ABSTRACT: Hydraulic fracturing is the best method to improve production in deep shale, and the complex natural fracture affects the fracture propagation. In this study, numerous true tri-axial experiments were conducted on hydrostone to study the effect of the natural fracture complexity. The position and the relative angle of natural fracture were mainly studied. The acoustic emission was used to detect the fracture propagation and the pumping curve was analyzed. The results showed that the three-dimensional fracture morphologies can be classified into five ways: arresting, bypassing, crossing, vertical extension and slipping. Symmetrical natural fracture around the wellbore helped to increase the fracture height extension. When natural fractures distributed asymmetrically, the hydraulic fracture showed obvious single wing crack. The hydraulic fracture tended to activate the natural fracture and changed direction along the natural fracture with a small approaching angle. The pumping curve rising in the propagation stage showed the fracture arresting the natural fracture. More complex fracture network corresponded to more fluctuated curve. The results give a deep understanding of the effect of the natural fracture complexity in 3D and help to predict the fracture intersection morphology in the field.

ABSTRACT: In this study, bump potential in the vicinity of a fault zone associated with a longwall mining excavation is evaluated with respect to the different stages of shear strain in the normal fault through the use of a numerical modeling technique. The influence of stored and dissipated energy resulting from fault deformation and timing of longwall excavation on the potential for a bump adjacent to the fault was also investigated by the numerical model. While faults are commonly simulated as a single planar or nonplanar interface for a safety or stability analysis in underground mining excavation, the real 3D structure of a fault is often very complex with different branches that activate at different times. Furthermore, these branches are zones of nonzero thickness where material continuously undergoes damage even during interseismic periods. The FLAC3D model successfully demonstrates and creates the 3D fault zone as a normal type structure in the entire thickness of the model. The normal fault structure and its displacement result in the induced deformation in the model. Several panels of a longwall excavation are virtually placed and excavated adjacent to the fault. The characteristics of stored and dissipated energy associated with the panel excavations are examined and observed at the different stages of shear strain in the fault to evaluate bump potential. Depending on the shear strain in the fault, the energy characteristics adjacent to the longwall panels present different degrees of bump potential, which is not possible to capture by conventional fault simulation using an interface.

ABSTRACT: Self-propped fracture is an important flow channel in the hydraulic fracturing process of shale. In this study, we have evaluated the flow capacity in the self-propped fracture and revealed its flow characteristics. On the basis of considering the plasticity characteristics, the elastic-plastic deformation model of self-propped fracture is established. Then, through the deformation model solution, the contact state of the fracture surface is analyzed, the deformation and stress of the fracture are calculated, and the model of fracture flow channel under different closure stresses is established. Finally, the three-dimensional incompressible Navier-Stokes equation is used to describe the flow, calculate the fracture conductivity, and analyze the flow characteristics. Compared with the traditional elastic model, the established model has a more accurate contact area of fracture under high closure stress. The size and shape of the contact area have an important influence on flow velocity, flow path and flow pressure difference. The “strip” contact is easier to form “dominant channels” than the “point” contact. In this study, a numerical simulation method is established to predict the self-propped fracture conductivity and reveal the flow characteristics. The results can provide a new approach for the study of self-propped fracture conductivity in deep shale reservoir.

ABSTRACT: Block caves and other mass mines are the future of global copper supply due to the depletion of surface mines with high production rates and the reduced rates of discovery of easily accessible copper resources. Many current and proposed caving mines are hosted within geological environments that have produced discontinuous fault systems in massive to moderately jointed but complexly veined rock masses. Rock mass characterization in such veined rock masses is challenging and requires attention to vein and open discontinuity data from both a data collection and interpretation standpoint. Depending on composition, abundance, and persistence, veins can work to weaken the rock mass to varying degrees. Although difficulties related to the characterization of veined rock masses are stated in the literature and solutions have been proposed, practitioners continue to struggle with the geomechanical significance of massive to moderately jointed, veined rock masses. In addition, the success of attempts to apply design methods that were developed originally for non-veined rock masses to veined rock masses is questionable. Assuming the rock mass can be characterized to gain both an appreciation of the veining and open discontinuities, a design engineer is left to decide how to incorporate this characterization into various design methods. While both veins and open joints need to be taken into consideration for excavation and cave design, there are limited guidelines on how to use the data reliably and effectively. In this article, both engineering design methods and caving assessment methods are discussed, and suggestions are provided to outline when and how to incorporate veins and joints into design.

