TY - JOUR
AU - Chen,, Mian
AB - Abstract Low porosity and permeability make it extremely difficult to develop shale oil and gas reservoirs. The stimulated reservoir volume is believed to have potential to obtain industry production by multi-stage or simultaneous fracturing in horizontal wells. The formation mechanism of network hydraulic fractures in fractured shale reservoirs remains poorly understood. In this article, a true tri-axial hydraulic fracturing system associated acoustic emission monitor was deployed to simulate hydraulic fracturing on shale outcrops. Results showed that the properties of natural fractures (such as aperture, orientation), compared to the viscosity and displacement of the fracturing fluid, affect the propagation direction of hydraulic fractures more predominantly. Each natural fracture in a natural fracture network can independently affect the hydraulic fracture. Low displacement (below the diffusion ability of a reservoir) fracturing tends to connect pre-existing fractures, while high displacement (surpass the diffusion ability of a reservoir) tends to create new fractures. After the breakdown pressure, an increase in injection rate results in more acoustic emission energy and induces new fractures. These results suggest that step-displacement fracturing technology is a possible mechanism to obtain effective fracture networks. Such an understanding would help to avoid unproductive, or sometimes destructive, costly segments of the hydraulic fracturing treatment design. hydraulic fracture, natural fracture, shale reservoir, horizontal well, fracture network, stimulated reservoir volume 1. Introduction Hydraulic fracturing is commonly used in low permeable reservoir stimulation. The objective of the fracturing process is to obtain a single or multiple cracks with enough flow conductivity, or even more complex fracture networks. Hydraulic fracture propagation in a naturally fractured reservoir is more complicated than those in reservoirs without any natural fractures (Beugelsdijk 2000, De Pater 2005). One of the mechanisms which cause hydraulic fracture complexity is the geomechanical interaction between a hydraulic and natural fracture. Whether a hydraulic fracture can cross the natural fracture or not is affected by many factors such as in situ stress field, natural fracture frictional properties, fracturing parameters etc. (Norman 1963, Blanton 1982, Warpinski 1987, Zhou 2008, Cheng 2014a, 2015a). The early criteria to identify the geomechanical interaction between hydraulic and natural fractures were based on the Mohr–Coulomb criteria (Blanton 1982, Warpinski 1987, Renshaw and Pollard 1995, Beugelsdijk 2000). Potluri et al (2005) summarized these criteria and compared their applicability. A criterion states that a fracture tip crosses a frictional interface when the compressive stresses acting perpendicularly on the interface are sufficient to prevent slip at the time the stresses ahead of the fracture tip are sufficient to initiate a fracture across the interface (Renshaw and Pollard 1995). Renshaw’s criterion (1995) was extended to identify a hydraulic fracture crossing a natural fracture with cohesion at non-orthogonal intersection angle by Gu and Weng (2010). The recent paper (Cheng et al2014a) extended Renshaw’s criterion (1995) for identifying a mode I hydraulic fracture crossing a natural fracture with arbitrary occurrence in 3D space and proved its high validity using extensive tri-axial fracturing tests. The above mentioned experiments and criteria focus on the geomechanical interaction between a hydraulic fracture and a single discontinuity. Researchers have recently realized that the hydraulic fracture networks are more significant than a single fracture for the exploitation of unconventional reservoirs (tight gas, shale gas, coal-bed methane, etc.). Liu (2014) inserted parallel concrete plates into the 300 mm  ×  300 mm  ×  300 mm specimen to simulate the natural fracture networks in a reservoir. Fan (2014a) studied the evolution of hydraulic fracture networks in a formation that contained two groups of orthogonal cemented fractures. They concluded that the hydraulic fracture networks are formed by the interactive process between the reopening and connecting of the natural fractures. Zhou (2008) also performed a series of experiments on Portland cement blocks to investigate the