Geophysical experiences of prospecting a
radioactive waste disposal site
LÁSZLÓ VÉRTESY, ÁGNES GULYÁS, ANDRÁS MADARASI, ENDRE HEGEDŰS, PÉTER
KOVÁCS, LÁSZLÓ SŐRÉS, JÁNOS KISS
EÖTVÖS LORÁND GEOPHYSICAL INSTITUTE OF HUNGARY H-1145 BUDAPEST COLUMBUS U. 17-23.
A great amount of geophysical work was involved in the programme of final disposal of low- and
intermediate-level radioactive waste in Hungary. The program started in 1993. Geophysics was
applied in the program since 1996, from the beginning of site selection phase. The investigations are
now concentrated on the granite body covered by 0–60 m thick Pleistocene loess. The repository is
planned in 250-300 m depth, in granite host rock (see http://www.rhk.hu).
Although a large number of boreholes were drilled, spatial information is expected from geophysical
methods. The geophysical exploration should help to set up the geological, hydrogeological model of
Variety of survey objectives and the geological model required application of many kinds of methods.
In some cases even new procedures, earlier not applied in Hungary, were used, because the set of
methods, which could routinely be applied to geophysical mapping of similar models has not
developed yet. In addition to using up the geophysical results in the present survey, the acquired
volume of data will serve as an information source in the future as well. This cannot be obtained later
any more due to the distortion effects caused by construction of further investigation facilities
(galleries), building of infrastructure (roads, power lines) and putting into operation of the repository.
To provide reliable spatial information for safety assessment is a challenge in lot of country, without
well established answers. That is the reason, the authors believe the topic will have the interest of the
In geology of the survey area rocks of the Mórágy Granite Formation and the overlying
unconsolidated sediments are significantly different in character, therefore their geophysical
investigation necessitated different methods. Geoelectric, electromagnetic, electric imaging and
magnetic methods as well as cone penetration tests were used in the study of the sedimentary
sequence. Magnetotelluric sounding, 3D seismic first break tomography and integrated geoelectric
fault detection were applied in the study of the granitic body.
In what follows these methods and their results are presented. Delineation of the granite body and
based on archive data, study of the major structural lines in the region including the planned site are
discussed in detail. A large variety of bore-hole-log and cross-hole methods, 2D reflection and
refraction seismics applied in the programme is out of the scope of the present lecture. The
geophysical model is based on the results of the 37 drillings and their bore-log data.
Geological surveys performed in the surroundings of the site extended to geological and tectonic
investigation of the larger vicinity, the region. To understand more accurately major structural features
in the wider vicinity of the Bátaapáti (Üveghuta) Site, i.e. in the region, re-processing of archived
gravity, magnetic, magnetotelluric and seismic data was performed. The objective of the study was to
provide geophysical data for delineation of the host granite body, for marking out regional structural
lines. Thanks to the fairly good gravity and magnetic data coverage, the delineation — using a range
of potential field processing methods — could be done.
EAGE 67th Conference & Exhibition — Madrid, Spain, 13 - 16 June 2005
Electromagnetic and geoelectric sounding
In the survey area maximum 50-60 m thick young Quaternary (loess) sediments cover the granite. To
map the granite’s surface geoelectric (VES) and transient electromagnetic soundings (TEM) were
performed. Transient survey was carried out with a Protem–37–47 (Geonics) and VES measurements
with a Syscal Junior R72 (IRIS) instrument. The measurement material at the same time proved to be
suitable to divide the young sediments as well. In the central area measurements were carried out in a
50×50 m regular grid, outside it in a 200×200 m quasi-regular grid. In those areas where the young
sediments’ thickness exceeds 10-15 m the transient electromagnetic soundings, where the cover is
thinner the DC soundings provided good results. Data of both methods were processed with 1D
Marquardt inversion. In the course of this a theoretical sequence of layers was determined whose
calculated reply sounding curve behaves similarly to the curve of field measurements performed over
the real sequence of layers. The theoretical sequence of layers is based on the geoelectric model that
was constructed using joint statistical analysis of well-log data and results of ground-based
geophysical survey performed in the vicinity of boreholes (Figure 1). This model can be extended to
the whole survey area. In it the resistivity well-logs of variable shape are replaced by three major
layers based on the interval resistivities.
Figure 1. The geoelectric model
1 — geoelectric layers separable on the basis of ground-based geoelectric survey (explanation in the text)
2 — borehole geophysical zonation (A-E sedimentological units and G I geotechnical unit)
General characteristics of the area’s geology are:
— resistivity of granite (G I, Figure 1) is significantly higher than that of the overburden, granite can
be considered as geoelectric basement,
— resistivity of the covering sequence decreases with increasing depth,
— this decreasing trend can be divided into two major intervals (marked with 1 and 2) and within each
of them there are two further intervals (marked with a and b).
From data of ground-based survey the three-fold (layers marked with 1, 2 and 3) division could
consistently be deduced. Based on the comparison with well-logging and geological divisions layer 1
can be identified as the Young Loess Sequence within the Paks Loess Formation, and layer 2 with the
Old Loess Sequence. Layer marked 3 is the granite’s (G I) upper zone. Using the above conclusions
relief map of the granite and the Old Loess Sequence could be constructed.
