double
the free swell tests caused almost a 50% reduction
layers. With more swelling, there is a possibility
in amount of swelling. This change was probably
that the porosity or free water is higher, thus a
ascribable to the collapse of the double layer
caused by adding the hydrogen chloride. It, thus,
Another factor is the gradient. In the laboratory it
appears that the addition of bleach to the Bento-
is common to use a gradient as high as 75 for
mat and Claymax GCL test pans, to control bio-
materials with a very low hydraulic conductivity.
logical growth, caused the large increase in the
The field gradients ranged from 5 to 15. The
hydraulic conductivity observed in June of 1994.
higher gradient in the laboratory means that a
Near the end of the field seepage tests, red dye
large pressure differential exists across the speci-
was placed in the three large GCL test pans to see
men. This may serve to further confine or squeeze
if preferential flow paths through the GCLs devel-
the specimen to keep it at a lower porosity and
oped during the freezing and thawing and seep-
lower hydraulic conductivity.
age processes. The dye appeared in the discharge
Test pans 4, 5, and 6 contained the Gundseal
water from the Claymax and Bentomat within 4
GCL and did not produce any seepage in either
days. Water was not seeping through the Gund-
the laboratory or the field. This is attributed to the
seal GCL, and the dye did not appear in its out-
HDPE geomembrane onto which the bentonite is
flow collection pipe.
fixed. The membrane prevents water from flow-
After draining the water from the test pans, we
ing through anything except the seams or a flaw
disassembled them and inspected them for clues
in the HDPE. We included the Gundseal GCL in
to show where the water was flowing. The Clay-
the study to see if freezethaw affected the ben-
max GCL test pan showed red dye staining
tonite in the seam. The two field test pan perme-
throughout the bentonite section. This indicated
ameters that contained the Gundseal GCL with
that the water was passing through the entire sec-
seams did not leak. During disassembly of the
tion; no preferential flow paths developed. The
pans, we observed that water had migrated a rel-
bentonite in the Claymax GCL was hydrated, but
atively short distance into the edge of the seam,
it appeared less plastic than is usually seen in a
but had not fully penetrated.
GCL hydrated under such low confining pres-
sure.
measured in the test pans increased dramatically
The Bentomat test pan showed red dye stain-
to nearly 1 107 cm/s in the Bentomat GCL and
ing through part of the seam area when it was dis-
Claymax GCL test pans. A visual characterization
assembled. This indicated that a preferential flow
of the bentonite in the GCLs at the end of the field
path existed and that water may not have been
test showed a difference in structure from that
flowing at the higher rate through the entire sec-
seen in GCLs from the test ponds. Instead of being
tion. As with the Claymax, the bentonite in this
a homogeneous, plastic, flexible paste, the bento-
test pan did not appear to be as fully hydrated as
nite in the test pan GCLs was found to be almost
is usually the case.
brittle, as though the double layer had reduced.
From the data collected in these field test pans,
As a contrast, specimens of GCL removed from
it is difficult to tell whether the higher-than-
the three large test ponds constructed in 1992 did
expected hydraulic conductivities were caused
not show the brittle structure. The change in
by freezethaw cycles, low confining pressures,
structure is different from that reported in the lab-
low gradients, or other factors. Further large-
oratory after freezethaw tests, so it may have
scale field study is recommended to sort out the
been caused by some factor other than freeze
causes of the increases. The increases in the
thaw.
hydraulic conductivities of the GCLs in the test
One possible explanation for the change in
pans were relatively small, however, and are not
structure may have been the addition of hydro-
considered consequential. The exception, of
15