influence of random natural fracture systems and horizontal stress contrasts on the geometry and propagation. Three types of geometries were observed in their tests: a vertical dominating fracture with multiple branches under a high stress difference; radial fractures around a wellbore at a low stress difference; a partly vertical fracture with random branches under a medium stress difference. These conclusions were experimentally proved by high-energy computed tomography (CT) scanning after tests on shale outcrops (Guo 2014). Numerical simulation (Weng 2011, Wu 2012, Kresse et al2011, 2013, Hall and Wang 2009, 2012, Rao and Wang 2009, Cheng 2015b) on hydraulic fracture propagation inside a fractured medium demonstrates a lot of advantages, such as low cost, oil-field scale and repeatability etc., compared with physical experiments on the real rock sample. However, the complexity of the natural fracture networks and the failure mechanism of rock make it impossible for numerical simulation to obtain precise results. Also, it is widely believed that a concrete block (Blanton 1982, Beugelsdijk 2000, De Pater 2005, Zhou 2008, Cheng 2014a, 2014b) is intrinsically different from a real rock, which results in a low reliability of the hydraulic fracturing experiment. In this article, the hydraulic fracturing experiments were conducted on shale outcrops by injecting slick-water fracturing fluid into a simulated wellbore, a hole drilled at the center of shale specimen. The hydraulic fracture networks of the shale specimen were monitored by acoustic emission during the experiment and indicated by a tracer through the rock breaking after hydraulic fracturing (Zhou 2008, Cheng 2014a, 2014b, Liu 2014, Fan 2014a, 2014b). Thus, these experimental results on the shale outcrops would make some contributions to shale oil and gas development. 2. Experimental procedures 2.1. Experiment setup The experiments were conducted in a true tri-axial hydraulic fracturing system (Zhou 2008). A cubic specimen with the dimension of 400 mm  ×  400 mm  ×  400 mm was positioned among steel platens; the platens created confining stresses on the specimen by the use of hydraulic pumps. The stresses were controlled by three individual pumps and hydraulic voltage stabilizers, respectively. The stresses can also be maintained at the desired values independently during the experiment. The three confining stresses can reach up to 28 MPa individually. The injection pressure of the fracturing fluid can reach up to 140 MPa maximally, which was provided by a hydraulic pump. The injection system is capable of pumping fluids with any injection scheme for continuous volumes up to 800 mL. The experimental control and data acquisition are conducted using customized software running on a personal computer. This system has been used successfully to initiate and propagate hydraulic fractures in a variety of materials using different fracturing fluids. Slick water fracturing fluid is widely used in actual shale gas development. Hence, slick water from the oilfield is used as fracturing fluid in our experiments. Acoustic emissions (AE) were monitored throughout testing in order to obtain source mechanisms. The AE event source mechanism locations were plotted in order to determine whether the spatial relationships existed. 2.2. Specimen preparation The shale outcrops used belong to the Longmaxi formation, Sichuan basin, China. The Longmaxi formation is generally 250–280 m thick. The organic-matter-rich shales at the bottom can be divided into deep and shallow sediments. The fractures develop in the shallow self-sediments, while the fissures develop in the bottom shale. The density of fractures is 1–2 m-1 with the opening 0.5–2 mm. They are filled with calcites. The bedding planes are mostly horizontal while the fractures are high-angle or vertical distributed, as shown in figure 1. Figure 1. Open in new tabDownload slide Shale outcrops in the Longmaxi formation, Sichuang basin, China. (1) Side view. (2) Top view. Figure 1. Open in new tabDownload slide Shale outcrops in the Longmaxi formation, Sichuang basin, China. (1) Side view. (2) Top view. The experimental specimen was made as a cubic shale outcrop with the dimensions 400 mm  ×  400 mm  ×  400 mm (figure 2). In order to observe the natural fractures clearly, the path of the natural fracture was noted by the purple line