Search

Fluid saturation plays an important role in rock weakening. However, few researches reveal the controlling factor(s) behind the rock weakening particularly from geochemical perspective. We performed rock failure/fracture reactivation tests on dry and wet samples (saturated with high/low salinity brine). We examined the effect of fluid saturation and salinity on rock mechanical properties through a unique triaxial compression system with combination of micro-seismicity monitoring. Besides, the induced time-dependent micro-seismicity events were located to characterize the fracture orientation together with medical CT imaging. Moreover, we performed geochemical modelling to link the surface energy to rock mechanics from geochemical perspective.Our results show that the peak stress of saturated samples is 11.7% to 15.0% less than unsaturated samples, confirming that fluid saturation can weaken rock. Lowering salinity appears to play a negligible role in the reduction of elastic modulus and peak stress of intact sandstones. We also did not observe the expansion of micro-seismicity cloud along main fracture plane, suggesting that lowering salinity may not trigger micro-fracture generation under confining condition. However, sample saturated with low salinity brine shows a much slower increase rate of differential stress during the plastic deformation to reach yield point during fracture reactivation process compared to the sample saturated with high salinity brine. This work contributes to the existing knowledge of rock weakening by providing quantitative geochemical characterization with implications in water-assisted enhancing hydrocarbon recovery and CO2 geological storage.

In this article we explore the change in elastic symmetry and anisotropy of two geological materials with stress-induced damage. Two independent uniaxial deformation experiments on two layered and natural geomaterials, a shale and a sandstone, support this analysis. Both samples were loaded along their bedding planes at a constant strain rate up to mechanical failure. During deformation, an array of ultrasonic P- and S-wave transducers were employed to monitor the evolution of the ultrasonic velocities up to and beyond sample failure. We report here the impact on elastic symmetry and anisotropy of a uniaxial load applied parallel to the bedding plane of the transversely isotropic shale and sandstone samples. Based on symmetry considerations we analyse whether this load and the resulting stress-induced damage preserve the original transverse isotropy of the rock prior to loading, or lead to orthotropy (orthorhombic symmetry). Our results show that the two transverse isotropic samples retained their transverse isotropy in the initial stages of loading/damage. However for the more anisotropic shale sample the main failure mode was splitting of the bedding planes. In contrast the sandstone sample failed along a shear plane inclined to the bedding plane. In addition, the experimental data are further analysed using continuum damage mechanics to identify the evolution of the general fourth order anisotropic damage tensor during loading. We found that the highest damage variable of the damage tensor was $$D_{11}$$ D 11 for the shale and $$D_{55}$$ D 55 for the sandstone sample. The next highest damage variables for the shale sample were related to the bedding plane ($$x_2$$ x 2 –$$x_3$$ x 3 plane) symmetry: $$D_{23}, D_{32}, D_{44}$$ D 23 , D 32 , D 44 . However the damage variables obtained for the less anisotropic and more brittle sandstone sample were harder to interpret due to the mixed mode of fracturing present.

Understanding the propagation of hydraulic fracture (HF) is essential for effectively stimulating the hydrocarbon production of unconventional reservoirs. Hydraulic fracturing may induce distinct failure modes within the formation, depending on the rheology of the solid and the in-situ stresses. A brittle-to-ductile transition of HF is thus anticipated with increasing depth, although only scarce data are available to support this hypothesis. Here we carry out laboratory hydraulic fracturing experiments in artificial geomaterials exhibiting a wide range of rheology: cubic samples 50x50x50 mm3 in size are subjected to true triaxial stresses with either a low (σv = 6.5 MPa, σH =3 MPa, and σh =1.5MPa), or a higher (15 MPa, 10 MPa, and 5MPa) confinement. The 3D strains induced by hydraulic fracturing are monitored and interpreted; X-ray Computed Tomography (CT) imaging is used to document the HF geometry; and viscoelastic modelling of the tested materials is also conducted to explain the distinct geometry of hydraulic fracture subjected to the stress state. Finally, a correlation between the normalized fracture area (AFN) and the brittleness index (BI) of tested samples is introduced. Our results reveal that: (i)The intermediate stress plays a profound role in hydraulic fracture propagation subjected to the normal faulting regimes (i.e., the transitional intermediate strain observed from brittle to ductile samples); (ii) The orientation angle of hydraulic fracture is highly inclined to the maximum horizontal σH (or vertical σv) stresses in brittle/semi-brittle samples; as BI decreases, the angle inclination is reduced for that of semi-ductile samples, finally reaches to zero (parallel to σH and σv) in ductile sample. (iii) The normalized fracturing area (AFN) decreases as the decrease of BI among different samples under either low or higher confinement. The results of viscoelastic modelling explain the distinct characteristics of hydraulic fracturing induced deformation among the tested samples subjected to true triaxial stress state. This study reveals the importance of understanding the underground brittle-to-ductile behaviour of hydraulic fracture prior to the field implementation.

Brittleness Index (BI) is a critical parameter characterising the deformation regime of geo-materials, covering the range from purely brittle (fractures) to ductile (plastic flow). A variety of BI models have been developed based on rock properties such as mineralogy, elastic parameters, or stress-strain data. However, very few of them are based on the deformation induced by hydro-mechanical interactions emerging in a wide range of underground engineering applications. In this study, we develop a BI model based on the partitioning of the injection energy EI into non-seismic deformation energy Ed associated with hydraulic fracture propagation. To calculate the Ed, we apply a model for temporal fracturing area (Ad) within the penny-shaped fracture; we also correlate the wellbore pressure and the three-dimensional strain induced by hydraulic fracturing of the different types of rock samples subjected to true triaxial stress conditions (TTSC), either σv = 6.5 MPa, σH = 3 MPa, σh = 1.5 MPa or σv = 15 MPa, σH = 10 MPa, σh = 5 MPa for all tests. As a comparison, the BI is also quantified based on the existing models: (i) acoustic measurement from Rickman et al (2008), and (ii) the Mohr-Coulomb’s criteria from Papanastasiou et al (2016). The non-seismic deformation energy Ed ranges between 32.4% and 90.6% of the total injection energy EI, which is slightly higher than reported from field-scale data (15% to 80%), and is comparable to other laboratory-derived data (18% to 94%). The results show that the predictions based on our newly proposed hydro-mechanical energy-based BI model are qualitatively consistent with Papanastasiou et al.’s, but less so with Rickman et al.’s. Our BI model is shown to be stress-dependent and capable of capturing the brittle-to-ductile behaviour within a wide range of rheological samples subjected to hydraulic fracturing. This study demonstrates that our BI model opens a new way for quantifying the brittleness index regarding to realistic propagation scenarios, showing its superior robustness for such underground applications.

