Geoscience Reference
In-Depth Information
Water pulse injection
R15
R16
R17
R18
R19
R20
Ref
0
SP1
SP32
15.26 m
17.58 m
SP17
SP18
0.2
Ground
Injection hole
Figure 5.43
Experimental setup. The
nonpolarizing electrodes were emplaced around
the injection well. The well consists of a plastic
tube with an open end localized at a depth of
65 cm from the ground surface. The
self-potential data were collected with a
32-channel ultrasensitive voltmeter (BioSemi).
The distance between the Ag and AgCl
electrodes was 7.5 cm, and the well is localized
between electrodes SP17 and SP18. The two
reference electrodes are located roughly 10 cm
close to the well on the side with respect to
the profile shown in this figure.
0.4
Stainless steel electrodes
Ag/AgCl electrodes
1
cm
0.6
Pulse low
0.8
7.0
7.5
8.0
8.5
9.0
9.5
10
Distance (m)
5.5.1 Material and methods
The experimental setup is shown in Figure 5.43 and
consists of a plastic tube in which water was injected
for 4 s with a flow rate of 0.6 L s
−
1
. The self-potential
response during the experiment was recorded using
the very sensitive multichannel voltmeter manufactured
by BioSemi, Inc., and was originally designed for EEG
(
http://www.biosemi.com
/). This instrument is the
same one used in Section 5.3 (Hydraulic Fracturing
Laboratory Experiment), subsection 5.3.2 (Material
and Method), and in Section 5.4 (Haines Jump Labo-
ratory Experiment), subsection 5.4.2 (Material and
Methods), and will not be described again here. Also,
this instrument has been used in numerous other inves-
tigations, including Crespy et al. (2008), Ikard et al.
(2012), and recently Haas et al. (2013) for laboratory
experiments.
Two DC resistivity profiles were also performed prior to
and after the water injection test with the ABEM SAS-
4000 Terrameter impedance meter. The DC resistivity
profile comprised a total of 32 stainless steel electrodes
(50 cm spacing) and 118measurements with theWenner
array data acquisition profile. The apparent resistivity
data were inverted with RES2DINV 3.4 (Loke and
Barker, 1996).
5.5.2 Results
The resistivity tomograms are shown in Figure 5.44. The
data shows small changes in the resistivity close to
the well. These resistivity data are essential for the
inverse calculation of the self-potential to localize the
causative source in the subsurface.
The full time series of the self-potential measurements
(called electrograms; see Crespy et al., 2008) are shown
in Figure 5.45. They are taken for a time window com-
prised between 140 and 148 s (Figure 5.45). We applied
a linear trend and DC offset removal process and brought
the time series of all 32 channels of the data acquisition
system to the same baseline during the preinjection time
window shown in Figure 5.45. The result of this process is
shown in Figure 5.45b and c. It can be seen that this offset
and trend removal process allows the characteristics of
the electrical pulses to be seen much more clearly across
all channels together, along with their correlations.
A snapshot of the resulting self-potential distribution
along with the least-squares inversion result is shown
in Figure 5.46b. It shows a negative baseline with a
positive trending anomaly on the top of this baseline.
The negative baseline corresponds to the fact that the
CMS and DRL reference electrodes are located close to
the injection well. What is really important is the positive