before the test. All the four specimens have similar bedding planes, which are orthogonal to the vertical principal stress. Six planes are divided by three pairs: back and front are denoted as P5 and P2, respectively; left and right are denoted as P6 and P3, respectively; up and down are denoted as P1 and P4, respectively. The fracturing fluid was injected into a casted center pipe to simulate an in situ wellbore, which had an outer diameter of 16 mm and an inner diameter of 10 mm (figure 3). The length of the open-hole was 60 mm and the depth of the hole was 200 mm. AE were monitored throughout testing from six observation points (figure 3) at the corners of each specimen, using wide-band piezoelectric transducers manufactured by Physical Acoustics Corporation. Figure 2. Open in new tabDownload slide Experimental specimens of hydraulic fracturing. Figure 2. Open in new tabDownload slide Experimental specimens of hydraulic fracturing. Figure 3. Open in new tabDownload slide Sketch of experimental specimen. Figure 3. Open in new tabDownload slide Sketch of experimental specimen. 2.3. Experimental parameters and outline The reservoir parameters in the Longmaxi formation, Sichuan basin are given in the first row of table 1. Note that the elastic module and Poisson’s ratio of shale outcrops are both different from those in a real reservoir. Experimental parameters can be obtained as follows after scaling transformation, as shown in the second row of table 1. Table 1. Reservoir and experiment parameters in Longmaxi formation. . VD(m) . E(GPa) . υ . σh(MPa) . σH(MPa) . σV(MPa) . P0(MPa) . Reservoir 2500 35 0.25 46.25 52.5 50 26.25 Experiment 0 40 0.18 19.1 26.3 22.1 0 . VD(m) . E(GPa) . υ . σh(MPa) . σH(MPa) . σV(MPa) . P0(MPa) . Reservoir 2500 35 0.25 46.25 52.5 50 26.25 Experiment 0 40 0.18 19.1 26.3 22.1 0 Open in new tab Table 1. Reservoir and experiment parameters in Longmaxi formation. . VD(m) . E(GPa) . υ . σh(MPa) . σH(MPa) . σV(MPa) . P0(MPa) . Reservoir 2500 35 0.25 46.25 52.5 50 26.25 Experiment 0 40 0.18 19.1 26.3 22.1 0 . VD(m) . E(GPa) . υ . σh(MPa) . σH(MPa) . σV(MPa) . P0(MPa) . Reservoir 2500 35 0.25 46.25 52.5 50 26.25 Experiment 0 40 0.18 19.1 26.3 22.1 0 Open in new tab Injection displacement is 0.163 ml s-1 until the rock is fractured, and then is improved to 0.326 ml s-1, quickly. After 25 s, the flow rate is set based on the pump pressure curve. The specimens were placed into the true tri-axial pressure equipment. Slick water with the viscosity of 2–3 mPa·s is utilized as fracturing fluid. The formula of fracturing fluid is shown in table 2. In order to track the fracturing fluid inside a hydraulic fracture, a yellow-green additive was mixed with slick water before injecting into the specimens. This additive doesn’t affect the rheological properties of slick water. Table 2. Formula of sliminess water fracturing fluid. Additives . FR-66 . Optikleen WF . BE-9 . Gasperm 1100 . Function Friction reducer Viscosity breaker bactericide Water lock preventer Concentration 0.075% 0.09 0.070% 0.2%–0.5% Unit vol% Kg m-3 vol% vol% Additives . FR-66 . Optikleen WF . BE-9 . Gasperm 1100 . Function Friction reducer Viscosity breaker bactericide Water lock preventer Concentration 0.075% 0.09 0.070% 0.2%–0.5% Unit vol% Kg m-3 vol% vol% Open in new tab Table 2. Formula of sliminess water fracturing fluid. Additives . FR-66 . Optikleen WF . BE-9 . Gasperm 1100 . Function Friction reducer Viscosity breaker bactericide Water lock preventer Concentration 0.075% 0.09 0.070% 0.2%–0.5% Unit vol% Kg m-3 vol% vol% Additives . FR-66 . Optikleen WF . BE-9 . Gasperm 1100 . Function Friction reducer Viscosity breaker bactericide Water lock preventer Concentration 0.075% 0.09 0.070% 0.2%–0.5% Unit vol% Kg m-3 vol% vol% Open in new tab The cubic specimens were loaded by three independent flatjacks to simulate in situ stress conditions. Once the stresses reached their desired values, a delay of approximately 30 min is allowed for the establishment of the stress equilibrium around the wellbore before the fracturing test is commenced. The yellow–green additive is non-penetrating and therefore highlights the fracture surface generated by the experiment. We increased the injection rate for several times, and so the displacement curves look like a step. Acoustic emission is utilized to monitor the fracture propagation between the pump startup until shutdown. Finally, we depressurized the bore-hole and unloaded the tri-axial equipment. After the test, the block was opened along the fracture path using a hammer and chisel. 