The understanding of coupled thermo-hydro-mechanical behaviour of fault zones or in naturally fractured reservoirs is essential both for fundamental and applied sciences and in particular for the safety assessment of radioactive waste disposal facilities. The overall objective of the CHENILLE project is to better understand the physical processes resulting from thermal and hydraulic loading in a small fault zone in a highly consolidated shale formation. Consequently, a thermally controlled in-situ fluid injection experiment is intended to be performed on a strike-slip fault zone outcropping at the Tournemire/France Underground Research Laboratory (URL). A heating system has been installed around the injection area to enable a precise and controlled incremental increase of the thermal load. Different monitoring systems are designed to measure the seismic and aseismic deformation induced either by thermal and/or by hydraulic loading. The seismic monitoring system is composed of Acoustic Emission (AE) and broadband seismic sensors enabling monitoring of seismic fracturing processes down to sub-decimetre scale as well as slow deformation processes. Furthermore, we are about to install an injection chamber allowing to perform a controlled gaz injection test. The injection borehole will also be partly equipped with fiber optics in order to measure temperature in a distributed manner in the borehole. Time-lapse active seismic surveys are scheduled for before and after the experiment to image the structural network but also to detect the appearance of new structures triggered from the hydro-thermal pressurization of the fault as well as eventual changes in the velocity field.
Numerous multidisciplinary studies have improved the understanding of hydro-mechanical behaviour of the geological subsurface, with immanent implications towards mitigating anthropogenic seismic hazard in the context of developing geothermal reservoirs (Kwiatek et al., 2019; Amann et al., 2018), or for the safe storage of radioactive waste (Armand et al., 2017) and carbon dioxide (Guglielmi et al., 2021), as well as for hydrocarbon extraction. However, due to the technical challenges involved, coupling these efforts to include the influence of temperature changes has so far been set aside despite its fundamental relevance.
In addition to its societal relevance, this topic is of paramount importance
to improve the understanding of the coupled Thermo-Hydro-Mechanical (THM)
behaviour of geological reservoirs (both for storage and production) and
along active plate-bounding fault zones. One method used by the Earth
science community to understand THM processes of reservoirs is to build
large-scale field approaches monitored through geophysical methods. However,
these non-intrusive geophysical technologies generally provide images
related to the spatial distribution of physical properties which rarely
allow to decouple the processes involved directly : geophysical monitoring
in general is a key topic for many purposes related to geosequestration
(Davis et al., 2019). In this respect, laboratory experiments can
be conducted at various pressure-temperature conditions to identify
mechanisms during deformation processes. Unfortunately, as of today, these
approaches are still limited to short time scales (between 1 d to a
month), very small spatial scale (e.g. sample sizes of the order of
centimeter). Moreover, these processes are rarely studied using a joint
laboratory and field research approach (e.g. in-situ injection experiment). The
here presented research program aims at closing this gap. We here intend to
address key questions related to the impact of high temperatures of the rock
mass(up to 150
The principal objectives of CHENILLE are:
Generate a profound understanding of the spatiotemporal
evolution of temperature and pore-pressure fields caused by induced heating
of an intact, damaged and intensely deformed shale formation crossed by a
tectonic fault (“fault-heating experiment”). Understand the coupled hydro-thermo-mechanical processes
occurring in and around a meso-scale fault in a semi-controlled environment,
bridging the knowledge gap between laboratory and field. Develop novel geophysical monitoring techniques and strategies
in different scales suitable for highly attenuating material (clay rich
rocks) in an URL and under natural conditions, i.e. with a fault zone in
place and under high-temperature conditions.
This experiment will allow us to test monitoring methods that have proven
their effectiveness in other geological environments and on laboratory
scale, and to illustrate the influence of temperature on natural geological
structures under representative in-situ conditions.
The experiment consists of stimulating a fault zone by injecting gas in its core and simultaneously heating or cooling the injection zone. Optimized geophysical monitoring methods have been installed for different scales and resolutions around the stimulation zone. These methods had been either tried or tested in other types of materials (combination of pneumatically driven seismic impact sources for application along the tunnel surfaces and single well imaging (SWI) tools for borehole seismic exploration, passive acoustic emission (AE) monitoring methods based on the use of acoustic/high-frequency sensors), or based on more recent methods (optical fibres for temperature measurement), but never in our type of environment combining thermal and hydraulic stimulation in clay rich rock.
