Content uploaded by Ferenc Deák
Author content
All content in this area was uploaded by Ferenc Deák on Jul 04, 2017
Content may be subject to copyright.
Rock Mechanics and Rock Engineering: From the Past to the Future – Ulusay et al. (Eds)
© 2016 Taylor & Francis Group, London, ISBN 978-1-138-03265-1
Examination of a granitic host rock behaviour around underground radwaste
repository chambers based on acoustic emission datasets
F. Deák
Mecsekérc Environmental Protection Limited Company, Budapest University ofTechnology and Economics,
Department of Engineering Geology and Geotechnics, Hungary
I. Sz˝ucs
Faculty of Engineering and Information Technology, Department of Environmental Engineering, LADINI
Engineering Co. Ltd., University of Pécs, Hungary
ABSTRACT: The program for the final disposal of low and intermediate level radioactive waste was established
by Paks Nuclear Power Plant, Hungary. The Central Nuclear Financial Fund and the Public Limited Company
for Radioactive Waste Management (PURAM) has been established to coordinate organizations and activities for
all tasks in connection with nuclear waste treatment. Based on a detailed geological screening a granite complex
has been chosen as potential host rock beside Bátaapáti village in the south-western part of Hungary. The goal
of this paper is to present shortly the used seismoacoustic systems in underground facilities and discussing more
detailed on the long term acoustic emission (AE) acquisition system installed in the repository area. The used AE
measurement method allowed a detailed short- and long-term, non-destructive supervision. With this technique
is possible to allocate zones of stress-related fracturing around the underground spaces and to correlate the
modeling results of the Excavation Damaged Zones (EDZs) around the Bátaapáti radioactive waste repository.
1 INTRODUCTION
There is no doubt about the worldwide renaissance in
nuclear industry based on its numerous strong eco-
nomical and environmental competitive advantages in
spite of some recent technical crashes generated by
devastating human or natural influences. While the
technical background of the nuclear industry exceeded
a half century the conscious nuclear environmental
protection has only some decades of experience. Con-
sequently the main questions in connection of siting
and construction of radwaste disposals are still in the
rank of pioneering challenges worldwide. In the lack
of generally adaptable methodology radwaste manage-
ment remains to be solved in national frames all over
the world, so this task in Hungary has an enhanced
priority too.
According to the managerial and research results in
all stages of the two decades process of the harmo-
nization of the natural environment and underground
objects confining radwaste disposal the role of inves-
tigation methods in tunneling design and construction
covers the:
– challenges of technical management background-
ing the special aspects of construction engineering
in connection with radwaste repository design and
safety,
– application results of some investigation methods
put into domestic practice for decision making
support.
The construction of the repository began in 2005
with the excavation of two access inclined tunnels.
During the excavation of underground facility 21 m2,
25 m2and 36 m2cross-sections were used. In Septem-
ber of 2011, two repository chambers, with 96 m2
cross-section (including final lining) were completed
(Fig. 1).
The construction process continued with excava-
tion of the third and fourth chambers of Bátaapáti
repository (intended for final disposal of low and inter-
mediate level radioactive waste) carried out between
2011–2015, (within the frame of the different con-
struction phases of the National Radioactive Waste
Repository) were continuously monitored by four seis-
moacoustic systems equipped with 24 accelerometers
surrounding the target area (Figs.2&9).
Geotechnical results of tunnel face documentations
were statistically analyzed and after every project
phase the results were published in summary reports.
During this period the results of the hypocenter dis-
tribution and the trend analysis of the stress changes
in the underground mining area supplied a substan-
tial amount of information for the decision makers
in charge of mining safety about the basic fracturing
1243
Figure 1. Schematicview of the National Radioactive Waste
Repository at Bátaapáti, magnified the investigation area
with just two repository chambers which after building began
a long period for AE monitoring without underground activity
until the next chambers excavation.
processes in the monitored host rock formation. This
monitoring phase also included the installation and
inspection of the systems in close connection with
other geotechnical investigations to be performed in
the frame of the investigation programme of the
National Radioactive Waste Repository in Bátaapáti
and financed by the Hungarian Agency for Radioactive
Waste Management (PURAM).
The investigation methods aiming decision support
of Bátaapáti L/ILW repository construction covered
the following areas:
– seismoacoustic monitoring the fracturing processes
in connection with the host rock stress changes,
– geotechnical investigations of the host rock
formations.