3. Experiment results and analysis 3.1. Geomechanical intersection of natural and hydraulic fractures The geomechanical interaction between the hydraulic and natural fracture is believed to be a complicated process, as shown in figure 4. In order to show these interactions vividly, this process was divided into three steps according to propagation tendency. Figure 4. Open in new tabDownload slide Sketches of interaction between hydraulic and natural fractures (PF—pre-existing fracture; HF—hydraulic fracture). (a) approaching. (b) non-crossing. (c) crossing. (d) dilation / arrest. (e) crossing. (f) dilation / crossing. (g) activation. (h) offset / activation. (i) activation and crossing. Figure 4. Open in new tabDownload slide Sketches of interaction between hydraulic and natural fractures (PF—pre-existing fracture; HF—hydraulic fracture). (a) approaching. (b) non-crossing. (c) crossing. (d) dilation / arrest. (e) crossing. (f) dilation / crossing. (g) activation. (h) offset / activation. (i) activation and crossing. First step; the hydraulic fracture approaches a natural fracture (figure 4(a)). The fluid frontier goes behind the fracture frontier because of fluid lag. Under this condition, the net pressure at the intersection point is zero, while the natural fracture is influenced by the stress field of the hydraulic fracture. Hence, whether the hydraulic fracture can cross the natural fracture only depends on the stress field adjacent to the hydraulic fracture tip and the property of the natural fracture. Their interactions are divided into two types: non-crossing (figure 4(b)) or crossing (figures 4(c) and (e)). A hydraulic fracture must meet two conditions to cross a natural fracture: (1) the maximum tensile stress at the hydraulic fracture tip is equal to the tensile strength of the rock on the opposite side of the natural fracture; (2) no shear slippage occurs in the natural fracture surface. The properties of the material that composes the interface control the degree of sufficiency of the frictional resistance or cohesion available to prevent slip and allow crossing (Wang et al2013). Therefore, a criterion to identify this interaction at an orthogonal intersection angle was first proposed by Renshaw (1995), which was extended to a hydraulic fracture crossing a natural fracture with cohesion at non-orthogonal intersection angles by Gu (2010). Recently, Cheng et al (2014a) developed a new criterion to identify a mode I hydraulic fracture crossing a natural fracture with arbitrary occurrence in 3D. Note that the natural fracture networks are composed of many natural fractures, and every natural fracture can independently affect the hydraulic fracture. Thus, these criteria can be useful to understand the formation with complex fracture networks by hydraulic fracturing and in applying to reservoir simulations after appropriate scaling transformation. Second step; the fluid pressure at the intersection point increases after the fluid frontier contacted the natural fracture. Fracture fluid leaks along the natural fracture. The natural fracture dilates when the fluid pressure in a natural fracture is higher than the normal stress, i.e. figure 4(b) goes to figure 4(d) (corresponding to the experimental photo in figure 6(d)), which produces bifurcation with two branches. Similarly, the fracture fluid at intersection point in figure 4(c) will also diffuse along the natural fracture. The natural fracture remains closed when the fluid pressure inside is lower than the normal stress, and the hydraulic fracture will propagate along the previous path (figure 4(e) and 5(e)). By contrast, when fluid pressure inside exceeds the normal stress acting on the natural fracture’s interface, the natural fracture will dilate (figure 4(f), 5(f) and 6(f)), which produces bifurcation with three branches. It is believed that fracture bifurcation in figures 4(d) and (f) have the potential to obtain fracture networks. Figure 5. Open in new tabDownload slide Photos of shale specimen 3# (‘(e)’, ‘(f)’ in figure 5 correspond to the figures 4(e) and (f) respectively). (1) Before experiment. (2) After experiment. Figure 5. Open in new tabDownload slide Photos of shale