The in-situ scale tackled here bridges the lab scale with its controlled boundary conditions with the field scale along plate-bounding fault zones where processes occur entirely out of control. CHENILLE thus allows to transfer knowledge gained at either end of the scale to derive a better understanding of deformation processes under realistic in-situ conditions. Finally, this study will allow to obtain results from controlled experiments to produce new methods or hypotheses to be implemented and investigated in the field. Conversely, observations in the field will potentially produce new hypotheses that will be in turn tested by controlled experiments. The objective of the present paper is to provide a detailed description of the CHENILLE experiment to give an overview of the combination of geophysical measurements that we are intending to perform.
Faults are mechanical discontinuities in the crust where deformation is localized and where elastic energy is released both seismically and aseismically (Ben-Zion and Sammis, 2003; Faulkner et al., 2008; Kaduri et al., 2017; Romanet et al., 2018). This denomination includes features with a size that may vary from meter scale (the so-called “meso-scale”) up to hundreds of kilometers (field scale along plate bounding tectonic faults). Due to their complex structure and mechanical behavior, their study requires a wide range of combined tools such as geomechanics, structural geology, geophysics, geodesy or geochemistry. Similarly, controlled experiments in the laboratory on specimens of cm-scale can be performed, allowing a better understanding of the physical mechanisms responsible for the failure, from the preparatory phase to the post rupture phase (Bonnelye et al., 2017a; Geng et al., 2017; Schuster et al., 2021). These approaches are very complementary and allow studying relevant processes and parameters in detail and with the option to repeat experiments and control the boundary conditions. Nevertheless, the question remains if up-scaling to meso- and field scale facilitates better understandingof the underlying processes controlling reactivation of faults in geological reservoirs and rupture initiation of large earthquakes along fault zones.
Over the past decades, numerous hydraulic stimulation experiments have been carried out in mesoscale in mines and URLs around the world (e.g., Kwiatek et al., 2019; Hadermann and Heer, 1996; Kneafsey et al., n.d.; Zang et al., 2017; Baisch, 2002). It is commonly accepted that these experiments are very important towards understanding processes governing induced seismicity for industrial applications such as geothermal energy, shale gas exploitation, subsurface gas storage and nuclear waste disposal. However, for technical reasons, most URLs are at shallow depth and the range of pressure and temperatures investigated in the frame of the experiments are low. Therefore, they are not representative of the conditions encountered in natural seismogenic zones, nor for industrial applications such as energy storage (geothermal), or conditions that can be encountered in the frame of nuclear waste disposal. For that reason, and even though laboratory experiments show that temperature plays a major role in the mechanical behaviour in clay rich materials, this parameter is rarely considered in mechanical models of fault zones. In spite of this, some in-situ experiments have been developed, allowing the monitoring of fluid potential overpressure under limited temperature conditions in unfractured zones (De Bruyn and Labat, 2002; Bernier and Neerdael, 1996). These experiments show that an increase of temperature, even if small, can lead to an increase of several bars in pore pressure, which could be of major importance for a critically stressed fault or in the frame of nuclear waste disposal safety.
Table presenting previous heating experiments performed in URLs dedicated to research on clay rich materials in western Europe.
Consequently, the project CHENILLE aims at studying the THM processes in the frame of a French German collaboration (the experiment has been delayed due to the pandemic situation and will take place in winter 2023).
Overview of the Tournemire URL (after Lefèvre et al., 2016).
The field test is located in the Tournemire URL (S France) owned by IRSN (French Institute for Nuclear Safety). The URL is composed of a 1.9 km long former old railway tunnel (excavated 135 years ago) and six other galleries (excavated between 1996 and 2008). The Tournemire URL is a key infrastructure enabling to study and understand the confining properties of shale rocks and the performance of certain components of a nuclear waste repository (engineered barrier systems or EBS) to provide an independent and robust technical assessment of France's future deep geological disposal facility for radioactive waste.
