Jökull - 31.12.2001, Page 50

Bryndís Brandsdóttir et al.
Microearthquake activity within geothermal sys-
tems in Iceland is highly variable. Some high-
temperature systems exhibit a fairly constant activity
(a few microearthquakes per week) whereas swarm
activity characterizes other systems. The Svartsengi
geothermal field has been seismically quiet for more
than a decade. An analog seismometer, operated at the
southern margin of the geothermal field since 1984,
has not recorded any seismic activity within the well-
field, although it has recorded many small swarms
within adjacent parts of the Reykjanes Seismic Zone.
The first seismograph in Iceland (Mainka, NS-
component) was installed in Reykjavík, approxi-
mately 40 km from Svartsengi, in 1909. An E-W
component was added in 1913. These were long-
period, low-gain sensors operated intermittently un-
til 1914. The station was shut down in the be-
ginning of the first world war and not reopened
until 1925. When three, high-gain, short-period,
three-component, Sprengnether seismometers were
installed in Reykjavík in 1951-1952 the detection
threshold of earthquakes at the Reykjanes peninsula
was lowered to magnitude 2-2.5. The catalogs of the
Icelandic Meteorological Office, list two earthquakes,
magnitude 4.1 and 4.2, which most likely originated
in vicinity of Svartsengi on Septemberl5, 1954 and
four earthquakes with magnitudes ranging between
4.0 and 4.6 on September30, 1967. According to the
bulletins of the Reykjavík seismic station earthquake
activity at the westernmost Reykjanes peninsula in-
creased markedly during 1967-1970 and remained
high during the period 1971-1975 during which many
dense swarms occurred along the plate boundary be-
tween 21.9°W and 22.7°W (Einarsson, 1991).
Microearthquake activity on the Reykjanes penin-
sula was first monitored during the summers of 1967
and 1968 when portable, one-component, analog seis-
mographs were operated for a few days near the
Reykjanes and Krýsuvík geothermal fields (Ward and
Björnsson, 1971). Considerable microearthquake ac-
tivity, close to 20events per day, was observed in
vicinity of the stations but no earthquakes were de-
tected near the Svartsengi field.
Accurate earthquake locations became available
in 1971-1972, when a portable network was installed
temporarily in the western part of the Reykjanes
peninsula (Klein et al. 1977). A few microearth-
quakes were detected in the Svartsengi field in 1972
but the seismic activity was lower there than in ad-
jacent areas, i.e. the Reykjanes and Krýsuvík geother-
mal fields. One focal mechanism was obtained, show-
ing normal faulting with a strike-slip component, i.e.
a left lateral shear assuming an ENE-striking nodal
plane for an earthquake at 3.2 km depth.
We operated an array of 15portable, digital seis-
mographs within and around the Svartsengi geother-
mal field, during May-August 1993 (Figure 1, mid-
dle), in order to monitor microearthquake activity
prior to and during the waste water injection test. Re-
fraction and seismic noise measurements were carried
out simultaneously in order to obtain information on
the crustal structure and ground noise associated with
the geothermal field.
Seismicity along the Reykjanes Seismic Zone dur-
ing the recording period in 1993 was mostly con-
fined to the Krýsuvík geothermal field and Fagra-
dalsfjall, 8-10 km east of Svartsengi (Figure 1). No
earthquakes originated within the Svartsengi field.
A few microearthquake swarms have originated at
Fagradalsfjall in recent years with hypocenters rang-
ing from 0-8 km depth. No focal mechanisms are
available for these earthquakes but their origin is
most likely related to strain at the plate boundary as
there is no surface geothermal activity in this region.
No spatial or temporal correlation can be drawn be-
tween the observed seismicity and the injection test.
The regional groundwater flow, from higher lying re-
gions towards the coast is perturbed by the NE-SW
and NNW-SSE lying normal faults and open fissures
which channel the flow. The seismicity thus occurred
“upstream” from the Svartsengi geothermal field. Fur-
thermore, geodetic measurements indicate that the
pressure drawdown has not affected the Fagradals-
fjall region. The subsidence anomaly is elongated
along the plate boundary, i.e. NNA-SSW from Svarts-
engi but gravity changes are not detected 6 km east of
the wellfield indicating that its eastern margin is less
permeable (Eysteinsson, 2000; Björnsson and Stein-
grímsson, 1991).
48 JÖKULLNo. 51