Geoscience Reference
In-Depth Information
signals can be easily observed. The fluid control system
injects fluid through stainless steel tubes (Figure 5.15)
using a computer-controlled Teledyne Isco 100DX
syringe pump that is able to control flow rate or pressure
during the various phases of the experiment. The injec-
tion tube was designed to have an open end at the bot-
tom; there were no side ports for fluid to flow through.
The system has a total fluid capacity of 103 ml and is
capable of achieving pressures up to 68.9MPa while
maintaining constant flow rates of 0.001
This did not affect the raw data saved or any subsequent
use of the data.
5.3.3 Observations
Figure 5.17 shows the temporal evolution of the electri-
cal potential for all of the electrodes, including the
occurrence of bursts in the electrical potential that
are similar in shape (but much larger in amplitude) to
the electrical field bursts observed by Haas and Revil
(2009) for Haines jumps during the drainage of an ini-
tially water-saturated sandbox. Figure 5.17a shows the
entire 2086 s record, while Figure 5.17b, c, and d zoom
in on specific areas of interest. There are seven major
events, of which three are highlighted (Events E1
through E3), and two will be used in the following text
to test our localization procedure. These events are
shown in the time series of Figure 5.17b, c, and d. All
major electrical potential events occurred during the
phase II constant flow injection period. During phase
I, the measured electrical potential gradually increases
as fluid is being injected into the cement block. No bursts
in the electrical field were observed during the constant
pressure phase (phase I). Other types of information are
contained within this part of the experiment, but are not
analyzed here.
Each major event is characterized by a rapid change in
the electrical potential time series followed by a slower
exponential-type relaxation of the potential with a
characteristic time comprised between several seconds
to several tens of seconds. This relaxation is believed to
be associated with the relaxation of the fluid flow under
pressure as shown later. Because the relaxation of the
potential distribution is relatively slow after each event,
a sequence of overlapping events causes a superposition
of the potentials from each event in the sequence to vary-
ing degrees (see Figure 5.17b and c). We term the super-
position of a past event decay response with a new
event a residual potential superposition. It can be clearly
seen from Figure 5.17 that the degree of residual poten-
tial superposition is dependent on event physics (hydro-
electric coupling), event magnitudes, event spatial
distribution, time of occurrence, and event decay rate.
Each of these factors is variable, and to localize and char-
acterize individual impulsive events, the influence of
residual potential superposition must be accounted for
and removed to complete a comprehensive analysis of
the data.
60 ml min
−
1
.
In this experiment, the injection tubes were initially pres-
surized to a PID controlled 1.17 kPa with the fracturing
fluid and maintained at that pressure for a period of time
to be sure that the system was maintaining pressure and
to measure the static fluid leak-off rate. If fluid was meas-
ured to be moving, then there was some sort of leak-off
into the block or the seal was failing to hold enough pres-
sure to commence hydraulic fracturing. A constant fluid
flow rate of 1 ml min
−
1
was then imposed on the system
with the intent of inducing hydraulic fracturing. Under
constant flow, either the cement block or the tubing seal
would eventually fail.
The test procedure began by preparing the cement
block for high-pressure injection (see Frash & Gutierrez,
2012, for details). The injector was filled with the saline
solution described earlier and coupled to the injection
tube that was also filled with the saline solution. The
injection system was purged of air and then subjected
to constant pressure of 1.17 kPa for about 30min to mon-
itor leak-off to be sure that there was no pressure loss. For
the experiment associated with Hole #9, a 60 s preinjec-
tion (termed phase 0) electrical potential measurement
period was acquired (Figure 5.17a). The goal of this phase
was to establish individual channel offsets and drift
trends for use during postacquisition signal processing.
Constant pressure fluid injection at 1.17 kPa (termed
phase I) was initiated at T0 = 60 s and terminated at
T1 = 1632 s. This phase was followed by phase II, a
1 ml min
−
1
constant flow rate initiated at T2 = 1795 s
(note that fluid pressure was maintained, but not actively
controlled between T1 and T2.). Fluid injection was
terminated well after the end of the electrical data acqui-
sition, when seal failure was confirmed through the
appearance of water on the surface of the block near
the injection hole. For this experiment, self-potential
data acquisition terminated at 2086 s, prior to completion
of phase II injection due to an accidental interruption of
the streaming data because of a poor USB 2.0 connection.
-