Experiences showed that the density of the applied geoelectric (VES and TEM) grid is not suitable for
division of the overlying sediments and for determination of the geoelectric horizons’ exposition line
on the hillsides. This data system was completed near to the valley bottoms with electrical imaging
along profiles starting from the valley bottom and running perpendicularly to the slope. This method,
in contrast with the soundings suitable for vertical division of the investigated half-space, can rather be
applied to investigate horizontal variability. A Syscal Junior R72 instrument was used that is capable
of switching 72 electrodes. 2D inversion was applied.
As an example a profile measured in a smaller valley west of the northern part of the central ridge is
presented (Figure 2). The high-resistivity granite is overlain by low-resistivity sediments (geoelectric
layer 2), which is covered by an upper loess sequence of higher resistivity (geoelectric layer 1). At the
right side of the section it can be observed that the re-deposited part of this latter formation covers the
underlying one on the steep slope. Shape of the granite surface offers the possibility that the valley is
tectonically preformed, but measurements along parallel profiles do not support this.
Figure 2. 2D inversion along a profile
Red dotted line –granite’s surface from transient and VES measurements. 1 — geoelectric layer 1; 2 —
geoelectric layer 2; G — granite; green triangle — end of geoelectric layer 1; blue triangle — end of geoelectric
layer 2; Scale of axes is given in m. The colour code shows the resistivity in Ωm
The preliminary reconnaissance geomagnetic survey performed in the area called the attention to the
interesting phenomenon that the values measured in the valleys where granite is almost in outcrops,
are systematically lower than those observed along the ridge lines of hills. From this and from the data
of susceptibility measurements carried out in boreholes the conclusion was drawn that granite can be
separated from the sediments based on its magnetic properties (Figure 3). The rather narrow range
(±27 nT) of the observed anomalies, their variability and connection with topography, and the
information derived from well-logging data show that the susceptibility of palaeosols exceeds on
average by 5×10–4 SI units the susceptibility of the host loess layers. This suggests that beside the
regional pattern caused by granite first of all magnetic sources within the sediments should be
assumed. Based on these phenomena, and on the geological model of sedimentary sequence (which
shows a sandwich-like variability of palaeosols and loess) the conclusion was drawn that with the help
of geomagnetic survey (along profiles perpendicular to the hill ridges) it can be made an attempt to
detect the expositions of the palaeosols of higher susceptibility. In the case of adequately dense profile
network a data set suitable to draw neotectonic conclusions can be obtained.
Anomalies were plotted along profiles, and the pattern was correlated all over the area, as magnetic
horizons. After marking out the magnetic horizons an attempt was made to connect the individual
horizons to zones determined from well logging. Disturbances of the horizons can be interpreted as
neotectonic events or landslide.
EAGE 67th Conference & Exhibition — Madrid, Spain, 13 - 16 June 2005
Figure 3. Magnetic profile across a hill built up from young sediments
Red dots with black code — magnetic horizons with code. a — Correlation of the two magnetic horizons marked
out along Profile P-22 on two sides of the hill using the magnetic susceptibility log measured in Borehole Üh-22.
b – Mutual identification of the well-log zones marked with letters and magnetic horizons with the assumption
that the levels are parallel with the granite’s surface
Cone penetration tests (CPTe)
For detecting of neotectonic elements a special version of the cone penetration tests (CPT) was
applied. Beside pressure and radioactive channels, geoelectric data was also recorded. That application
is the scope of two more lectures of this conference.
Granite host rock
The detect and trace fracture zones, possible channels of water, in the granite body (covered with
sediments) is a real challenge for geophysics. Several methods were applied.
Complex geophysical survey at the valley bottoms
To investigate the homogeneity of the granite body complex geophysical profiling was performed
close to the bottom line of valleys surrounding the Site. Basic objective of the survey was to detect the
possible loose zones. In certain sections of valleys of the area, over the granite being in outcrops or
expectedly covered by only a few meters of sediments measurements were carried out with a station
spacing of 5-10 m. In a 400 m long section, which contains anomalies, measurements were carried out
on the hillside as well beside the profile running at the valley bottom, parallel with it. Based on the
anomalies, which can be correlated based on the parallel profiles, strike directions were estimated. To
investigate the chosen profiles electromagnetic (EM) methods ensuring three different investigation
depths and features (EM–31, Slingram EM, VLF) and geomagnetic methods were applied. Based on
these an image of the resistivity distribution in the investigated half-space, of its homogeneity was
created. In the course of geomagnetic survey natural magnetic field of rocks was observed and bodies,
veins having a magnetic susceptibility significantly different from their vicinity were looked for.
Studying the data it was found that based on the changes in different parameters along the profile the
investigated valley sections are not homogeneous from geoelectric viewpoint. The granite body can be
divided into (several times 10-100 m wide) parts, blocks of different average resistivity and average
variability. This image is similar to that, which can be seen in the well logs of boreholes. VLF and
EM-31 measurements provided the most important information from the viewpoint of the set
objective. Based on the resistivity image – within this primarily on that of the VLF ensuring the
deepest penetration – units, blocks were separated. Wide fracture zones detected in trenches could be
correlated with geoelectric anomalies. According to magnetic measurements the loose or fracture
zones do not give any anomaly. Only the biggest loose zones could be traced bellow 10-60 m thick
The high frequency magnetotelluric survey and 3D seismic velocity tomography, performed to study
the homogeneity of the host rock in the 0–300 m depth range are the scope of another lecture.