Jökull - 31.12.2001, Page 51

Seismic monitoring during a wastewater injection experiment in Svartsengi
day 225 250 275 300 325 350
Figure 3. A histogram showing the weekly injection
volume into borehole H-6, from July 21 through
December 1993 (Julian days 200-365) and the cumu-
lative injection volume through the injection period.
- Súlurit yfir daglegt rúmmál niðurdœlingarvatns í
H-6, frá 21. júlí til desemberloka 1993 (dagar eru
númeraðir 200-365) og heildarrúmmál vfir sama
tímabil.
RESULTS
A total of 217,6 thousand tons of water (217.6
million liters) of 70-80°C wastewater were injected
into borehole H-6 in Svartsengi from July 19, through
December 1993, at a maximum rate of 301/s (Fig-
ure 3 and Table 1), without generating any detectable
microearthquakes within the geothermal field. The
2000 m deep borehole accepted the injection fluid un-
der gravity, with no pumping being required. The wa-
tertable in H-6, at an average depth of 200 m before
the test, was elevated by approximately 1 m for every
11/s increase in injection rate. The water is consid-
ered to have flowed southwards from the hole as water
from an earlier injection into a borehole 200 m north
ofH-6 was detected in H-6 (Vatnaskil, 1989).
Annual wastewater injection volumes have varied
between 0.07 and 1.73 million tons during 1984-2001
(Vatnaskil, 2002). A total of 9.45 million tons has
been injected into the geothermal reservoir during this
period. A fairly constant reduction in drawdown rate
by about 1.2 bars/year during 1984-1990 is not lin-
early related to the annual injection volume, which
ranged from 0.32 to 1.73 million tons during this pe-
riod amounting to a total of 6.46 million tons. The
Svartsengi geothermal field has been devoid of earth-
quakes since 1984.
Wastewater injection volumes have been shown to
affect geothermal fields in various ways. Relatively
small changes within the local stress field can have
drastic effects on pore pressure and thus completely
alter the failure criteria for a given field. Various rein-
jection tests have, however, failed to show a linear cor-
relation between levels of microseismicity and injec-
tion volume into geothermal fields (e.g. Batani et al.,
1985; Sherburn et al., 1990).
A marked increase in microseismicity was ob-
served within the Wairakei geothermal field in New
Zealand during an injection experiment in 1984 with
pressures reaching 20-30 bars (Sherburn et al., 1990).
Injection into the same field in 1988-1989 did not
trigger any microearthquakes. Pore pressures de-
creased markedly in 1984-1988 and as a result the
Wairakei region accepted the injection fluid under
gravity in 1988-1989, no pumping was required. An
injection volume of 2milliontons or more into the
Larderello-Travale geothermal area in Italy caused
hundreds of microearthquakes there in 1978-1982
(Batini et al., 1985). Furthermore, an injection of
5001/min or less into the KTB deep drilling site
in Germany generated microearthquakes at 8-9 km
depth only 2 hours after the initiation of injection
(Zoback and Harjes, 1997). It is not clear where the
JÖKULL No. 51 49