Hydraulic fracturing is a coupled multi-physics and scale-dependent process requiring an extensive numerical-laboratory appraisal to assess feasibility in the field. Developing a robust model of hydraulic fracture propagation requires knowledge of the time evolution of the fracture’s geometrical attributes, e.g., width/aperture and length/radius. However, it is inherently challenging to directly measure even the simplest fracture attribute (i.e., radius) within a rock sample subjected to in-situ stresses in the laboratory, let alone in the field. In this study, an analytical model (Rd) is developed based on Poiseuille’s law. Scaling laws and dimensional analysis are used to define propagation regimes; and non-linear hydro-mechanical coupling is accounted for the near-tip region. This model aims to predict the time evolution of radius for a homothetic penny-shaped hydraulic fracture when the fracture opening, and internal pressure gradient are known. Based on the available experimental data from literature, we quantify the growth of the fracture radius using linear elastic fracture growth model (RE); tip asymptotic solutions (RV and RT); semi-analytical solutions (RS); and the model Rd. A comparison of the four analytical models with published experimental data reveal that (i) the asymptotic solutions are limited to linearly elastic and homogeneous materials, i.e., PMMA; (ii) the semi-analytical solutions (RS) is only suitable for late-time propagation (iii) the performance of the linear elastic model (RE) poorly matches the experimental data, especially for unstable propagation situations; (iv) the new Rd model takes advantage of a robust reconstruction of the temporal radius growth of hydraulic fracture problems under realistic stress conditions, and including multiscale propagation regimes, cohesive effects, as well as stable and unstable propagation regimes of geomaterials.

Hydraulic fracturing is a coupled multi-physics and scale-dependent process requiring an extensive numerical-laboratory appraisal prior to a field trial. Developing a robust model of hydraulic fracture propagation requires knowledge of the time evolution of the fracture’s geometrical attributes, e.g., width/aperture and length/radius. However, it is inherently challenging to directly measure even the simplest fracture attribute (i.e., dynamic radius) within the rock sample subjected to in-situ stress conditions. In this study, two analytical models are developed to predict the time evolution of the radius of a penny-shaped hydraulic fracture. In the first model (RE), the incremental fracture propagation of an infinite hydraulic fracture is described using linear elasticity, modified to incorporate a realistic distribution of internal fluid pressure using a Taylor series. The second model (Rd) predicts the fracture opening and internal pressure gradient using Poiseuille’s law and assuming a self-similar propagation. Scaling and dimensional analysis are used to define propagation regimes; non-linear hydro-mechanical coupling is accounted for in the near-tip region (process- or cohesive-zone). Finally, the two analytical models are compared against published experimental data. Our results show that the radius prediction based on linear elasticity (RE) is suitable only for limited conditions, i.e., non-viscous dominated propagation regimes with negligible fluid lag and a stable fracture propagation. On the other hand, the Rd model takes advantage of a fast reconstruction of the dynamic radius of finite hydraulic fracture problems under more realistic conditions, i.e., including multiscale propagation regimes, cohesive effects, stable and unstable propagation in different geomaterials.

Typical ultrasonic laboratory measurements of rock physical properties are conducted with ultrasonic transducers that are relatively large compared to the rock sample. We have determined the impact of the transducer size on the resulting velocity estimations. To improve data interpretation, we explore two complementary avenues: (1) explicitly account for the finite size of the transducers as part of the data interpretation/correction workflow, rather than assuming point sources, and (2) reducing the effective size of the transducers at the hardware design level. Both approaches as well as their combination have been tested successfully on multiple data sets including artificial homogeneous and isotropic samples as well as natural anisotropic rocks such as shales.

Mudstones are often anisotropic, which complicates depth conversion in seismic exploration and monitoring subsurface reservoirs during injection or production. In addition, the physical and mechanical properties of mudstones are highly sensitive to their water content. The elastic anisotropy of mudstones is not well understood because of their complex nature and the lack of laboratory experiments performed on well-preserved samples. Triaxial deformation tests were performed on mudstone core plugs to investigate the impact of water saturation on the elastic anisotropy of the Whitby Mudstone (United Kingdom). The mechanical and physical properties of the Whitby Mudstone were estimated from stress-strain and ultrasonic wave velocity data obtained on core plugs with different water saturations under isotropic and anisotropic stress conditions. The Whitby Mudstone has extremely high intrinsic elastic anisotropy (0.3–0.4) due to its composition and lamination. This elastic anisotropy increases with decreasing water content. There are three competing mechanisms that play a key role in the anisotropy increase due to dehydration such as (1) density contrast in the pore space (i.e., the presence of purely brine or a mixture of brine and air in the pore space), (2) formation of dehydration fractures, and (3) frame stiffening. Increasing the mean effective stress leads to a decrease in Thomsen’s anisotropy parameters [Formula: see text] and [Formula: see text] because of the closure of defects, such as natural and dehydration fractures, and the formation of stress-induced fractures. The relationship between the wavefront anellipticity factor [Formula: see text] and the mean effective stress is nonmonotonic and can be related to the onset of inelastic deformation.

International audience; In the present work, we investigate the mechanical weakening and deformation induced by water on a microporous carbonate rock, the Obourg Chalk from Mons Basin (Belgium), through conventional triaxial tests and injection tests. The injection tests were conducted by waterflooding critically loaded rock samples, initially in dry condition, in a way to minimize the variations in the effective pressures. Furthermore, the samples were instrumented with P-wave piezoelectric transducers to provide active ultrasonic monitoring while injecting. The results show a significant reduction in the mechanical strength of this chalk. Analysis of the mechanical tests and the associated deformation allows us to describe the mechanical behavior as a function of the confining pressure, which draws a brittle-ductile transition spanning from low to high confining pressure. The injection tests, moreover, revealed that the amount of water injected before triggering mechanical instability decreases exponentially with respect to the applied differential stress on the rock sample. The data, therefore, suggest that the failure might be controlled by a mechanical coupling between the water-invaded zone and the dry one. Since water-weakening plays an important role in several fields like oil industry, through secondary and tertiary recovery of hydrocarbons, Enhanced Geothermal Systems (EGS), as well as in the mechanical stability of underground cavities, the outcome of this work is of primary importance in mitigating any kind of problems related to these operations.