specimen 3# (‘(e)’, ‘(f)’ in figure 5 correspond to the figures 4(e) and (f) respectively). (1) Before experiment. (2) After experiment. Figure 6. Open in new tabDownload slide Photos of shale specimen 2# (‘(d)’, ‘(f)’, ‘(h)’, ‘(i)’ in figure 6 correspond to figure 4(e), (f), (h), (i) respectively). (1) Before experiment. (2) After experiment. Figure 6. Open in new tabDownload slide Photos of shale specimen 2# (‘(d)’, ‘(f)’, ‘(h)’, ‘(i)’ in figure 6 correspond to figure 4(e), (f), (h), (i) respectively). (1) Before experiment. (2) After experiment. Third step, as the fracturing fluid pumps on, the fluid frontier reaches the natural fracture tip, which activates the natural fracture. The expanding angle of hydraulic fracture depends on stress intensity factors (Cheng et al2015a). The roughness of a natural fracture causes the stress concentration at the defect. New fractures appear and then bifurcate with multi-fractures, as shown in figure 4(h) and 6(h). Additionally, when fluid pressure inside a natural fracture (figure 4(f), 5(f) and 6(f)) is higher than the normal stress acting on the natural fracture’s interface, the natural fracture will be activated to expand along its tip, which leads to a triple-branched fracture (figure 4(i) and 6(i)). For convenience, figures 4(i) and (e) are classified into one type. As a result, the geomechanical interaction between the natural and hydraulic fracture could be classified into three types as follows: Crossing, such as figures 4(e) and (i), when the hydraulic fracture continues its propagation in the same direction. This usually happens for the large approaching angles. Non-crossing (arrest), such as figure 4(g), hydraulic fracture could not cross the natural fracture and continues its propagation along the natural fracture; Offset, such as figure 4(h), hydraulic fracturing fluid enters the natural fracture, propagates along it for some distance and reinitiates the fracture on the opposite side of the natural fracture in a direction normal to the minimum far-field stress. The above analysis shows that the interaction mechanism in step one and two (connecting the natural fracture) is suitable not only for the interaction between hydraulic and natural fractures, but also for the interaction between hydraulic fractures and other discontinuities, such as bedding planes/faults. Compared to bedding planes and faults, a natural fracture has a finite length. Hence, the propagation mode in step three (activating the natural fracture) is key for hydraulic fracturing to develop fracture networks in naturally fractured formations. On behalf of complex fractures in naturally fractured reservoirs with low porosity and low permeability, hydraulic fracturing is expected to create new fractures, as well as extensive natural fractures. These complex fracture networks will improve the hydraulic conductivity for oil and gas. When a hydraulic fracturing process is carried out on this type of naturally fractured formation, the hydraulic fracture will connect natural fractures preferentially. The overall propagation direction of a hydraulic fracture network still follows the path of least resistance. The study of how a hydraulic fracture geomechanically interacts with the natural fractures also offer many insights in understanding the interaction of hydraulic fractures with all forms of discontinuity such as lithologic boundaries, bedding planes, joints, cleats and other interfaces that may exist underground. 3.2. Influences of viscosity and injection rate In situ stresses and natural fractures in a petroleum reservoir are usually uncontrollable. Engineering factors controlling the fracturing of petroleum reservoirs are typically managed by the field engineer. The viscosity and injection rate of fracturing fluid are the main factors in fracturing design. In a reservoir with natural fracture networks, the fracturing fluid will diffuse inside the natural fractures. It means that fluid pressure inside a wellbore will not reach the breakdown pressure if the injection volume is less than the diffusion volume. A high flow rate yields fluid-driven fractures, while a low flow rate just opens an existing fracture network (De Pater 2005). If all the natural fractures inside a reservoir are distributed randomly, in situ stress (compressive stress) will generally affect the opening. The fracture opening, orthogonal