The old railway tunnel crosses a Lower Jurassic shale formation (Toarcian, 180 My) that has been extensively studied over the past 30 years in terms of physical properties, mechanical behaviour, chemical composition etc. (Boisson et al., 2001; Bonin, 1998; Matray et al., 2007; Dick et al., 2016; Moreno et al., 2018; Niandou et al., 1997; Bonnelye et al., 2017b; Masri et al., 2014). The Tournemire shale rock has the particularity to be imbedded with faults and fractures of different sizes that extend from the surface (sedimentary cover) to the basement (crystalline). Two minor faults (F1 and F2, see Fig. 1) outcropping in the URL have been extensively studied (Boisson et al., 2001; Bonin, 1998; Matray et al., 2007; Dick et al., 2016; Moreno et al., 2018) and thus offer a unique opportunity to understand under various conditions a well-documented and well-preserved 3D clay-rich fault rock.
The architecture and geometry of these fault zones were determined by
small-scale mapping of the galleries' walls and floor and through the
geological and geophysical logging of boreholes and core samples (Fig. 1).
F1 and F2 have a similar dip and dip-direction varying spatially between
170 to 010
This THM field test is located at the depth level of the F2 fault at 10 m distance from the northern wall of the Gallery East_03. A view of the borehole setup is shown in Fig. 2, which will host the different monitoring systems that will be installed. The borehole layout consists of four types of boreholes: (i) one injection borehole (BH1) that crosses the entire fault centered and equipped with a hydro-mechanical probe centered along the fault's core; (ii) four heating boreholes (BH2 to BH5) dedicated to host electrical heaters, (iii) 4 boreholes (BH6 to BH9) dedicated to the geophysical monitoring of seismic and aseismic fracturing processes and (iv) 17 boreholes for active seismic imaging (BH10 to BH26).
Simplified map of the CHENILLE borehole layout. The geometry of the fault zones was determined from borehole logging and geological mapping from the galleries' floor, walls, and roof. The three main components of the fault zones are represented in this sketch: the fault core (red), the damage zone (grey), and the undisturbed protolith (white).
The target area is located at the level of the F2 fault 10 m from the
northern wall of the Gallery East_03. The borehole layout of
the test (Fig. 2) consists of (i) one (BH1) 30 m long injection borehole
(borehole core diameter: 96–63 mm) located in the Gallery
East_03 following a N72
Before being equipped, detailed petrophysical studies as well as
Scheme of the injection chamber that is about to be installed in BH1.
The injection borehole (BH1) is about to be equipped with chamber delimited by two temperature resistant packers (Fig. 3). The upper part of the borehole (between the chamber and the borehole mouth) will be cemented and equipped with an injection tube and fiber optics for distributed temperature measurements (see Sect. 3.4). The chamber is designed to perform pressure tests during the heating phase.
The heating boreholes will each house a 4 m long stainless-steel heating
canister. The heaters will contain spiral resistors enabling a heating power
of 4000 W at 230 V. The heating canisters were designed to generate,
incrementally, temperatures in the order of 150
The temperature evolution along the four heating boreholes will be monitored using distributed temperature sensing (DTS). This is probably the most frequently applied fibre-optic sensing method, which has also been used with increasing success for geophysical applications in boreholes during the past 30 years (Henninges and Masoudi, 2020).
Using DTS technology, an optical fiber acts as the sensing element. As there
are no electronic parts required, fiber-optic sensors have some advantages
over conventional sensors, including ruggedness, flexibility and ease of
deployment, high temperature tolerance, and immunity to electromagnetic
interference. With a proper DTS installation and system, a measurement
accuracy of
Under harsh environmental conditions such as high temperature and excessive
pressure, the fiber attenuation can be altered through a process known as
hydrogen ingression or darkening, which can lead to a rapid degradation of
the measurement data (e.g. (Smithpeter et al.,
1999). The coating material protects the optical fiber against
adverse chemical influences and primarily determines the temperature
tolerance. Metal-coated fibers have the highest temperature tolerances, but
usually also exhibit high attenuation, especially at low temperatures
(Reinsch and Henninges, 2010), which limits
their use to very short sensing ranges. A coating comprised of an outer
high-temperature tolerant polyimide layer and an additional carbon layer,
which offers some protection against hydrogen ingression, is referred to
hermetic coating. During a field test in a high-temperature geothermal well
in Iceland, Reinsch et al. (2013) have
successfully deployed hermetic coating fibers at temperatures up to 230
View into the gallery east-03 from the main tunnel showing the impact seismic source mounted at a fork lifter. In the foreground the tail end of a seismic receiver is located at the left tunnel wall.