Monitoring of fluid movements in the crust is one of the most discussed topics in oil & gas industry as well as in geothermal systems and CO2 storage, but still remains a challenge. The seismic method is one of the most common ways to detect the fluid migration. However, the use of ultrasonic monitoring at the sample scale in laboratory experiments persists as the most effective way to highlight large scale observations in which the boundary conditions are not well constrained.To unravel the fluid effect on P-wave and S-wave velocity, we performed mechanical experiments coupled with ultrasonic monitoring on Obourg chalk from Mons basin (Belgium). Water injection tests under critical loading, imbibition tests and evaporation tests provided a full spectrum of observations of fluid-induced wave alteration in term of propagation time and attenuation.The analysis of these experimental results showed that significant velocity dispersion and attenuation developed through variations in water saturation, and that these processes are linked to the presence of patches of water and air in the pore space.We used the White’s formulation to model the relaxation effects due to spherical pockets of air homogeneously distributed in a water-saturated medium. In this framework, the pressure induced by the passing wave, produces a fluid flow across the water-air boundary with consequent energy loss.This model reproduces both qualitatively and quantitatively the experimental results observed on the water injection tests. Indeed, it is shown that the progressive water saturation or desaturation of this chalk, generates a shift of the critical frequency (from the undrained relaxed towards unrelaxed regimes) which at some point matches the resonance frequency of the piezoelectric transducers used in the experimental setup (0.5 MHz). This phenomenon allowed us to get a continuous recording of the relaxation processes induced by saturation variations.The outcomes of this work can significantly improve the actual knowledge on coupled effects of waves and fluids which is a crucial aspect of fluid monitoring in the context of reservoir evaluation and production.

Large-scale geological carbon storage in the North Sea will involve injecting CO2 in a super-critical state into deep saline aquifers. As CO2 is injected, stress changes due to fluid pressure and temperature changes may result in deformation and potentially may result in microseismicity if existing faults are reactivated. Microseismicity may provide additional information on in situ stress conditions and the progression of the CO2 plume. To assess failure criteria in the reservoir, we must know the mechanical response of reservoir sandstone due to injection of super-critical CO2 (scCO2) under reservoir stress conditions. This includes fracturing processes at the microstructural scale as well as at the macroscopic scale by reactivation of existing fractures. We conducted three triaxial tests on a sandstone analogue to deep North Sea reservoir rock: once pressurized, samples were saturated with either scCO2, brine, or a brine-scCO2 mixture. Each sample was then deformed axially to induce a through-going fracture, which was then reactivated—first by increasing the pore pressure under anisotropic stress conditions, then by axial reloading with a constant pore pressure. Throughout testing, we monitored acoustic emissions (AE) and measured ultrasonic P-wave velocities. AE were located within the sample, and their relative magnitudes were calculated, as well as their source mechanisms for a subset of high signal-noise ratio events. We observed: • Higher strength and stiffness for the sample saturated only with scCO2 than for samples containing brine, and a brine-scCO2 mixture. • No clear difference in fracture strength for samples with different fluid saturations. • Higher energy events released from fracturing of the stronger and stiffer scCO2-saturated rock. • For all samples, the AE response during fracture reactivation was more energetic and had a higher frequency content at higher effective stresses (by axial loading) than at lower effective stresses (by pore pressure increase). We found fluid saturation to have only a slight influence on the mechanical and failure properties. We observed no significant weakening with scCO2 saturation that might otherwise cause undesired inelastic deformation within the reservoir.

Summary Distributed temperature (DTS) and distributed acoustic (DAS) fibre optic sensing have been permanently deployed for downhole monitoring at the In-Situ Laboratory (ISL) in Western Australia. This site, primarily developed for field-scale CO2 controlled-release experiments, was designed as a research facility for remote and 24/7 monitoring using fibre optic sensing technology with a minimum footprint and low maintenance requirements. DTS, DAS combined with pressure gauges are instrumenting an injection well, a monitoring well and a shallow groundwater well. The benefit of integrating discrete pressure measurements with distributed temperature and acoustic sensing was demonstrated over the lifetime of the site. This paper presents such benefits through examples and case studies. DTS enabled continuous recording from the well completion (monitoring of mud circulation and cement hydration) to CO2 injection operations (detection of underground CO2 motion), while DAS allowed for an accurate spatial-temporal location of expected or unexpected events, such as air-lift operations

Summary Applications of fibre optic sensing technology to the energy resource sector are now very common. Yet, they often rely on an analysis of an independent measurement such as DTS or DAS alone. In this work, we present the case study of a water injection and artificial lift operations monitored simultaneously and continuously using surface and downhole pressure gauges, DTS and DAS in a well doublet: injection and monitor at the CSIRO In-Situ Laboratory (ISL) research facility. The results shows that discrete sensors recording at key locations combined with distributed measurement enable further understanding of the water injection and artificial lift operations.

The aim of the current study, IGCCS (2017–2020), is to evaluate the feasibility of micro-seismic (MS) monitoring of CO2 injection into representative storage candidates in the North Sea, based on broad and quantitative characterization of relevant subsurface behavior with respect to geology, geomechanics and seismicity. For this purpose, we first group potential CO2 storage sites in the North Sea into three different depths. Then, advanced triaxial rock mechanical tests are performed together with acoustic emission (AE) acquisition under representative loading for CO2 storage sites in the North Sea and for formations of each depth group, covering shale, mudstone and sandstone cores. Our work focuses particularly on quantifying the effects of injected fluid type and temperature on mechanical behavior and associated MS response of subsurface sediments. The experiment results show that each depth group may behave differently in responses to CO2 injection. Particularly, the occurrence of detectable MS events is expected to increase with depth, as the combined effects of rock stiffness and temperature contrast between the host rock and injected CO2 are increasing. In addition, lithology plays an important role in terms of the MS response, i.e. high AE event rate is observed in sandstones, while aseismicity in shale and mudstone. The test results are then scaled up and applied to advanced coupled flow-geomechanics simulations and a synthetic field-scale MS data study to understand micro-seismicity at fracture, reservoir and regional scales. The numerical simulation of scCO2 injection scenario shows quite different stress-strain changes compared to brine injection, resulting mainly from the thermally-induced behavior. Furthermore, the numerical simulation study via so-called Cohesion Zone Modeling (CZM) approach shows strong potential to improve our understanding of the multiphase-flow-driven fracture propagation. Our synthetic MS data study, focused on slow-earthquake scenario, also suggests that sensors with high sensitivity at low frequency might be necessary for better signal detection and characterization during CO2 injection. This manuscript covers the main findings and insights obtained during the whole study of IGCCS, and refers to relevant publications for more details.

During the production of hydrocarbon reservoirs, EOR operations, storage of CO2 underground or geothermal fluid exchanges at depth, fluid substitution processes can lead to significant changes in rock properties which can be captured from the variations in seismic waves attributes. In the laboratory, fluid substitution processes can be investigated using ultrasonic monitoring. The motivation of our study was to identify the seismic attributes of fluid substitution in reservoir rocks through a direct comparison between the variation in amplitude, velocity, spectral content, energy, and the actual fluid distribution in the rocks. Different arrays of ultrasonic P-wave sensors were used to record at constant time steps the waveforms during fluid substitution experiments. Two different kinds of experiments are presented: (i) water injection experiments in oil-saturated samples under stress in a triaxial setup mimicking EOR operations, (ii) spontaneous water imbibition experiments at room conditions.In the water injection tests on a poorly consolidated sandstone saturated with oil and loaded at high deviatoric stresses, water weakening triggers mechanical instabilities leading to the rock failure. The onset of such instabilities can be followed with ultrasonic monitoring either in the passive mode (acoustic emissions recording) or in the active mode (P wave velocity survey).In the water imbibition experiments, a methodology based on the analytical signal and instantaneous phase was designed to decompose each waveform into discrete wavelets associated with direct or reflected waves. The energy carried by the wavelets is very sensitive to the fluid substitution process: the coda wavelets are impacted as soon as imbibition starts and can be used as a precursor for remote fluid substitution. It is also shown that the amplitude of the first P-wave arrival is impacted by the upward moving fluid front before the P-wave velocity is. Several scenarios are discussed to explain the decoupling between P wave amplitude and velocity variations during fluid substitution processes.