to the minimum horizontal in situ stress, will be larger than that along the other directions because of their lower compressions. It can be easily predicted that most fracturing fluid tends to diffuse along the maximum horizontal in situ stress. The spatial envelope of fracturing fluid will be an approximate ellipsoid with the major axis along the direction of maximum horizontal stress, which has been experimentally proved by (Fan and Zhang 2014a). It is assumed that volumetric fluid flow rate in a given natural fracture is governed by Poiseuille’s classical parallel-plate flow equation. High viscosity produces high resistance for fracturing fluid flowing, which achieved only moderate activation of natural fractures with a limited stimulated network, indicating limited penetration of the fracturing fluids into the natural fracture and rock matrix. Hence, the fluid in the wellbore will hold high pressure and drive fluid into the larger opening natural fractures, i.e. orthogonal to the minimum horizontal in situ stress. The overall propagation direction of the hydraulic fracture network still follows along the maximum horizontal in situ stress because of its low resistance. By contrast, the lower-viscosity fracturing fluid achieved a larger network indicating the penetration of fracturing fluids through a wider network of natural fractures (Kresse et al2013) Two groups of hydraulic fracturing experiments were conducted on concrete blocks under tri-axial stresses in large-sized fracturing systems (Cheng 2014b). The pre-existing fracture and properties of blocks are the same. Different pump displacements with the same viscosity of fracturing fluid were applied in the first group, whereas different viscosities with the same the pump displacement were applied in the second group. Results showed that the hydraulic fracture can cross the natural fracture under large injection rate and large viscosity, while hydraulic fractures can’t cross natural fractures under low injection rate and low viscosity. When the viscosity stays the same, hydraulic fractures can cross natural fractures above the critical injection rate; below the critical displacement, hydraulic fractures only propagate along the surface of natural fractures to their end, rather than across them. These two critical values offer insights in the design of fracturing parameters. In general, high injection rate (usual about 10–15 m3 min-1) and low viscosity (water-based fracturing fluid) were commonly used in hydraulic fracturing stimulations of shale gas reservoirs. Actually, the injection rate in the oil-filed increases gradually. Therefore, step-displacement hydraulic fracturing was used in our experiments (figures 7–9). After the wellbore is broken down, the treatment pressure increases as the flow rate increases (figures 7–9). It can be observed that the pump pressure fluctuation (pressure pulse) of figure 7 is more drastic than that in figure 9, and the pressure fluctuation in figure 8 is more drastic than that in figure 9. As a result, complexity of the hydraulic fracture networks after experiments in these specimens satisfy the relationship: 2# (figure 6)  >3# (figure 5)  >4# (figure 12). Fan et al (2014b) indicated that pump pressure profiles could reflect different characteristics of extending behaviors. Pressure pulse inside the fracture will create and connect micro-cracks in the rock mass. Figure 7. Open in new tabDownload slide Pump curves during the experiments in shale specimen 2#. Figure 7. Open in new tabDownload slide Pump curves during the experiments in shale specimen 2#. Figure 8. Open in new tabDownload slide Pump curves during the experiments in shale specimen 3#. Figure 8. Open in new tabDownload slide Pump curves during the experiments in shale specimen 3#. Figure 9. Open in new tabDownload slide Pump curves during the experiments in shale specimen 4#. Figure 9. Open in new tabDownload slide Pump curves during the experiments in shale specimen 4#. 