ABSTRACT The aim of this paper is the proposition of a new approach and a specific workflow to characterize and model the natural fractures using as case study the Cambro-Ordovician reservoir analog in the Mouydir basin, Algeria. The fracture analysis is based on two approaches. The first approach is the analysis of global fractures that affect the area of study. The fracture map is generated from the combination of curvature attribute, illumination attribute, geological maps, satellite images, and digital elevation models. The outcomes are the determination of fracture sets, fractures length's distributions, correlation coefficients, power law coefficients, and fractal dimensions. The second approach is the analysis of the fractures affecting the basement and the different Cambro-Ordovician units, by using the same input data and determining the same parameters and outputs. The 3D deterministic fracture models for each unit are built to show the fractures distribution in space and their superposition helps to determine their origin, their relationship, their kinematics, which illustrate the impact of the basement's fractures on the sedimentary cover.

ABSTRACT: This paper shows how to integrate several disciplines to characterize and model one of the most important tight oil fractured reservoirs in Algeria to explain its productivity. Seismic data, seismic attributes, well correlation, core, and borehole data are integrated to understand the type and origin of fractures, their spatial distribution, and their impact on the reservoir parameters. The Discrete Fracture Network (DFN) models are generated to display the fault and the fracture sets and show the different mechanical units and compare their relationship with the reservoir units. DFN model at field scale show a lateral variability due to the impact of the Hercynian Unconformity (HU) on the different Hamra Quartzite’ units. The overall DFN shows that the upper unit of Hamra Quartzite appears intensely fractured and fracture sets appear connected and can contribute significantly to the productivity of this reservoir. The existence of the upper unit if the Hercynian Unconformity does not erode it would be the best indicator of the sweet spot in the Hassi Guettar field.

ABSTRACT: Discrete Fracture Network (DFN) modeling is an effective tool to estimate the distribution of rock block sizes. However, for stochastically generated DFN, the clustering of rock blocks is often difficult to emulate, and the locations of high fracture intensity areas are often randomized or arbitrary leading to not-so-realistic DFN models and potentially inaccurate estimates of block sizes. This paper follows a geostatistical approach to model the spatial variability of fracture intensities. This approach is demonstrated through a basement excavation project in Sydney and provides a comparison case study of two DFN modelling methodologies using (1) Geostatistical interpolation of lineal (1-D) fracture intensity (P 10 ) using Ordinary Kriging (OK) and, (2) Random Distribution (RD) of P 10 . Four validation criteria were established to calibrate and provide quality assurance for the DFN models. The results of DFN validation indicate the OK model was able to return better DFN validation measures in P 10 , P 21 and visual sections. For P 32 validation measure, both the OK and RD methodologies are comparable. The Block Size Distribution (BSD) curves generated using the OK model have significant differences when compared to the RD model. This is especially true when assessing the BSD curves for each individual excavation face.

ABSTRACT: An approach for constructing a multi-scale parametric Rock-Physics model of effective elastic properties of fractured reservoir rock is presented. The approach provides a link between three scales - core, log and seismic that correspond to conventional works with reservoir rocks in geophysical prospecting practice. At each scale the Rock-Physics model incorporates parameters specifying rock microstructure at the respective scale including shape, orientation, volume concentration and connectivity of voids (pores, fractures, and cavities). The modeling follows the principle “from small to larger heterogeneities” and incorporates elastic anisotropy of any type caused by preferential orientation of fractures. Evidence for fractures at the core and log scales are provided by different techniques commonly used for the analysis including electronic and optical microscopy, and computer tomography for the core scale and results of FMI and UBI for log scale. Fractures at the seismic scales are inserted in the Rock-Physics model based on the results provided by the geomechanical modeling. At each scale a probabilistic solution for the model parameters is obtained.

ABSTRACT: A new method is introduced for modeling fluid-driven fracture propagation by coupling explicit fracture propagation in MPM with a lower-dimensional diffusion flow model along the fracture. This new method was tested with multiple numerical experiments of fracture propagation in layered materials with interfaces. Observations that are reported by laboratory and field observations (fracture arresting, fracture curving, and reduced fracture growth) in stacked layered media, were obtained in the multiple numerical tests analyzed in this paper.

ABSTRACT: Shear fracture geometry including dilation, roughness, and anisotropy exerts a primary control on fracture permeability. Predictive models of the geometry of shear fracture formation are few, resulting in limited understanding of the permeability of subsurface shear fractures as a function of stress conditions, rock type and other factors. We conduct experiments with triaxial direct shear methods at subsurface stresses that provide new insights into intrinsic mechanisms controlling fracture roughness and the significant anisotropy of fracture aperture distributions. Observed fracture systems contain multiple segments at a variety of scales that initiate as shear fractures and which are out-of-plane with the imposed direct-shear plane and interconnect via tensile fractures. We use a combination of theory and finite-discrete element models to explore the origins of this behavior and the significance for fracture permeability. We find that the simulations are consistent with a theory that predicts that multiple segment fracture systems (en échelon structures) are favored (in terms of stress to cause failure) relative to simple planar fractures. These en échelon structures create permeability anisotropy as well as imparting roughness consistent with field and laboratory observations.