Since the performances of geological reservoirs are continuously changing during the production history, tools capable of characterizing these changes are becoming day by day of relevant importance. A time-lapse monitoring is an important and widely used way to explore the variations induced by the oil depletion. In an enhanced oil recovery scenario, the seismic survey method has been mostly used to monitor the remaining oil fraction with respect to the injected water, however no particular attention has been addressed to the effects generated by the water-rock interaction, which might induce deformation with no stress variation. Indeed, it is well known that water can induce important mechanical weakening in reservoir rocks.For that purpose, we performed injection tests on carbonate rocks in a conventional triaxial cell. An essential characteristic of these tests is the very low injection pressure, in order to minimize changes in the effective stresses and focus specifically on the rock - fluid interaction. The test consists in injecting water from the bottom in a critically loaded sample, initially in a dry state, until deformation is induced by the water-air substitution and failure is reached. While testing, the rock sample is instrumented with either 6 (at UCP, GEC lab) or 16 (at CSIRO, Geomechanics and Geophysics lab) P-wave transducers allowing us to perform an active ultrasonic survey with a narrow time intervals.The described methodology allowed us to monitor how P-wave attributes (amplitude, velocity and frequency), elastic moduli, as well as, permeability and injection rate change while water is flooding the sample increasing the water saturation, and damage is produced by the water – rock interaction. For instance, injecting water into a dry rock sample could produce several patterns of variations in the P-wave velocity, which we ascribed to 1) partial water saturation; 2) water-induced damage with no failure; 3) water-induced failure and, in some cases, 4) total water saturation. More experiments are planned to mimic real EOR operation, like injecting water in an oil-saturated rock sample, with acoustic monitoring as well.The outcome of this study indicates that combining multiple data sets from different sources is an effective tool for monitoring the exploitation of underground resources. This can certainly enhance our understanding of reservoir properties changing over time and target the attention toward the areas of greatest interest.

The physics of gas uptake in coal constrains the feasibility of coal seam gas extraction and carbon sequestration in coal measures. While X-ray tomography using xenon as an X-ray opaque tracer gas has been used in the past to study gas uptake in coal specimens, synchrotron sources enable high-resolution micro-tomography datasets of high quality to be acquired in minutes adding the ability to quantify the dynamics of gas sorption over time in five coals of similar rank on the microscopic scale. In the present work we describe a synchrotron micro-computed-tomography (micro-CT) study of gas uptake in coal using Xe, Kr and CO2 gases. Measurement of gas uptake over time for Xe and Kr in different specimens and of CO2 induced swelling are reported, together with an analysis of gas diffusion profiles and their correspondence to those expected from diffusion models.The rate of penetration was CO2 > Kr > Xe in all cases, though the rates of penetration varied enormously between different coals. In some cases even 2 months was not enough to allow the Xe to equilibrate fully. The gas distribution was also shown to respond rapidly to flushing with CO2.

CO2 injection and storage in geological reservoirs is an attractive prospect for mitigating the anthropogenic production of greenhouse gases and global warming. The technology could lead to mineral precipitation and therefore stable storage over geological time scales. This contribution investigates the evolution of three calcite-rich reservoir rock analogues during injection of and exposure to supercritical CO2 (scCO2), i.e., two limestones (Tuffeau and Savonnieres) and a synthetic calcite-cemented sandstone (CIPS). Three types of exposure protocols have been conducted: (i) scCO2 injection and a four-hour residence time in an initially dry rock; (ii) scCO2 injection and a two-hour residence time in an initially brine-saturated rock; and (iii) scCO2 injection and a four-hour residence time in an initially brine-saturated rock. Two aspects are monitored during these experiments: (i) the evolution of the pore fluid chemical composition; and (ii) the evolution of the rocks’ physical properties (i.e. porosity, permeability, P-wave velocity and electrical resistivity). Additionally, some scCO2 injection and exposure experiments in the brine-saturated rocks have been conducted using X-ray tomographic monitoring. X-ray tomographic monitoring suggests that scCO2 first displaces the water, leading to an average water saturation of about 70–90%. Then, scCO2 dissolves in the pore brine, leading to a homogeneous decrease by about 3% in water saturation of the sample. As a result, the pore brine acidifies even after 2 h of exposure only, which leads to calcite dissolution and a significant increase in the brine’s concentration in calcium cations. For the samples and most exposure experiments, evidence of calcite dissolution is inferred from the measured physical properties. For the brine-saturated Tuffeau limestone and CIPS sandstone, calcite dissolution leads to significant mechanical weakening. For the brine-saturated Savonnieres limestone, the sample subject to two-hour residence time shows evidence of calcite dissolution, whereas the sample after four-hour residence time does not. Calcite re-precipitation could be the cause of this unexpected response after four hours.

Capillary rise experiments (spontaneous imbibition tests) were conducted in the laboratory with ultrasonic and X-ray monitoring on the Sherwood sandstone and the Majella grainstone. The aim was to provide a direct comparison between the variation in seismic attributes (amplitude, velocity, spectral content, and energy) and the actual fluid distribution in the rock. Two pairs of ultrasonic P wave sensors located at different heights on a cylindrical rock specimen recorded every 5 s the waveforms when capillary forces make water rise up into the rock from the bottom in contact with a water tank. Simultaneously, computerized tomography scan images of a vertical cross section were also recorded. Two important results were found. (i) The amplitude of the first P wave arrival is impacted by the upward moving fluid front before the P wave velocity is, while the fluid front has not yet reached the sensors level. In contrast, the P wave velocity decreases when the fluid front reaches the Fresnel clearance zone. The spectral analysis of the waveforms shows that the peak frequency amplitude is continuously decreasing without noticeable frequency shift. (ii) A methodology based on the calculation of the analytical signal and instantaneous phase was designed to decompose each waveform into discrete wavelets associated with direct or reflected waves. The energy carried by the wavelets is very sensitive to the fluid substitution process: the coda wavelets related to reflections on the bottom end face of the specimen are impacted as soon as imbibition starts and can be used as a precursor for the arriving fluid.