AE monitoring provides the localization and exam-
ination the mechanics of damage in the Excavation
Damaged Zones (EDZs) around the excavatedcaverns,
allowing it to be mapped spatially and temporally with
high resolution.
2 GEOLOGICAL CONDITIONS
The Palaeozoic granite is an intruded and displaced
batholithic body. The Mecsekalja-belt is located near
this area which is an extended tectonic zone. The
geotechnical background of this area is very com-
plex. At shallow depths, the tunnels crossed through
an absolutely decomposed, weathered rock mass. The
granite is crisscrossed randomly by trachyandezite &
aplitic veins indicating a lower degree of weathering,
followed by tectonically determined tension joints.
Also, at deeper regions, the granite is intersected by
fault zones, interlaced with clay and composed brec-
cia zones the width of which could be 5–10 m. The
rock mass is stochastically fractured with limonite,
hematite, chlorite, carbonate fracture infillings. Some
apparent trends can be discovered in the inclined
exploratory tunnels, e.g. at the veins the fracturing has
a NE-SW strike with middle and flat dip angles, while
igneous structural set is represented at NE-SW strike
and 65–75◦dip angle.
Geotechnically, 4 main rock groups can be dis-
tinguished: monzogranite, monzonite, veins, hybrid
rocks. The percentage of these rocks in the reposi-
tory area test and access tunnels appeared as follows:
monzonite-granite 62%, monzonite 1%, veins 10% &
hybrid rocks 27%. The four repository chambers con-
structed crossed through a rock mass which contained
mainly monzogranite with aplitic veins and scarce
monzonite inclusions (Deák et al. 2014).
3 SEISMOACOUSTIC MONITORING
3.1 Activities carried out so far in the facility
The introduction of the regular installation of seismoa-
coustic monitoring systems established the observa-
tion conditions of host rock formation:
– initially in the inclined shafts and connecting gal-
leries approaching to the target area;
– later inside of the surrounded chamber field, in the
target area;
– by means of imaging acoustic emissions;
– accompanied by increasing and releasing rock
stress changes and;
– in connection with the precursor phenomena of
fracturing and rock movements.
The seismoacoustic monitoring systems have been
installed in the connecting galleries of the parallel
inclined shafts approaching to the target area estab-
lished and fulfilled their geotechnical and mining
safety role.
The application of the method proved to be a rich
information source (of the order of 104 acoustic emis-
sions during this period) for monitoring observation
and imaging of the rock stress change processes gen-
erated in the interaction of the natural and constructed
environment.
The increasing distance of the excavation frontal
face from the monitoring positions during tunneling
process is generally demonstrated with a decreas-
ing trend of the detected acoustic emission’s energy
content.
The aging process of the already constructed mining
areas is generally connected with a decreasing trend of
the detected acoustic emissions both in frequency and
intensity.
The monitoring results of the continuous seis-
moacoustic imaging in time and space (hypocenter
distribution, spectral content) may have an important
1244
Figure 2. The general screen view of the specific program,
developed for the preliminary visualization of AE database.
There is a possibility to choose some registered points and
check the source location on the schematic planar repository
area. Beside the point appearance are notified the arrival time,
coordinates of AE points, waveforms etc. (Sz ˝ucs 2012).
Figure 3. AE source locations around the repository area
(4179 hypocenters).
role in supplying substantial amount of information
(∼Tbyte/month) for the decision makers:
– in identification of types of the different phenom-
ena connected with the interaction of the natural
and the constructed environment,
– in the trend analysis of the host rock stress changes,
– in the underground mining area,
– in the empirical qualification and feedback of the
engineering operations.
The AE monitoring results were logged and visu-
alized previously with a self developed software and
after the processing of the raw data is possible the
detailed examinations (Fig. 2).
3.2 AE database used during the examination
We were used for our research a dataset collected
between 2011.03.01–2014.07.31. During this period
were detected and localized 4179 hypocenters (Fig. 3).
We cumulatively studied the above mentioned
period, but finally we have been concentrated on
time phase when was no excavation in the facility.
In that period between 2011.09.11–2014.05.29, 2532
hypocenters were registered.