Raw data example of seismograms recorded by the radial component in BH14. Arrows mark later P-wave arrival times which indicates the location of fault core and possible fractured zones within the damaged zone along in gallery East-03.
For deployment in boreholes, sensor cables must be designed to protect the optical fibers against mechanical stress, and to withstand the environmental conditions like temperature, pressure and corrosive fluids. For installation in the four heating boreholes, a sensor cable design with an inner stainless-steel tubing hosting the optical fibers and an outer steel wire jacket for additional mechanical strength and protection has been chosen. The fiber-optic sensor cables will be installed in a loop configuration, allowing to interrogate the optical fiber from both sides. They will be fixed to the outside of the heater elements, extending as long as possible towards the tip of the borehole, using suitable devices to protect the cables from mechanical damage during installation.
For these reasons, two kinds of arrays will be deployed: one allowing for the temperature distributed monitoring (Raman) (Henninges et al., 2005b), and one for deformation measurement (Brillouin). The first array will provide accurate information on the temperature evolution along the heating boreholes. The DTS data will allow to derive petrophysical properties such as thermal conductivity and heat capacity of the surrounding host rocks. The second array will be an essential tool to understand the influence of thermal loading on a fault zone.
Monitoring seismic deformation and resulting high-frequency signals in clay
rich rocks is challenging due to strong attenuation and anisotropy present
in the rock mass (e.g. Plenkers
et al., 2022; Kwiatek et al., 2011; Bohnhoff et al., 2009). During
the Faults&Fluids (De Barros et
al., 2016) project at the Tournemire URL, accelerometers sensitive
in a frequency range between 2 Hz and 4 kHz were installed in boreholes
within the same fault zone. These allowed for the detection of 32 events
with magnitudes ranging between
Attempts to monitor slow deformation processes using broadband seismic sensors and tilt sensors have been performed beforehand in hydraulic fracturing experiments performed in volcanic and metamorphic rocks. The tilt of sensors was observed at Äspö/Sweden (Zang et al., 2017) and at Grimsel Test Site/Switzerland (Gischig et al., 2018). Within the STIMTEC/Germany project (Boese et al., 2022), low-frequency signals responding to injection pressure have been observed during some hydraulic fracturing or refracturing stages. The interpretation of such recorded signals is still challenging due to their overall scarcity and ambiguity of interpretation. Within the CHENILLE project, the six-component broadband sensor ASIR A-SiA-ULN-G4.5-GS-70 will be installed in four boreholes located very close to injection intervals. The sensor consists of a 3-component 4.5 Hz geophone and a 3-component ultra-low-noise optical accelerometer covering the range from 0.01 to 100 Hz. For time synchronization purposes, one component will be concurrently recorded using AE acquisition system operating at higher sampling rates.
Controlled seismic experiments using different types of impact sources (Richter et al., 2018) will be conducted to image the velocity field and the fault network in and around the rock formation targeted by the CHENILLE project. The results from the active seismic survey will complement the surface-based geological mapping and analysis of the drill cores and will be intergrated towards a background geological model as baseline for time-lapse investigations planned later on. A baseline seismic survey along the main tunnel and gallery east-03 was performed in June 2021 to explore the structures and to map the seismic rock conditions before the hydraulic stimulation experiment is taken place (see Fig. 2, short boreholes along the galleries). The survey length is about 100 m with a source point interval of 0.5 m. Two different pneumatically driven impact sources (Fig. 4) were used in a comparative test in 1 m interval each.To prove the repeatability of the sources and to increase the signal-noise ratio by vertical stacking three and five shots were fired per source point. In total, 821 shots were recorded by 15 three-component geophone receivers (28 Hz) of about 6 to 7 m interval installed in 2 m deep boreholes (BH10 to BH26 in Fig. 2). The selection of the source type for the baseline Tournemire survey is based on a previous seismic experiment conducted in the Mont Terri URL (Esefelder et al., 2021; Wawerzinek et al., 2022).
Material properties used in the numerical simulations.
Evolution of the temperature along BH1 during the thermal stimulation.
Evolution of the temperature along the heating boreholes
(BH6 to BH9, panels
Raw data recordings are characterized by P-, S- waves and surface waves in the frequency range of 30 to 1000 Hz in an offset range of up to about 5 m. Beyond this range the maximum signal frequencies decrease quickly below 500 Hz. Figure 5 depicts a raw data example of the radial component of the receiver in borehole BH 14 (Fig. 2). The fault core zone (see Fig. 5) is characterized by delayed arrival times of the direct waves as well as back-propagated surface waves (arrows in Fig. 5). These delayed arrival times can be observed also at other locations along gallery East-03 which indicates the complexity of the fault zone.