We have developed a new numerical algorithm for inversion of ultrasonic data in transversely isotropic media. This algorithm is able to determine from the measured P-wave velocities the orientation of the symmetry axis of a rock sample and the Thomsen’s parameters, only assuming transverse isotropy. The inversion of ultrasonic data acquired on natural and potentially heterogeneous shale samples produced reasonable results. In addition, the algorithm was successfully tested on ultrasonic data acquired on synthetic samples with predefined orientations of the symmetry axis. An additional outcome of the algorithm is a simple approximation of Thomsen’s formulation, which can be effectively used for interpretation of seismic data in transversely isotropic media.

We studied the effect of fluid injection on the mechanical behaviour of the poorly consolidated and layered Sherwood sandstone under varying stresses, with micro-seismic (MS) monitoring. In order to highlight possible weakening effects, water and inert oil have been injected into critically-loaded samples to assess their effect on strength and elastic properties, derive the ultrasonic signature of the saturation front for each fluid, and the potential development of damage. To this end, the specimens were instrumented with 16 ultrasonic P-wave transducers used for both passive and active monitoring during loading and fluid injection. A first set of injection tests in hydrostatic conditions, using either water or inert oil, has been performed on samples subjected to low confining pressure. Water invasion in the pore space induces a significant decrease of the P-wave velocity, whereas oil invasion shows a velocity increase. The velocity decrease associated with water injection is analysed in terms of attenuation mechanisms and corresponding critical frequencies. A second series of injection tests with the same fluids has been performed during creep tests on critically-loaded samples. The development of mechanical instability inducing micro-seismic activity is observed only when water is injected into the sample. The recorded micro-seismic events are spatially and temporally located thanks to the dedicated velocity models accounting for the initially homogeneous sample anisotropy and for the heterogeneous velocity field associated with fluid migration within the sample. The consistency between the relocated clusters of events and the final damage pattern is verified thanks to X-ray computed tomography images of the samples taken post-mortem.

Monitoring microseismic activity provides a window through which to observe reservoir deformation during hydrocarbon and geothermal energy production, or CO2 injection and storage. Specifically, microseismic monitoring may help constrain geomechanical models through an improved understanding of the location and geometry of faults, and the stress conditions local to them. Such techniques can be assessed in the laboratory, where fault geometries and stress conditions are well constrained. We carried out a triaxial test on a sample of Red Wildmoor sandstone, an analogue to a weak North Sea reservoir sandstone. The sample was coupled with an array of piezo-transducers, to measure ultrasonic wave velocities and monitor acoustic emissions (AE) – sample-scale microseismic activity associated with micro-cracking. We calculated the rate of AE, localised the AE events, and inferred their moment tensor from P-wave first motion polarities and amplitudes. We applied a biaxial decomposition to the resulting moment tensors of the high signal-to-noise ratio events, to provide nodal planes, slip vectors, and displacement vectors for each event. These attributes were then used to infer local stress directions and their relative magnitudes. Both the AE fracture mechanisms and the inferred stress conditions correspond to the sample-scale fracturing and applied stresses. This workflow, which considers fracture models relevant to the subsurface, can be applied to large-scale geoengineering applications to obtain fracture mechanisms and in-situ stresses from recorded microseismic data. Inferring microseismic source mechanisms and in situ stresses during triaxial deformation of a North-Sea-analogue sandstone

Mudstones are characterized by their tight matrix and are therefore of interest in various industries, including the petroleum and underground repository industry. These clay-rich rocks often show dynamic elastic anisotropy, which causes significant problems in geophysical interpretations. In addition, pore water has a significant effect on the bulk properties of mudstones. However, the degree of saturation is often not reported in the literature. This study investigates the impact of water saturation on the elastic anisotropy of the Whitby Mudstone. Four core plugs with different water saturations were deformed until failure and tested ultrasonically at effective confining pressure conditions of 25 MPa. P-wave and S-wave velocities were monitored along the symmetry axis, across the core diameter, and at ~49˚ to the horizontal bedding plane to calculate the full elastic tensor and subsequently the Thomsen anisotropy parameters. The degree of saturation highly affects the rock strength and static elastic properties and leads to significant changes in the elastic anisotropy parameters.

Mudstones are of great interest in the petroleum industry, since they can act as a source rocks, reservoirs and seals. Understanding the mechanical behaviour of these clay-rich rocks is crucial for successful hydrocarbon exploration and production. This study examines the impact of water saturation on the mechanical behaviour of the Whitby Mudstone. Triaxial compression tests are performed on mudstone samples with different water saturations at effective confining pressures of 5 MPa, 15 MPa, 25 MPa, and 40 MPa. The results show that the degree of saturation significantly affects the rock strength and static elastic properties (Young’s modulus and Poisson’s ratio). The mechanical properties obtained from partially-saturated mudstones should be treated with care when extrapolating them to in situ conditions.

Fluid-shale interactions appear to significantly affect shale micromechanical properties, which regulate fracture network propagation during multistage hydraulic fracturing in shales and thus the gas production. While published works confirm that fluid-shale interactions can reduce Young's modulus and rock strength compared to dry shale samples, few attentions have been paid to unveil the nature of the physics with a combination of geochemical modelling and disjoining pressure isotherm, and fewer works have envisaged the effect of mineralogy (calcite-rich or quartz-rich rock) on shale weakening with presence of aqueous liquids. We hypothesize that fluid-mineral interactions likely generate electrical double layer force between mineral surfaces which would shift the disjoining pressure from strongly negative (mineral-air-mineral) to less negative or even positive (mineral-fluid-mineral) thus triggering Young's modulus reduction. To test the hypothesis, geochemical modellings including minerals dissolution (calcite, dolomite, quartz, pyrite and illite) and surface complexation were performed together with disjoining pressure isotherm using literature experimental data, accounting for the Young's modulus reduction induced by fluid-shale interactions. The geochemical modelling shows that the amount of mineral dissolution at in-situ condition is negligible, and thus may play a minor role in Young's modulus. Disjoining isotherm shows that both calcite and quartz (plane-plane geometry) give a strong negative disjoining pressure generated by van der Waals force and structural force in the presence of air, whereas the disjoining pressure can be shifted from strongly negative to less negative for calcite and even positive for quartz in presence of fluids due to the electrostatic force, suggesting a Young's modulus reduction due to the weaker adhesion in line with literature experimental data. This work provides an overall conceptual framework, which supports the hydraulic fracturing fluid design and treatment in shale reservoirs from geochemical and disjoining pressure isotherm perspectives.