ABSTRACT: Pressure interference data between fractured wells in tight formations during hydraulic fracturing (fracing) has been shown to reflect the geometry of propagating fractures. Interpretation of such data can be used to estimate fracture parameters such as azimuth, length, width and height of the fracture. This pressure interference is ascribed to the poroelastic impact of the propagating fracture's stress shadow. Numerical modeling of this observation using fully-coupled geomechanical simulators has been shown to capture field observations. Numerical modeling, however, can be time consuming and not feasible for application to on-the-fly solutions during the frac treatment. There is a need for fast simple models or simulation tools that can provide a very quick, near real-time interpretations of the position and geometry of the propagating fractures. In this work, we present a new analytical model that provides a quick method of fracture diagnostics during a frac treatment.

ABSTRACT: Source ground vibration in fault blocks of granular fault with respect to stick-slips is important to understand the dynamic behavior of sheared granular fault system. Here, a two-dimensional implementation of the combined finite-discrete element method (FDEM), which merges the finite element method (FEM) and the discrete element method (DEM), is used to explicitly simulate a sheared granular fault system. In the FDEM model, each plate and each particle is represented by a discrete element, which allows for the tracking of their motion and interactions with neighboring objects, and each discrete element is further discretized into finite elements to capture its deformation to external forces. We focus on investigating the stick-slip induced plate vibrations at places adjacent to the gouge in terms of velocities in directions parallel (x-direction) and perpendicular (y-direction) to the shear direction. The simulation shows that the x- and y-velocities on the upper and lower plates have both positive and negative values. The positive x-velocities with large magnitude are mainly seen on the upper plate and the negative x-velocities with large magnitude mostly occurred on the lower plate; whereas the y-velocities show the opposite characteristics. The x-velocities of both the upper and lower plates are highly concentrated around the half of the shear velocity, while the y-velocities of both plates are stabilized around zero. The average vibrations in the y-direction is about 68% of that in the x-direction, which means the places on plates nearby the gouge vibrate in x- and y-direction in a similar pace, but with different magnitudes. The simulations may help reveal the complexities of earthquake source dynamics.

ABSTRACT: Fracture surface features and proppant are the two essential factors controlling hydraulic fracture conductivity. In this work, we measured the effect of fracture surface and proppant parameters on fracture conductivity based on duplicated rough fracture surfaces. To simulate the real hydraulic fracture, large number of artificially split rock samples were collected from shale outcrops. Rough fracture surface was scanned and carving technique was utilized to duplicate rough fracture surfaces on flat rock samples. After that, unpropped and propped fracture conductivity was tested. Experimental results indicated that shale samples perpendicular to bedding exhibited higher strength and rougher fracture surface, resulting in higher unpropped fracture conductivity than samples parallel to bedding. There was not a strong positive relationship between displacement and unpropped conductivity. Partial-monolayer proppant placement could bring high conductivity under low closure stress, while multilayer proppant placement sustained enough conductivity under high closure stress. Optimal proppant parameters were changed with fracture surface roughness and closure stress. Taking initial roughness of fracture surface into account, experimental results can allow us to understand the collation of rough fracture surface and proppant, and help field engineers to make rational choices for the selection of proppant parameters in shale gas fracturing.

ABSTRACT: We present a new fault injection model for induced seismicity in the context of geothermal systems. The characteristic and novelty of our model is that it is both (1) fast and thus allows a probabilistic assessment and (2) purely physic-based in the sense that none unconstrained engineering approach to mimic processes or parameters distributions are used.

ABSTRACT: Subsurface fluid injection can disturb the effective stress regime by elevating pore pressure and potentially reactivate faults and fractures. Laboratory studies indicate that fracture rheology and permeability in such reactivation events are linked to the roughness of the fracture surfaces. We construct discrete element method (DEM) models to explore the influence of fracture surface roughness on the shear strength, slip stability, and permeability evolution during such slip events. For each simulation, a pair of analog rock coupons (3D bonded quartz-particle analogs) representing a mated fracture are sheared under a velocity-stepping scheme. The roughness of the fracture is defined in terms of asperity height and asperity wavelength. Results show that (1) samples with larger asperity heights (rougher), when sheared, exhibit a higher peak strength which quickly devolves to a residual strength after a threshold shear displacement; (2) these rougher samples also exhibit greater slip stability due to a high degree of asperity wear and resultant production of wear products; (3) long-term suppression of permeability is observed with rougher fractures, which is plausibly due to the removal of asperities and redistribution of wear products, which locally reduces porosity in the dilating fracture. This study provides insights into the understanding of the mechanisms of frictional and rheological evolution of rough fractures anticipated during reactivation events.