Due to the AE investigation we were observed the
following:
– the process of the stress redistribution is good
followable with the used AE monitoring system;
– the most intensive rock mechanical occurrences
(mainly the stress redistribution) are visible around
the repository chambers and at the cavern’s
intersections;
– the AE intensity decreasing in time, but this phe-
nomena is not monotonic, some decreased intensity
stages are followed by increased intensity;
– at the preliminary phase seemed that the AE points
distribution is stochastic. The experts who car-
ried out the measurements didn’t observed any
deterministic connection with any rock mechan-
ical process based on hypocenter density, signal
frequency (Bakai et al. 2012);
– later we found very good correlations between the
AE locations and spatial extension of EDZs, as well
as the relationship between AE effect and mapped
discontinuities is discoverable.
4 SPATIAL EXTENSION OF EDZSAND
CORRELATION WITH AE SOURCE
LOCATIONS
4.1 Theoretical background based on the
laboratory measurements
The effects of stress induced brittle fracturing on
the progressive degradation of intact rock (and rock
mass) strength is an important concern in assessing
the degree of damage that occurs around underground
excavations (Eberhardt et al. 1998).
During the laboratory research the most reliable
methods are the crack volume strain method (with
strain gauges) and the acoustic emission method (with
AE transducers) for the determination of crack clo-
sure (CC), crack initiation (CI) and crack damage (CD)
thresholds (Diederich & Martin 2010).
In our case the damage thresholds were determined
following the methods outlined by Eberhardt’s &
Martin’s instructions (Eberhardt et al. 1998).
The summarized results showed, that the monzo-
granite samples behaves as brittle material during the
UCS laboratory tests (between the CD and failure
stage rarely could be detected “creeping”, or ductile
indications) (Deák et al. 2013).
Because of the sub-granular structure of the mon-
zonite, it deforms ductile before the failure. This stage
is somewhat larger than in the case of monzogranite.
The results are summarized in Table 1.
4.2 The failure and post failure behaviour of rock
masses
Very low- or no-confinement induces purely brittle
failure. In this case the complete loss of strength,
resulting spalling and slabbing on the excavatioan
walls. The mentioned failure mechanism is dominated
1245
Table 1. Monzogranite and monzonite thresholds.
Threshold CC CI CD UCS
Monzogranite
Mean (MPa) 24.45 52.14 145.60 165.11
% (of UCS) 14.81 31.58 88.18
Monzonite
Mean (MPa) 24.22 53.30 134.25 163.72
% (of UCS) 14.81 31.58 88.18
Figure 4. Schematic draw on the rock mass behaviour as a
function of confinement comparison with an intact rock sam-
ple’s result (in Bátaapáti, around the two built chambers the
monzogranite maximum Ei=65 GPa) and rock mass mod-
ulus (maximum Erm =16 GPa, the used value for modeling
Erm =14 GPa), the proportional rock mass stiffness is based
on the GSI ∼45 (after Coulson 2009, based on data Kovács
et al. 2012).
by dilatational extension fracturing, which is similar to
axial splitting in laboratory UCS test, in this case AE
events are mainly tensile in nature (Cai et al. 1998).
From rock mechanical studies has been observed, that
the onset of microseismicity occurs close to the equiv-
alent laboratory point of CI (Martin 1997, Diederich
et al. 2002, Coulson 2009) (Fig. 4). Based on field
observations and micromechanical studies, the crack
initiation level is “universal” as mentioned Diederich
(2000) at all rock scales and relative intensive to
confinement.
With the below approach is possible to describe the
near field of excavations (e.g. EDZs) and the same time
the far-field rock mass behaviour (e.g. shear zones,
fracture coalescence).
4.3 Rock mechanical characterization of EDZ zones
In summary, the tunnel excavation process affects the
rock mass producing zones of damage. These are the
Excavation Damaged Zones (EDZs). Moving away
from the excavation, the EDZs can be distinguished
as the Highly Damaged Zone (HDZ), the inner EDZ
(EDZi) and the outer EDZ (EDZo). Beyond the EDZ is
a stress and/or strain influence zone that involves only
elastic change, the Excavation Influence Zone (EIZ)
(Perras et al. 2012).
TheUCS and Braziliantensile tests results likeinput
parameters and triaxial tests determine the true intact
rock strength (σci) and the intact rock material constant
miare needed for EDZs numerical modeling (Deák
et al. 2013).
Diederich (2007) has developed a method to
describe brittle behaviour with the generalized Hoek-
Brown peak and residual parameters. He referred to
this method as the Damage Initiation and Spalling
Limit (DISL). This method requires CI, UCS and
tensile strength as input parameters.
To get the adequate results, we were using the
prediction of spalling approach (Diederich & Martin
2010).