The technical and economic success of a CO2 geological storage project requires the preservation of the site injectivity and integrity properties over its lifetime. Unlike conventional hydrocarbon gas injection, CO2 injection may imply geochemical reactions between acidified pore fluids and target reservoir formations, leading to modifications of their poromechanical properties. To date, the chemical effects on the host rock mechanical behaviour are not satisfactorily taken into account in sitescale numerical models of CO2 injection, mainly due to a lack of quantitative data. The present experimental work aims at characterizing the evolution of carbonate poromechanical properties induced by acid alteration. Unlike standard experimental approaches, the implemented alteration method induces a homogeneous dissolution pattern, which ensures reliable poromechanical measurements on altered samples. These well-controlled alteration conditions allow a proper interpretation of the test results through the macroscopic continuous approach of poromechanics. Petrophysical, geomechanical, and petroacoustic properties of outcrop carbonate samples have been measured for different levels of alteration to mimic long-term exposure to reactive brine. The obtained experimental data show clear trends of chemically induced mechanical weakening. Nuclear magnetic resonance measurements and microscanner imaging performed before and after alteration have provided complementary insights into the alteration effects at the microscopic scale.

Understanding of microstructural changes in gas shales caused by their thermal maturation is of practical importance for evaluation of extractability of hydrocarbons from these low permeability reservoirs through methods such as sweet spot mapping from surface seismic. Two organic-rich shales (ORS), one with extremely high total organic carbon (TOC) and the other extremely low TOC are chosen for this study. The Upper Jurassic Kimmeridge Shale from and the Upper Cretaceous Mancos Shale contain around 23% and 1% TOC, respectively. Samples are subjected to temperatures in the range of 300 to 510°C. Changes in their mineralogical composition, TOC, weight and microstructure with temperature increase are monitored. The Kimmeridge Shale shows rapid decomposition of the organic matter at the temperatures of 370-390°C. This process is accompanied by fracture development and propagation. The Mancos Shale exhibits shrinkage of the solid organic matter with mobile bitumen expulsion and relocation. No fracture development is directly observed in microtomograms. Further work has to be done to understand whether the ability of shale to develop a fracture network depends on its TOC content, the mineralogical composition of its inorganic matrix or on other parameters.

International audience; We tested on a data set including 12 different rocks the assumption that, during water imbibition experiments, the interaction between direct P waves and reflected P waves on the imbibition front could explain the early P wave amplitude variation occurring before any velocity variation. Our calculations show that (i) the observed distance between the water front and the ultrasonic sensors is always larger than the estimated critical distance below which the first peak amplitude is impacted by the reflected P wave, (ii) the magnitude of the coefficients of reflection is generally very small, and (iii) using simulations of composite waveforms, an impact on the first peak amplitude is expected only in 25% of the tested samples .P wave interaction is definitively a mechanism which can lead to amplitude variations, but it is very likely that such a mechanism is not responsible for the early amplitude variation detected in most of the tested rocks. Moisture diffusion is still a very relevant mechanism to account for our observations on P wave amplitude variations during capillary rise

To assess water-weakening effects in reservoir rocks, previous experimental studies have focused on changes in the failure envelopes derived from mechanical tests conducted on rocks fully saturated either with water or with inert fluids. So far, little attention has been paid to the mechanical behavior during fluid injection under conditions similar to enhanced oil recovery operations. We studied the effect of fluid injection on the mechanical behavior of the weakly consolidated Sherwood sandstone in laboratory experiments. Our specimens were instrumented with 16 ultrasonic P wave transducers for both passive and active acoustic monitoring during loading and fluid injection to record the acoustic signature of fluid migration in the pore space and the development of damage. Calibration triaxial tests were conducted on three samples saturated with air, water, or oil. In a second series of experiments, water and inert oil were injected into samples critically loaded up to 80% or 70% of the dry or oil-saturated compressive strength, respectively, to assess the impact of fluid migration on mechanical strength and elastic properties. The fluids were injected with a low back pressure to minimize effective stress variations during injection. Our observations show that creep takes place with a much higher strain rate for water injection compared to oil injection. The most remarkable difference is that water injection in both dry and oil-saturated samples triggers mechanical instability (macroscopic failure) within half an hour whereas oil injection does not after several hours. The analysis of X-ray computed tomography images of postmortem samples revealed that the mechanical instability was probably linked to loss of cohesion in the water-invaded region.

This paper presents a novel rock mechanics testing method for remotely monitoring, detecting and assessing mechanical weakening and possible sand production in sandstone reservoirs during water flooding or during drilling with water-based muds. Rock mechanics triaxial tests are conducted on sandstone reservoir samples subjected to realistic in situ stresses, pore pressure, and temperature. During these tests, in addition to the conventional stress and strain monitoring techniques, the new approach involves a combination of ultrasonic and micro-seismic monitoring techniques allowing for a remote detection and spatial localisation of the possible mechanical weakening that ultimately leads to sand production. The original testing protocol allows for the substitution of the in situ fluid by a water-based fluid with minimal impact on the in situ effective stress in the rock sample. The weakening effect of the substituting fluid can therefore be discriminated from any mechanical artefact associated with a possible reduction in the effective pressure during the substitution. Thanks to the conventional stress-strain monitoring technique, it has been observed that (i) the strength and static elastic moduli decrease when water saturates the pore space; (ii) the injection of water in a critically loaded sample initially saturated either with air or inert oil induces a drastic increase in the creep rate and the development of a mechanical instability. Thanks to the ultrasonic and micro-seismic monitoring of the fluid substitution process, it has been observed that (iii) the mechanical instability is mainly localised in the water-flooded zone; (iv) this instability corresponds to a loss of cohesion between the sandstone grains; (v) during oil substitution by water, the insability is delayed in comparison with the water injection in the dry sandstone and the loss of cohesion is less pronounced. This loss of grain-to-grain cohesion is expected to lead to sand production in weakly consolidated siliciclastic reservoirs. In addition, the injected fluids significantly affect the ultrasonic P-wave velocity: (vi) increases during oil injection; (vii) decrease during water injection. This helps in detecting the advancement of the fluid front within the rock's pore space.

International audience; The influence of hydrostatic and uniaxial stress states on the porosity and permeability of sandstones has been investigated. The experimental procedure uses a special triaxial cell which allows permeability measurements in the axial and radial directions. The core sleeve is equipped with two pressure samplers placed distant from the ends. They provide mid-length axial permeability measure as opposed to the overall permeability measure, which is based on the flow imposed through the pistons of the triaxial cell. The core sleeve is also equipped to perform flows in two directions transverse to the axis of the sample. Two independent measures of axial and complementary radial permeability are thus obtained. Both Fontainebleau sandstone specimens with a porosity of about 5.8% to 8% and low permeability ranging from 2.5 mD to 30 mD and Bentheimer sandstone with a porosity of 24% and a high permeability of 3D have been tested. The initial axial permeability values obtained by each method are in good agreement for the Fontainebleau sandstone. The Bentheimer sandstone samples present an axial mid-length permeability 1.6 times higher than the overall permeability. A similar discrepancy is also observed in the radial direction, also it relates essentially to the shape of flow lines induced by the radial flow. All the tested samples have shown a higher stress dependency of overall and radial permeability than mid-length permeability. The effect of compaction damage at the pistons/sample and radial ports/sample interfaces is discussed. The relevance of directional permeability measurements during continuous uniaxial compression loadings has been shown on the Bentheimer sandstone until the failure of the sample. We can efficiently measure the influence of brittle failure associated to dilatant regime on the permeability: It tends to increase in the failure propagation direction and to decrease strongly in the transverse direction.