3.3. Characteristics of acoustic emissions Micro-seismic monitoring is an accepted tool used to evaluate the geometry and evolution of a fracture network induced during a hydraulic treatment (Fan and Zhang 2014b). Induced micro-seismic events are those associated with the opening and slip of fractures caused by the stimulation fluid (Tan et al2014). Triggered micro-seismic events are driven by the change in stress field associated with the induced strain within and surrounding the zone of stimulated fractures. Triggered events can give a false sense of the stimulated reservoir volume (SRV). On the contrary, the presence of the treatment fluid is direct evidence for fractures interconnected with the wellbore. The principle of micro-seismic events is similar to acoustic emissions during rock tests. AE were monitored in order to characterize the stimulated reservoir and perform simplified moment tensor analysis to obtain source mechanisms (Chitrala 2012, Hampton 2013). The AE event source mechanism locations were plotted in order to determine the position of fractures. As shown in figure 10, the area inside yellow ellipse figure 10(2), corresponding to the natural fracture in figure 10(1), only produces a few AE events. By contrast, the zone where new fractures are created produces more AE events. Figure 10. Open in new tabDownload slide AE event source distribution of specimen 3#. (1) Before test. (2) AE events distribution. Figure 10. Open in new tabDownload slide AE event source distribution of specimen 3#. (1) Before test. (2) AE events distribution. The number of accumulated AE events associated with the pump pressure is demonstrated in the figure 11. During the period of pressure increase, AE event numbers increase drastically. This period occurs earlier than the moment of rock breakdown, which means extensive micro-cracks were created before the peak pressure. Around the peak pressure, there are few AE events shown in the accumulated AE events. This phenomenon was observed in all the experiments. Rock breakdown should be the process used to connect all the micro-cracks. After the peak pressure, the increasing of injection displacement results in more AE events (figure 11). These AE events prove that more new fractures are created by large injection displacement. Figure 11. Open in new tabDownload slide AE event number and pump pressure of specimen 3#. Figure 11. Open in new tabDownload slide AE event number and pump pressure of specimen 3#. 3.4. Influences of in situ stresses It is widely known that hydraulic fractures open against the minimum principal stress. Hydraulic fractures will mainly propagate horizontally in a reverse fault stress regime. On the contrary, hydraulic fracture will mainly propagate vertically in a normal or strike-slip fault stress regime. Three types of geometries were observed in the experiments conducted by Zhou (2008): a vertical dominating fracture with multiple branches under high stress difference; radial fractures around a wellbore at low stress difference and a partly vertical fracture with random branches under medium stress difference. Basically, this opinion holds when their cement blocks are only under the condition of tectonic stress regime for normal faults. In a 3D reservoir, three in situ stresses should be considered together to determine the fracture propagation inside a fractured reservoir. Only the horizontal stress difference coefficient (Guo 2014, Zhou 2008) roughly predicts the fracture propagation. The normal and shear stiffness of a natural fracture mainly controls its initial opening under a specific in situ stress regime. The opened fracture or fault will not only change the local stress state, but also absolve the fracturing fluid, which leads to the failure of reservoir stimulations. A large width natural fracture was present in specimen 4#. When a hydraulic fracture approaches this natural fracture, the fracturing fluid will quickly leak and lead to a simple fracture (figure 12). It is recommended that hydraulic fracturing in an oilfield should avoid faults or open natural fracture systems. Figure 12. Open in new tabDownload slide Photos of specimen 4# after experiment. Figure 12. Open in new tabDownload slide Photos of specimen 4# after experiment. 