Project no.: 518378
Project acronym: I-GET
Project title: Integrated Geophysical Exploration Technologies for
Deep Fractured Geothermal Systems
Instrument:
Thematic Priority:
STREP
Sustainable development, global change and ecosystems
Publishable final activity report
Period covered: from 01/11/2005 to 30/04/2009
Date of preparation: 27/08/2009
Start date of project: 01/11/2005
Duration: 42 months
Project coordinator name: Ernst Huenges
Project coordinator organisation name: GFZ Potsdam
Revision [draft]
Publishable final activity report
Part of the project was to evaluate existing exploration methods, applied at representative sites
throughout Europe. For this purpose, a database has been compiled containing all available
information on previous exploration efforts (workpackage 2). This database is on the project
website (www.i-get.it) and available through I-GET partner BRGM, who were responsible for
the compilation of the report. The success or failure of these methods are discussed and
recommendations have been derived from these experiences as well as from those gained in
the field experiments performed within the project in a best practise guide for geothermal
exploration (workpackage 6).
The Travale test site is part of the extensive reservoir used since 1904 for energy production
at Larderello. The productive zones of the deep reservoir occur mainly within contact
metamorphic carbonate rocks above a granite intrusion. Less productive fractured levels are
also found inside deeper granitic bodies, where temperatures do not exceed 330°C. Deep
wells did not meet productive levels where recorded temperatures exceed 350 °C. The lowest
boundary of this reservoir still represents an open question.
To better understand the extent of the reservoir at Travale and to predict the sustainability of
the resource extensive field operations were performed at the site prior to and within the IGET project.
Figure 1: 3D distributions of temperature, pressure and steam quality for natural state.
Existing seismic lines were reprocessed by partner UNIPI. In addition, magnetotelluric (MT)
data were acquired by partner IGG parallel to some of the seismic profiles. Geophysical
measurements were compared with information from well-logging, which was also performed
within the project. The existing geophysical data were the basis for a final reservoir model at
the regional scale, calculated and developed by I-GET partner ENEL (Fig.1). This numerical
simulation granted a better understanding of the geothermal processes that are the basis of the
continuous steam supply in the Travale system.
The comparison between distributions of the simulated and observed values of temperatures
and pressures, both for natural state and production history, provided satisfactory results.
In addition, temperature vertical profiles and pressure evolution with time show good fitting
between simulated and observed data.
The good results have allowed using this regional modeling to predict future evolution
with time. In particular, Travale field production has proved to be sustainable for at least
the next 50 years.
The volcanic site at Hengill, Iceland is currently expanded to increase energy output for the
rapidly growing Aluminium industry in Iceland. MT surveys at Hengill have been performed
by IGET partner ÍSOR. At Hengill, an extensive MT survey was carried out. A total of 70 MT
soundings were made within the framework of the I-GET project. 117 pre-existing soundings
and additional 30 soundings funded by Reykjavik Energy are available to the project. The IGET project therefore includes a total of 217 MT soundings in the Hengill area. Almost all
the MT soundings were made at locations previously visited by TEM soundings which was
used for static shift corrections of the MT data. Joint 1D inversion of all MT and TEM
soundings in the Hengill area was performed and a 3D reisistivity structure was compiled
from the individual 1D models (Fig.3).
Finally, a full 3D inversion was done for the full MT tensor, corrected for static shifts by
TEM. The 3D inversion revealed very interesting resistivity structure, showing deep
conductors (3-9 km) under the geothermal system. These conductors correlate with transform
tectonics revealed by intense seismicity from 1991 to 2000 and an area where an uplift
occurred in the same period. The deep conductors are thought to be very hot intrusions and
heat sources for the geothermal system(s) in the Hengill area.
Figure 2 Comparison of electrical resistivity and seismic velocities (P-waves) measured at the Hengill
volcanic site. Isolines of P-wave velocities in black.
In addition to electromagnetic measurements, a broad-band passive seismic survey was
carried out (Partners: BRGM and ÍSOR). Seven broad-band seismic stations were installed in
and around the most seismically active part of the area and recorded continuously for four
months. The data base includes 662 earthquakes recorded by at least 4 of the stations. 424 of
the recorded events were micro-earthquakes with clear P and S waves. Finally, 19 low
frequency earthquakes were recognised with sharp onset and resonance having a large peak at
about 1.5 Hz. The velocity structure of the Hengill area, as revealed by the joint 3D
tomography inversion, was compared to other geophysical data. The location of high P-wave
velocity coincides with an area where a deep conductor is at shallowest depth (about 3 km),
according to the results from the 1D inversion of TEM and MT soundings and 3D inversion
of MT soundings. No sign of attenuation of seismic S-waves is observed under the Hengill
area, indicating absence of extensive magma chambers at depth.
At Gross Schoenebeck, seismic experiments included two parts: First a 40 km long profile to
derive a regional 2-D seismic model of the potential reservoir layers and overlying sediments.
The profile is centred at the geothermal drill sites GrSk 3/90 and GrSk4/05 and is oriented
parallel to the estimated strike of the regional stress field. Second, a star-like arrangement
consisting of 4 profiles each of 6 km length was deployed and a low-fold (low budget) 3-D
seismic experiment was conducted to identify fractures around the geothermal well location
(partners: GFZ and UNIPI).