Summary To predict the effects of stress on rock permeability, the authors propose an integrated approach based on an extended rock characterization, an experimental investigation of pressure dependency of directional rock permeabilities and finally a pore-scale simulation of this dependency using equivalent pore network extracted from microtomography analysis. This study has been conducted on two analog reservoir rock types: the high-permeability Bentheimer Sandstone and a dual-porosity bioclastic carbonate, the Estaillades Limestone, having an intermediate permeability. Compression tests have been conducted using a new triaxial cell specially designed to measure directional permeabilities along and transverse to direction of maximum stress application. We measured the pressure dependency of porosity, directional permeabilities, compressibilities, and elastic moduli of the tested samples. We also performed computed microtomography (CMT) imaging of the rock samples, from which we extracted the poral skeletons and the associated characteristics lengths. Then, we calculated the macroscopic transport properties using Pore Network Modeling (PNM) based on the real pore geometry. We included a model of pressure dependence of pore and throat sizes based on pressurized cavity models derived from elasticity theory to simulate the evolution of porosity and permeability with pressure. First, we show that the experimental determination of the evolution of directional permeabilities under hydrostatic and deviatoric loading is feasible. Finally, we show that the PNM coupled with μ-tomography can be a promising tool to forecast the evolution of transport properties under stresses representative of reservoir conditions, at the condition of integrating more advanced pore-scale compaction models.

Several methods are compared for the detection of moving capillary fronts in spontaneous imbibition experiments where water invades dry porous rocks. These methods are: (i) the continuous monitoring of the mass increase during imbibition, (ii) the imaging of the water front motion using X-ray CT scanning, (iii) the use of ultrasonic measurements allowing the detection of velocity, amplitude and spectral content of the propagating elastic waves, and (iv) the combined use of X-ray CT scanning and ultrasonic monitoring. It is shown that the properties of capillary fronts depend on the heterogeneity of the rocks, and that the information derived from each method on the dynamics of capillary motion can be significantly different. One important result from the direct comparison of the moving capillary front position and the P wave attributes is that the wave amplitude is strongly impacted before the capillary front reaches the sensors, in contrast with the velocity change which is concomitant with the fluid front arrival in the sensors plane.

14 pages; International audience; he ferrofluid impregnation technique combined with anisotropy of magnetic susceptibility measurements (AMSff) is one of the ways to analyze the 3-D geometry of the pore space in a rock and indirectly to infer the anisotropy of permeability. We applied this method on different types of rocks (sandstones and carbonates) with a range of different porosity values (10-30%) and permeability (1 mD to 1 D). To get additional information on both the pore aspect ratio and the directional anisotropy we used another technique, measuring the anisotropy of P-waves velocity (APV) in dry and water saturated conditions. Comparing between both methods shows that despite the good agreement in directional data, inferring the true shape of the porosity is not straightforward. Modeling the presence of an elastic anisotropy in the solid matrix for sandstones allows one to get more consistent values for the pore aspect ratio obtained from both APV and AMSff. However for the carbonate rocks, due to an intricate distribution of microstructures, the aspect ratios obtained show significant discrepancies between the two methods. The ferrofluid method is very sensitive to the quality of the impregnation and suffers from a major drawback which is the threshold size of investigation, limited by the size of the magnetite nanoparticles (10 nm) and probably this method doesn't see all the porosity. On the other hand with acoustic methods, the range of porosity investigated is probably larger but several microstructural attributes can contribute to the elastic anisotropy which makes the pore shape effect more difficult to decipher. Therefore, we promote the combined use of both methods in order to get more reliable information on the pore shape in porous media.

Summary We have studied changes of static and dynamic elastic properties of the organic-rich Mancos shale caused by heating to the temperature within the range 380 °C-550°C. An increase of ultrasonic velocities and static Young’s moduli has been observed in the samples that experienced temperatures up to ~400°C. Further heating to temperatures up to 550°C has led to the decrease of elastic moduli. The samples subjected to temperatures above 430°C show an increase in stress sensitivity, which implies microcrack generation. Micro-CT images of small samples have been obtained to inspect the microstructural changes caused by this dry pyrolysis. Comparison of the registered micro-CT images of Mancos shale samples subjected to temperatures up to 470°C shows significant shrinkage of the kerogen that fills pore spaces resulted from bitumen expulsion into the pores.

International audience; The influence of stress paths on the mechanical behavior and coupled permeability evolution of quartz sand packings and soft sandstone have been investigated. The proposed approach is mainly based on the performing drained compression tests keeping constant a stress path parameter K, defined as the ratio of the horizontal to vertical effective stress increments. Macroscopic mechanical data and the stress path dependency of permeability have been measured in the elastic, brittle and compaction regimes. The micro-structural characteristics of unconsolidated materials--coarse rounded grains (glass beads--GB) and angular grains (Durance sand--DS)--and a weakly consolidated sandstone (Otter Sherwood sandstone--OSS) have been quantified in intact and deformed states combining petrophysical core analysis and multi-scale imaging according to the feasibility of sub-sampling of the different materials. Two different types of hydro-mechanical behavior were evidenced for each material. The weakly consolidated OSS presents a mechanical behavior similar to the consolidated rocks with two distinct regimes of deformation; the transition between these two regimes is a function of the stress path. On the other hand, the sand shows a gradual transition regime requiring the use of a curvature criterion to pick yield stresses; this criterion has been validated on the basis of acoustic emissions analysis. The modified Cam-Clay model and the elliptical cap model allow capturing efficiently the yield envelopes. The permeability measurements were performed with intermediate pressure ports to measure accurately the high permeable sand and to avoid end-effects. For OSS, permeability evolutions follow closely the volumetric strain evolutions in both the elastic and plastic regimes and are mainly controlled by the mean pressure before yield. At these critical stresses, the permeability drops drastically under the influence of deviatoric stress. Conversely for DS, the correlation between strain and permeability is not obvious as permeability reduction is pronounced at an early stage of loading.

