Geotechnical News • March 2015
35
GEOTECHNICAL INSTRUMENTATION NEWS
bottom to prevent vertical movement
of the piezometers before the grout
set. The top of the tremie pipe was
surveyed after grouting to provide an
accurate location and elevation.
Grouting
Portland cement, water, and sodium
bentonite powder were blended with
a cement to water ratio by weight of
1:2.5, per DGSI (2009). The cement
and water were mixed first, with
bentonite blended in afterwards as
required to achieve a consistency
suitable for tremie pumping. A hose
was connected to the tremie pipe and
the grout pumped in as the drill casing
was slowly removed.
Cabling and data collection
The cables from the vibrating wire
piezometers were threaded through
galvanized steel pipes for protec-
tion and weight and then laid on the
channel bottom to the bank as shown
in Figure 4. A data acquisition system
was installed in a steel job box as
shown in Figure 5. The job box was
weighted with concrete blocks and
padlocked to discourage theft. The
data acquisition system was pro-
grammed to take readings at 15-min-
ute intervals to provide adequate data
during storm events, which typically
cause the creek elevation to peak in
3 to 6 hours. The stilling wells and
on-shore standpipe piezometers were
monitored using vented water level
loggers, also programmed to collect
readings every 15 minutes.
Evaluation of in situ hydraulic
conductivity
The on-shore open standpipe piezom-
eters in each cross-section were
tested using falling and rising head
tests. These tests showed that the
soils around these piezometers have
hydraulic conductivities that range
from 3 x 10
-5
cm/s to 2 x 10
-2
cm/s,
with most between 5x10
-4
cm/s to
2x10
-3
cm/s. Grain-size analyses of the
materials obtained during the drilling
of the in-stream piezometers indicated
that the creek sediments in which the
fully-grouted piezometers were bed-
ded would also likely be in this range.
Since we expected that the grout mix
permeability would be about 1x10
-6
cm/s, we determined that the fully-
grouted piezometers should provide
accurate readings with good response
times. The research of Contreras et al
(2102) confirms that this assumption
was appropriate.
Data analysis
Barometric pressure measurements
were obtained from a local meteoro-
logical station and used in the calcu-
lation of the piezometric pressures
measured by the unvented vibrating
wire piezometers. This permitted
direct comparison of the piezomet-
ric data between the fully-grouted
piezometers and the vented water level
loggers in the standpipe piezometers
and stilling wells. Contreras et al.
(2012) provide a good discussion on
the importance of incorporating baro-
metric measurements into vibrating
wire piezometer measurements.
Only one fully-grouted piezometer of
the 12 installed showed anomalous
results. A bottom piezometer had sig-
nificantly higher piezometric pressures
than the on-shore piezometer at about
the same elevation, and it showed an
upward hydraulic gradient greater than
1. The boring log for the vibrating
wire piezometer installation indicated
a 0.1 m layer of running sand, and
water inflow was observed during
drilling at the installed elevation.
We were unable to determine if the
anomalous readings were a real local
phenomenon, or simply an instru-
mentation error. During design of the
stream remedy, neither interpretation
created a challenge so the issue could
remain unresolved.
Figure 2. Drill rig on barge.
Figure 3. Installation of vibrating
wire piezometer and tremie pipe.
Figure 4. Stilling well and pipe pro-
tection of cables.
Figure 5. On-shore monitoring loca-
tion and on-shore open standpipe
piezometers.