ABSTRACT: The Yangbajing geothermal field occurred in Yangbajing Cenozoic basin in Lhasa-Gangdese Magmatic Belt between Yalu-Brahmaputra River Suture Zone and Bangong Nujiang Suture Zone, controlled by strike-slip and detachment fault system south the Nyainqentanglha Mountain.Formation of hydrothermal system in the geothermal field has direct relationship with secondary faults of the fault system south the Nyainqentanglha acting as a perfect geothermal channel, in trending of northeast, northwest and north northeast in chronological order respectively. The hydrothermal activities display strongly in intersection areas of the several groups of faults, and have joined in geothermal circulation since about late period of Middle Pleistocene - early period of Late Pleistocene.There are two types of geothermal reservoir in the Yangbajing geothermal field, such as shallow basement fracture and Quarternary porous reservoir, and deep-seated hyperthermia structural basement fracture reservoir.

ABSTRACT: Off-fault fracturing that occurs out of the main fault is observed at all scales, from laboratory to plate boundaries. Understanding these co-seismic fracturing phenomena in the laboratory under controlled conditions can provide insight on the u dynamic stress and fault conditions of the earthquakes. Studies regarding the relative amount of energy consumed by fracturing (E G ) is inconclusive, and the disagreement about the relative size of E G is related to the difficulty in assessing the fracture surface area. We conducted a rotary shear experiment under X-ray micro-CT, which allowed not only the measurement of macroscopic stresses but also the imaging of the newly formed fractures inside the sample. With the careful analysis of the micro-CT images, we quantitatively assessed the fracture surface area. Then, we used Griffith theory of brittle fracture to estimate E G , which accounted for only 0.15–0.43% of the total energy consumption during slipping. The E G we estimated may imply the lower bound of the actual fracture energy, because it did not include the micro-fractures below the resolution of the micro-CT or the energy consumed by non-elastic deformation. Friction work, which is calculated from the macroscopic shear stress and angular slipping distance, consumed most (>80%) of the total energy. Less than 18% energy may have radiated in form of stress waves.

ABSTRACT: Using examples from Cretaceous Frontier Formation, we document quartz volumes deposited during and after opening in fractures formed at shallow (<2 km), intermediate (>4 km) and deep (>6 km) conditions. Inconspicuous rinds >0.02 mm thick predominate at shallow depths. At intermediate depths, rinds and porosity are interspersed with narrow ~0.04-mm wide spanning bridges that thicken with time. At >6 km, spanning bridges are thicker and more pervasive (>50%), but retained porosity is extensive. Quartz volumes are governed by l, which reflects temperature, thermal exposure, opening rate, and rock composition. Fluid inclusion assemblages trapped during fracture growth record porosity creation and persistence at depths of >4-6 km for >5 to 40 m.y. Results show that open fractures persist at high temperatures despite gradual cement accumulation. Where amounts of quartz deposited during propagation is high, fracture clustering is more pronounced, suggesting interaction between cement accumulation and fracture affects spatial arrangement.

ABSTRACT: The goal of this study is to explore potential stress shadow patterns associated with hydraulic fracture swarms such as those encountered in the recent stimulated rock volume (SRV) drill-through projects in the Eagle Ford and Midland Basin, TX. Whereas stress shadow modeling typically follows simulation of simplified hydraulic fracture planes, the modeling presented here focuses on morphological complexity similar to what has been observed in the subsurface core and image log data. Conceptual hydraulic fracture swarms are created by merging iterative variants of modeled hydraulic fractures. Then a proprietary boundary element modeling code based on the triangular elastic dislocation method is used to simulate stress perturbation effects due to pressurization of the fracture network. Internal pressure in the fractures is modeled as a crack-normal dislocation with a smooth aperture decay function. The wide range of stress shadow results complement subsurface data (e.g., distributed acoustic sensing - DAS, time-lapse microseismic, and downhole pressure gauges, etc.) that reveal pressure interference during stimulation and depletion to be non-uniform, dynamic, and difficult to predict using deterministic approaches.

ABSTRACT: Rocks typically consist of continuous intact rock and discontinuities. Fracture intensity, one of the geometrical properties, plays an important role in mechanical behaviors and hydraulic conductivity of discontinuous rock masses. However, discontinuous rock masses exhibit spatial variabilities in the fracture intensity due to the variation of geometric constituents. The accurate fracture intensity estimation is needed. The two-dimensional fracture intensity (P 21 ) measurement is an efficient approach to predict the fracture intensity in a rock mass. This study proposes analytical solutions for the uncertainty in a fracture intensity measurement. The Poisson distribution is used for modeling the uncertainty of a P 21 measurement. To validate the analytical solutions, this paper uses a commercial discrete fracture network (DFN) code, along with an in-house DFN code, to simulate P 21 measurement. The results showed that the uncertainty of a P 21 measurement can be expressed by the size of sampling window, fracture diameter, and fracture intensity. An illustrative example is provided to illustrate the proposed method.