International audience

This paper proposed an integrated experimental approach to the characterization of CO2 injection chemical impact on rock dynamic poroelastic properties. Acoustic measurements have been performed on intact and homogenously altered carbonate samples. The obtained results show a decrease of both the drained bulk modulus and the shear modulus under the effect of alteration. NMR measurements and microscanner imaging performed before and after alteration have provided an insight into the alteration impacts at the microscopic scale. The substantial reduction of CO2 emissions into the atmosphere implies the injection of huge quantities of CO 2 in appropriate deep geological formations (IEA, 2009). Due to their significant storage capacity, deep saline aquifers range among the most promising options for long-term CO2 storage. The viability of a CO2 geological storage rests on the combination of adequate storage capacity, injectivity and integrity. Those characteristics need to be assessed at the beginning of the project and updated over its lifetime. The injected CO 2 will notably have to be followed through geophysical monitoring to ensure that it remains sealed within the expected area (Lumley, 2010). This paper first presents the poromechanical framework used to interpret seismic data in terms of CO2 presence. This analysis requires the quantification of CO2 injection chemical impact on the host rock acoustic signature. An integrated approach allowing the characterization of these specific effects at the laboratory scale is therefore proposed. Each step of the experimental workflow is detailed. The obtained results show the weakening of the rock frame dynamic elastic properties and illustrate alteration impacts at the microscopic scale.

[1] The influence of stress paths representative of reservoir conditions on the poromechanical behavior and coupled directional permeabilities evolution of a heterogeneous carbonate has been studied. Our experimental methodology is based on performing confined compression tests keeping constant a stress path coefficient K = Δσr/Δσa ratio of the radial and axial stress magnitudes, commonly assumed to be representative of reservoir stress state evolution during production. The experiments are performed in a triaxial cell specially designed to measure the permeability in two orthogonal directions, along and transverse to the direction of maximum stress. The tested rock is a heterogeneous bioclastic carbonate, the Estaillades limestone, with a bimodal porosity, of mean value around 28% and a moderate permeability of mean value 125 mdarcy. Microstructural analyses of initial and deformed samples have been performed combining X-ray tomography and microtomography, scanning electron microscopy (SEM) observations, and mercury injection porosimetry. The microstructural heterogeneity, observable by SEM, is characterized by the arrangement of the micrograins of calcite in either dense or microporous aggregates surrounded by larger pores. The spatial distribution of the two kinds of aggregates is responsible for important density fluctuations throughout the samples, recorded by X-ray tomography, which characterizes the mesoheterogeneity. We show that this mesoheterogeneity is a source of a large directional variability of permeability for a given specimen and also from sample to sample. In addition, the fluctuation of the porosity in the tested set of samples, from 24% to 31%, is an expression of the macroheterogeneity. Macroscopic mechanical data and the stress path dependency of porosity and permeability have been measured in the elastic, brittle, and compaction regimes. No significant effect of the stress path on the evolution of directional permeabilities is observed in the elastic regime. At failure, according to the selected stress path, either a limited or a drastic permeability decrease takes place. From the postmortem observations at different scales, we clearly show the impact of the mesoheterogeneities on the localization of compaction, and we identify the precursor of the shear-enhanced compaction and pore collapse mechanisms (for K ≥ 0.25) as an intense microcracking affecting only the denser aggregates. Applying an effective medium theory adapted to our observations, we propose a porosity scaling to normalize the pressures at failure. It is then found that the normalized critical pressures evolve linearly with the stress path coefficient. Consequently, we put forward a new definition of the yield cap for this type of carbonate, which is parameterized by the stress path coefficient.

The influence of stress paths, representative of reservoir conditions, on the hydromechanical behavior of a moderately heterogeneous carbonate has been investigated. Multiscale structural heterogeneities, common for instance in carbonate rocks, can strongly alter the mechanical response and significantly influence the evolution of flow properties with stress. Using a triaxial cell, the permeability evolutions during compression and the effects of brittle (fracture) and plastic (pore collapse) deformations at yield, were measured. A strong scattering was observed on the mechanical response both in term of compressibility and failure threshold. Using the porosity scaling predicted by an adapted effective medium theory (based on crack growth under Hertzian contact), we have rescaled the critical pressures by the normalized porosity deviation. This procedure reduces efficiently the scattering, revealing in the framework of proportional stress path loading, a linear relation between the critical pressures and the stress path parameter through all the deformation regimes. It leads to a new formulation for the critical state envelope in the 'mean stress, deviatoric stress' diagram. The attractive feature of this new yield envelope formulation relies on the fact that only the two most common different mechanical tests 'Uniaxial Compression' and 'Hydrostatic Compression', are needed to define entirely the yield envelope. Volumic strains and normalized permeabilities are finally mapped in the stresses diagram and correlated.

International audience; In a previous study (Dautriat et al., 2008), we have studied the macroscopic hydro-mechanical behavior of a moderately heterogeneous carbonate reservoir analogue, undergoing triaxial testing along several proportional stress paths, ranging from hydrostatic compression -to-axial compression. Evolutions of permeabilities and compressibilities during loading have been measured and correlated and the yield envelope has been determined. Structural heterogeneities have been shown to strongly affect the initiation of brittle and plastic damages. There is a trend to interpret the macroscopic response in term of micro-mechanisms without actual observation and identification. While post-mortem characterization techniques (HPMI, SEM and XR CT / μ-CT images) inform qualitatively on the damage mechanisms activated at the grain and aggregate scales, a quantitative and continuous micromechanical investigation is needed to follow the history of the deformation and the localization during compression. We have therefore performed in-situ observations during loading at different scales. First, small samples have been deformed by simple compression inside a SEM, in order to identify the physical deformation and damage micro-mechanisms responsible for the evolutions of the transport properties. Then, larger samples have been subjected to axial compression in a hydraulic press and have been observed by optical methods, in order to better understand the complex interactions governing the macroscopic behavior. Observations of the micro-mechanisms during mechanical tests are difficult for geomaterials because the levels of deformation are low (below the percent), but feasible with some care and appropriate recording devices; different regions of small parallelipipedic samples have been imaged using different magnifications to focus either on grain or contact, or to visualize the aggregates. Direct optical observations on bigger cylindrical samples, mechanically loaded on conventional UCS testing machines, have also been performed at two different scales by means of high resolution digital cameras. On one side, the full sample is imaged (≈20μm resolution) in order to characterize the overall response. On the opposite spot side, a centimetric area has been considered (≈3μm resolution), at which scale the composite nature of the rock made of microporous and dense calcite grains is revealed. Those two scales imaging experiments have been combined with efficient Digital Image Correlation (DIC) post-treatments, able to detect very small displacements and evolutions of the microstructures (strain accuracy better than 10-3). The comparisons of SEM images taken stepwise, reveal deformations hardly detectable by conventional observations, such as: opening or closure of pre-existing microcracks, nucleation of new microcracks and relative movements at grain interfaces. Different strain accommodation regimes are also observed in dense and microporous grains, respectively brittle and diffuse. The movies at sample scale show that the heterogeneity of strain is correlated to the local distribution of the aggregates, which confirms the post-mortem observations.