Independently of the applied excavation modes –
spalling occurs immediately and it develops notches,
wedges and immediately fracturing parallel to the
tunnel axis.
If the wedges are not developed by the fractures
which intersect the rock mass, then the characteristic
overbreaks are caused by the immediate spalling and
blasting processes.
At low and intermediate in situ stress levels, the
stress magnitude may reach the level, that cause local-
ized spalling and notch formation depending on the
rock strength. At greater depths the brittle failure may
involve the whole boundary of the opening.
The development of brittle failure appears parallel
to the excavation boundary due to the increasing of the
load (Ghazvinian et al. 2011).
The existing fractures start to extend and form wing
cracks parallel to the excavation surface are formed in
the absence of confining stress and due to the increase
in the stress magnitude around the excavation surface.
The cracks propagate until they coalesce and start to
interact. The interaction of the cracks causes slabbing
and/or spalling around the boundary.
4.4 Numerical modeling of EDZs
During the modeling work we were using the soft-
ware package developed by Rocscience Inc. For the 2D
approach Phase2, for the detailed AE and EDZs visu-
alization Examine3D and for checking RS3programs
were used.
The above presented approaches to identify the
EDZs, have a main disadvantage, these methods can
be used only in the case of circular shape tunnels (Deák
et al. 2013).
Hence we carried out calibration modeling in
Phase2with circular cross-section.
Circular tunnels were modeled using characteristic
rock mechanical input parameters which of the stor-
age chambers environment. In the output results the
1246
Figure 5. Extension of the EDZs around the repository
chambers (threshold isosurfaces: HDZ with red, EDZiwith
violet and EDZowith gray) (Deák et al. 2013).
Table 2. The distribution of EDZs based on empirical and
circular tunnel modeling determinations.
σmax =65.3MPa,σmax/CI =1.31
Monzogranite FOS
EDZ-type r/a σ3/σ1(around the chambers)
HDZ 1.10 0.084 1.13
EDZi1.25 0.187 1.41
EDZo1.32 0.290 1.80
limits of EDZs were identified. The obtained val-
ues and modeling experience were applied during
the modeling of repository chambers. Using the real
geometries it was possible to define the limits and spa-
tial extent of EDZs, which theoretically occurs as a
result of tunneling stress redistribution.
Using the characteristic horizontal/vertical stress
ratio (K =1.35) the expected results were obtained.
Around the excavation floor and in the roof the exca-
vation damaged zone appeared in a larger dimension.
Our results are shown in Figure 5 andTable 2.
Due to the control modeling we were used the real
residual rock mechanical parameters (strain softening
failure mechanism) and the prescribed sequential tun-
neling mode with the required rock support system
(Fig. 6). The obtained results showed that the spatial
extent increased considerable at the crown and less
at the invert (originally for the EDZo: invert =3.7 m,
crown =3.4 m; with rock support and without residual
parameters: invert =2.3 m, crown =1.5 m).
4.5 Representation of AE events with the main
modeling results
Our present investigation is limited only to the spatial
expansion and distribution of the AE events.
If we try to superficially analyze the distribution of
the detected hypocenters, then it becomes clear that
the events are attached to the vicinity of the excavated
areas, and these point’s density increases significant
Figure 6. Control modeling of the spatial extent of EDZs
with the used rock support.
Figure 7. Spatial distribution of AE hypocenters with EDZo
isosurface (no excavation period).
Figure 8. Spatial distribution of AE hypocenters with EDZo
isosurface (with all hypocenters).
around the excavated chambers (Fig. 3). This charac-
ter can be discovered during the tunneling work break
even,whenthestress redistribution and fracturingarise
caused by no technological influences.
Using the factor of safety (or strength factor) gives a
better view of AE points concentrated around the exca-
vations. The most AE events can be localizable with
stress redistributions, caused by the excavation pro-
cess (Figs. 7 & 8). The remaining hypocenters should
be in contact with structural discontinuities or weak-
ness zones and those are arises due to the macro cracks
coalescence or shearing of the existent discontinuities.
1247
Figure 9. Actual schematic view of the Bátaapáti radioac-
tive waste repository area with the new built chambers and
the extended seismoacoustic monitoring system.
4.6 Further consideration in the AE research from
Bátaapáti radwaste repository
Simultaneously with the expansion of the repository
with the new chambers and access tunnel, the
seismoacoustic monitoring system has been ex-
tended with two more monitoring devices (Szeiz-11
– TOVD access tunnel, Szeiz-12 – TESZV access
tunnel) (Fig. 9).