The signals of the first arriving P-waves from the long profile are used for the inversion of
travel times to determine P wave velocities and vertical gradients, and represent the input data
for the attenuation tomography. Both the P wave velocity tomography and, in particular, the
derived vertical gradient of the velocity reveal new aspects how sedimentary sections can be
imaged.
In addition to the surface geophysical measurements, borehole geophysics was used to
complement the set of field data. For this purpose, a chain of seismometers was located in the
well GrSk 3/90 while stimulation tests were performed in the other well at the same site. The
data of this test are currently processed.
MT measurements were carried out along a 40 km long profile coinciding with the seismic
profile, with 58 MT stations; 20 new MT sites were recorded along a 20-km profile located 5
km to the east of the long profile. In addition, 159 new sites were recorded on a 3 km x 28 km
grid around the Gross Schönebeck borehole, with a site spacing of 500m x 500m. Data
analysis indicates a mainly two-dimensional conductivity structure. Interpretation of the
results yields generally a good agreement with the known geology of the area. Several zones
of high conductivity coincide with structural lows, probably due to sedimentation of more
porous material or fluid accumulation. Two distinct conductors located at depths of 4-5 km
coincide with the Lower Permian reservoir rocks
An interpretation of both seismic and MT measurements is shown in figure 3. The
geophysical data are fit to a pre-existing geological model. For a joint interpretation of
seismic and MT data a cross-plot was generated (Fig. 4a), where seismic velocities were
plotted against electrical resistivity data for various sections of the geological profile. Specific
domains, indicating a common source for specific seismic velocity and electrical resistivity
values could be determined that way. They were used to identify the corresponding units in
the geological profile (Fig. 4c).
Rock physics experiments at high pressure and temperature were performed at the high
pressure laboratory at GFZ Potsdam, where conditions in the reservoirs can be simulated, with
controlled pressure, temperature and pore pressure as well as pore fluid salinity (WP 4). Rock
samples from Iceland, Germany and from Italy were subjected to conditions comparable those
in the geothermal reservoirs investigated, while electrical conductivity and seismic velocities
were determined on the cores. By varying the difference between pore pressure and confining
pressure at a given temperature, the transition from liquid to steam in the pore fluid was
Figure 3 Gross Schönebeck: Comparison of velocity (Vp) model obtained from travel time inversion
and electrical resistivity model obtained from inversion of the magnetotelluric data.
Figure 4 Cross-plot of the probability density function in gray-scale plan view (a) and three-dimensional
view (b). Spatial distribution of classes (c). Colours correspond with the ellipses in (a) defining
the class boundaries.
simulated while changes in physical properties were observed. The raw data acquired by these
experiments have been further analysed and interpreted by partner FUB in terms of their
elastic properties.
These measurements serve to calibrate results from well logging and from field geophysical
observations and help our understanding of processes leading to the observed results.At
Skierniewice, field work around the deep well Kompina-2 was completed within the third
year of the project. The seismic measurements, performed by a commercial company yielded
some completely new information about the potential reservoir. Interpretation of the highquality data set of a low-density 3D-grid provides the basis of ongoing work, beyond the end
of the project. First results show structural elements dominated by vertical movements and
salt tectonics. Structural maps of the Triassic horizons revealed normal faults along the NWSE direction. The main fault and four minor ones were recognized, all of them can be traced
up to the top of the Middle Jurassic formations. In the western part a wide syncline extends in
NW-SE direction (Fig. 5)
The MT survey was very much influenced by artificial electromagnetic noise. The main
sources of noise are two electric railway lines running south and west of the survey area, highvoltage power lines that crossed the survey area, and numerous sources of cultural noise.
Strong noise signals are very difficult to remove with the use of applied processing
procedures. They affect mainly low frequency data over the whole survey area and cause a
great statistic scatter of estimated parameters as well as probable substantial curves
deformation in some measurement sites. As a result, the reliability of low-frequency data is
significantly lowered. Unfortunately, the main target of the survey is located at depths relating
to low-frequency range of data.
To improve data quality, CSAMT measurements were performed to complement the MT data
with a controlled source of electromagnetic signals (Fig. 6). This procedure improved data
quality in the upper section of the geological profile significantly. For the lower part, two
options of robust data processing were applied to improve the quality of calculated
parameters. First, the coherence between time series of magnetic field components recorded
on survey and remote sites was applied. Then data were reprocessed without the use of the
coherence function. The reprocessing results appeared to be a little better than results of the
first applied option, but finally, the quality of obtained estimations of magnetotelluric
parameters is rather poor for low-frequency range and the results of interpretation can prove
unreliable.
A joint interpretation of all available data included the newly acquired seismic and MT
profiles as well as information from well-logging. Sets of resistivity maps and structural maps
related to horizons interested for geothermal investigations were prepared. Zones of lower
resistivities in central and southeastern parts of the area are probably connected with strong
fracturing or high porosity forming ways for filtration of geothermal water. Such
interpretation is confirmed by the Kompina-2 borehole and location of faults in the seismic
map. Vertical boundaries of resistivity more or less correspond to faults interpreted from
seismic data.
These encouraging results give a much clearer picture of the potential reservoir. The quality
of the seismic data and the information compiled within the project provide a useful basis for
an evaluation of the further development of the well.
Fig. 5 Spatial view of the top seismic boundary of Buntsandstein in depth version (Kompina-2 well marked by
red line, data: Geofizyka Krakow Ltd.)
Fig. 6
Complementary results of 2D inversion of MT profile in Skierniewice - A (top)
Including CSAMT - B (bottom)