It is intended to apply 3D-imaging techniques combining tomographic and migration inversion techniques to gain images for P- and S- waves in the area of about 50 m around the tunnel with resolution of about 0.5 m (Ciese et al., 2005; Lüth et al., 2008). Recent experiments performed with the same system in the Mont Terri URL (Switzerland) revealed a transmissivity of seismic waves for frequencies of up to 1000 Hz. The width of seismic imaging in the frame of the CHENILLE project will be completed by tomographic inversion of attenuation values (Krauß et al., 2014). The seismic attenuation is a valuable parameter to understand the rocks mechanical behaviour during hydraulic stimulation experiments.
In order to image eventual variations of the velocity field at and near the stimulation a repeat (time lapse) seismic survey will be conducted after the hydraulic stimulation of the fault core zone using the same survey layout to enable a 4D-seismic data processing. Additionally, a deeper horizontal borehole will be integrated into the survey layout applying recoverable 3C-geophone chain levels and a newly developed single-well imaging tool. In this way the resolution of the seismic imaging methods shall be increased to gain a deeper insight into fault core zone structure and possible changes due to the hydraulic stimulation.
In order to optimize the CHENILLE stimulation experiment, we performed a numerical simulation of the heating procedure for the thermal stimulation using a finite element commercial software COMSOL Multiphysics 5.6.
For these simulations, we considered a volume of rock of
The thermal stimulation is then simulated as a time-dependent thermal
diffusion process as described by the following heat transfer equation:
Figures 7 and 8 present the result of the numerical simulations during the thermal stimulation. These results clearly show that the four heaters set-up will allow the thermal stimulation of the fault core and damage zone up to 150
In-situ experiments have proven to be important tools for testing monitoring
devices under real conditions, but also for a better apprehension of all the
key parameters that need to be accounted for in the frame of nuclear waste
disposal such as evolution of hydraulic properties, mechanical behaviour and
geochemical reactions under realistic conditions. The CHENILLE project aims
at participating at the efforts that were developed by the community over
the past decades towards the understanding of clay rich material with
investigating the influence of relatively high temperatures (up to
150
Temperature map on a plan in the
The outcome of the CHENILLE experiment is expected to have key implications and a broad range of applications: the safety in geo-engineering (nuclear waste storage, energy storage, geothermal energy), but also for fundamental understanding of coupled THM behaviour of fault zones. The monitoring protocol developed uses state-of-the-art technologies that will be tested in clay rocks for the very first time under challenging (temperature) conditions as proposed here. By combining the results obtained in the frame of this study and experimental techniques allowing for precise description of deformation mechanisms from micro scale to centimetre scale, with the systematic use of acoustic data at all scales (and its integration) will ensure the generation and comprehension of a unique dataset.
The code is provided in open access through the a Zenodo repository and can be found at:
Not applicable in our case as the paper just presents the combination of different techniques, no data is shown.
AB and PD are the PIs of the project and wrote the manuscript. RG and SL provided the set up for the active seismic part, JH provided the set up for fiber optic part, GK provided the set up for the passive seismic part, DJ made the numerical simulation. All the authors reviewed the final manuscript.
The contact author has declared that none of the authors has any competing interests.
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the special issue “European Geosciences Union General Assembly 2022, EGU Division Energy, Resources & Environment (ERE)”. It is a result of the EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022.
We would like to acknowledge Jonathan Dreux and Bruno Combes for their patience and enthusiasm during the drilling operations. We thank Amberg Technologies for funding the development of the new seismic impact source and Helmholtz Innovation Lab “3D-Underground Seismic” for technical and financial support for the active seismic experiment. This document is a deliverable of the European Joint Programme on Radioactive Waste Management (EURAD). EURAD has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 847593.
This research has been supported by the Horizon 2020 (EURAD (grant no. 847593)).The article processing charges for this open-access publication were covered by the Helmholtz Centre Potsdam – GFZ German Research Centre for Geosciences.
This paper was edited by Christopher Juhlin and reviewed by Roberto Emanuele Rizzo, Sarah Weihmann, and one anonymous referee.