The presented investigation will be followed by
examination of AE waveforms, the space-time distri-
bution of the hypocenters and the source mechanism.
5 CONCLUSIONS
In a previous work we highlighted the importance of
EDZs (Deák et al. 2013). In the Bátaapáti project the
dimensions of the EDZs are identical in both 2D and
3D numerical models. During the new investigations
we observed a good correlation between the EDZs spa-
tial extent and AE events distribution. When is viewed
parallel to the axis of the repository chambers there
is a very clear relationship between AE source loca-
tions and induced stresses. A significant part of the AE
sources are in relationship with the EDZs and with the
stress redistribution caused by the excavation, other
part may be contact with effect of generation of new
micro-cracks, due to the coalescence appears macro-
fractures in weakness zones of the rock mass. Sliding
and shearing can generate stress disturbance.
REFERENCES
Bakai, J., Kardos, J. & Sz ˝ucs, I. 2012. Operation of
seismoacoustic system (Szeiz-08, Szeiz-09, Szeiz-10) (in
Hungarian) – RHK-K-045/12 – PURAM report, 47 p.
Cai, M., Kaiser, P.K. & Mar tin, C.D. 1998. A tensile model for
the Integration of microseismic Events Near Underground
Openings. Pure Appl. Geophys. 153: 67–92.
Coulson, A.L. 2009. Investigation of the pre to post peak
strength state and behaviour of confined rockmasses. PhD
Thesis, Department of Civil Engineering, University of
Toronto. 457 p.
Deák, F., Kovács, L. & Vásárhelyi, B. 2014. Geotechni-
cal rock mass documentation in the Bátaapáti radioac-
tive waste repository. Central European Geology, 57(2):
193–207.
Deák, F., Kovács, L. & Vásárhelyi, B. 2013. Modeling the
Excavation Damaged zones in the Bátaapáti radioactive
waste repository. EUROCK2013, ISRM,Taylor & Francis
Group, London, 603–608.
Diederich, M.S. 2000. Instability of hard rock masses:
The role of tensile damage and relaxation. PhD Thesis,
Department of Civil Engineering, University of Waterloo,
Ontario. 566 p.
Diederich, M.S. 2002. Keynote: Stress Induced Damage
Accumulation and Implications for Hard Rock Engineer-
ing. Proc 4th northAm. Rock Mech. Symp., NARMS ’02,
Mining and Tunneling Innovations and Opportunities,
Toronto, July 8th–10th. 1:3–12.
Diederich, M.S. 2007. Mechanistic interpretation and practi-
cal application of damage and spalling prediction criteria
for deep tunneling. Can. Geotechn. J. 44(9): 1082–1116.
Diederich M.S., Martin C.D. 2010: Measurement of spalling
parameters from laboratory testing. Zhao, Labiouse,
Drudt & Mathier (eds). EUROCK2010, ISRM, Taylor &
Francis Group, London, 323–326.
Eberhardt, E., Stead, D., Stimpson B. & Read, R.S. 1998.
Identifying crack initiation and propagation thresholds in
brittle rock. Canada Geotech J. 35: 222–223.
Ghazvinian, E., Diederich, M. & Archibald, J. 2011. Chal-
lenges related to standardized detection of crack initiation
thresholds for lower-bound or ultra-long-term strength
prediction of rock. 2011 Pan-Am CGS Geotechnical
Conference. 8p.
Kovács, L., Deák, F., Somodi, G., Mészáros, E., Máté, K.,
Jakab, A., Vásárhelyi, B., Geiger, J., Dankó, Gy., Korpai,
F., Mez˝o, Gy., Ván, P., Fülöp, T. & Asszonyi Cs. 2012.
Review and extension of the Geotechnical Integration
Report (in Hungarian). RHK-K-032/12, PURAM report,
2012 December, 312 p.
Martin, C.D. 1993. Seventeenth Canadian geotechnical Col-
loquium: The effects of cohesion loss and stress path on
brittel rocj strength. Can. Geotech. J. 34(5): 698–725.
Sz˝ucs, I. 2012. Visualization of database for seismoacoustic
monitoring system from Bátaapáti (Szeiz-08, Szeiz-09,
Szeiz-10) (in Hungarian). Geotechnical Environmenal
Observation Production and Research Development
report, 2012 December, 18 p.
1248