Page 56 - 83

ENGINEERING

L-CO (Liquid CO ), water and viscous oil, besides
2
2
SC-CO . The all fracturing fl uids were injected un-
2
der the same fl ow rate of 10 mL/min into the cen-
tral interval of 60 mm length of the hole, sealed by
Z a straddle packer with O-rings in the both ends.
Experimental results
Figure 4 shows the changes in injected fl uid pres-
sure, temperature, confi ning pressure, and AE
count rate in the case of SC-CO injection. The
2
temperature shown in the fi gure was measured
Y with a thermocouple glued on the injection pipe just
above the packer in the hole.
The AE sources can be located by using arrival
times of an AE wave to AE sensors, in a similar
way to that for an earthquake. Figure 5 shows the
X
locations of AE sources observed during the fl uid
injections projected onto the three orthogonal pla-
nes, XY, YZ, and ZX. Cracks visible on the two op-
posite surfaces of the XY plane are distinguished by
using dark lines for the near plane that Z = 170 mm

Figure 3 Figure 3 - Coordinate system and loading condition of the specimen. The fi lled circles indi- and light lines for the far plane that Z = 0 mm. The
AE sources were distributed along the cracks with
cate AE sensors at open positions, while the open circles with broken lines indicate those
at hidden positions behind the specimen. (Ishida et al. 2016) scattering, as expected from the observed surface
cracks.
Z The same data as these shown in Figure 4 except
temperature were measured for all specimens
nitored to clarify feature of cracks induced by HF. using the other fracturing fl uid. Their AE sources
Here, AE is defi ned as elastic wave emitted with distribuytions and visible cracks werw shown in the
fracturing of rock and it is basically the same phe- same way as those in Figure 5 (Ishida et al. 2016).
nomenon as the micro-seismicity induced with fi eld
HF operation. Figure 3 shows coordinate system,
Y
the positions of AE sensors and loading condition We can consider SC-CO2 shale
of the specimen. We applied confi ning pressures
gas extraction as a box where
of 3, 6, and 4 MPa in the X, Y, and Z directions,
respectively, to simulate underground rock stress. “the inputs are SC-CO
X
2

The granite specimen has the rift plane where and suitable proppants and
many inherent preexisting microcracks align. The outputs are shale gas and CO
rift plane, which would be a weak plane in HF, was 2

Figure 3 oriented to correspond to the YZ plane in the Car- sequestration
tesian coordinate system.
To compare to effect of HF using SC-CO , we
2
injected fracturing fl uids having different viscosities; Discussion

In Figure 6, we show the dependency on the frac-
turing fl uid viscosity of breakdown pressure, the
AE distribution, and fracturing mechanism based
on ratios of the P wave fi rst motion polarity of AE
events. Because we made two experiments for
each fracturing fl uid in the same condition to con-
Figure 4 fi rm reproductivity of the results, two points are

plotted for the same fracturing fl uids in Figure 6.
The difference in viscosity of the same fracturing
fl uid indicated by the two points was due to the
difference in temperatures when each experiment
was conducted.

1

Breakdown pressure

Breakdown pressure is defi ned as a peak pressure
immediately before the injected pressure shows an
Figure 4 - Changes in injected fl uid pressure, temperature, confi ning pressure, and AE abrupt drop due to fracture initiation in the pres-

count rate during injection of SC-CO2. (Ishida et al. 2016) surized central interval accompanying a burst of
Figure 4

54 54 Impiantistica Italiana - Settembre-Ottobre 2020

1