4. Discussion and conclusion This article provides an experimental work on hydrofracking with specimens from shale outcrops and experimental parameters from shale gas reservoirs, avoiding the difference between concrete blocks and real shale. Results showed that the geo-mechanical interaction between hydraulic and natural fractures should be divided into three steps. The first two steps are also suitable to describe the interaction mechanism between hydraulic fractures and other discontinuities. The third step is critical to obtain a hydraulic-fracture network. Every natural fracture can independently affect a hydraulic fracture, but the overall propagation direction of a hydraulic fracture network still follows the maximum horizontal in situ stress because of its least resistance in naturally fractured formations. After the breakdown pressure, pressure fluctuation increases due to an increase in injection rate, leading to the creation of more micro-cracks. The geometry ofhydraulic fracture networks is dominated not only by in situ stresses, but also by the properties of the natural fractures. Hydraulic fracturing by step-displacement can create complex fracture networks that increase the stimulated reservoir volume in naturally fractured reservoirs. The fracturing mechanism in a natural fracture network totally differs from early theories, which predict the propagation of a single fracture by the tensile failure of the rock and fluid conservation. Although it requires further studies to quantify fracture network initiation, propagation and interaction, this method is still advantageous to model and research the fracture geometry in a real reservoir. Acknowledgments This research was supported by the Natural Science Foundation of China (51325402), China Scholarship Council (201406440009). Thanks to China Scholarship Council for supporting the first author of this paper to research in Georgia Institute of Technology, Atlanta, GA, USA. References Beugelsdijk L J L , Pater C J , Sato K . , 2000 Experimental hydraulic fracture propagation in a multi-fractured medium SPE Asia Pacific Conf. on Integrated Modelling for Asset Management Yokohama, Japan 25–26 April paper #59419 Blanton T L . , 1982 An experimental study of interaction between hydraulically induced and pre-existing fractures SPE Unconvential Gas Recovery Symp. Pittsburg, PA 16–18 May paper #10847 Cheng W , Jin Y , Chen M . , 2014a A criterion for identifying hydraulic fractures crossing natural fractures in 3D space , Pet. Explor. Dev , vol. 41 (pg. 371 - 376 ) 10.1016/S1876-3804(14)60042-2 Google Scholar Crossref Search ADS WorldCat Crossref Cheng W , Jin Y , Chen Y , Zhang Y , Diao C , Wang Y . , 2014b Experimental investigation about influence of natural fracture on hydraulic fracture propagation under different fracturing parameters 8th Asian Rock Mechanics Symp Sapporo, Japan Cheng W , Jin Y , Chen M . , 2015a Reactivation mechanism of natural fractures by hydraulic fracturing in naturally fractured shale reservoirs , J. Nat. Gas Sci. Eng. , vol. 23 (pg. 431 - 439 ) 10.1016/j.jngse.2015.01.031 Google Scholar Crossref Search ADS WorldCat Crossref Cheng W , Jin Y , Li H , Chen M . , 2015b A novel linear triangular element of 3D displacement discontinuity method , Eng. Anal. Bound. Elem , vol. 59 (pg. 89 - 96 ) 10.1016/j.enganabound.2015.04.020 Google Scholar Crossref Search ADS WorldCat Crossref Chitrala Y , Sondergeld C H , Rai C R . , 2012 Acoustic emission studies of hydraulic fracture evolution using different fluid viscosities The 46th US Rock Mechanics /Geomechanics Symp Chicago 24–27 June 2012 De Pater C J , Beugelsdijk L J L . , 2005 Experiments and numerical simulation of hydraulic fracturing in naturally fractured rock The 40th US Symp. on Rock Mechanics Anchorgae, Alaska 25–29 June 2005 Fan T , Zhang G . , 2014a Laboratory investigation of hydraulic fracture networks in formations with continuous orthogonal fractures , Energy , vol. 74 (pg. 164 - 173 ) Google Scholar Crossref Search ADS WorldCat Fan T , Zhang G , Cui J . , 2014b The impact of cleats on hydraulic fracture initiation and propagation in coal seams , Pet. Sci. , vol. 11 (pg. 532 - 539 ) Google Scholar Crossref Search ADS WorldCat Gu H , Weng X . , 2010 Criterion for fractures crossing frictional interfaces at nonorthogonal angles The 44th US Rock Mechanics Symp. and 5th US-Canada Rock Mechanics Symp. Salt Lake City, Utah 27–30 June 2010 Guo T , Zhang S , Qu Z , Zhou T , Xiao Y , Gao J . , 2014 Experimental study of hydraulic fracturing for shale by stimulated reservoir volume , Fuel , vol. 128 (pg. 373 - 380 ) 10.1016/j.fuel.2014.03.029 Google Scholar Crossref Search ADS WorldCat Crossref Hall F , Wang Y . , 2009 Elastic wave modelling by an integrated finite difference method , Geophys. J. Int. , vol. 177 (pg. 104 - 114 ) 10.1111/j.1365-246X.2008.04065.x Google Scholar Crossref Search ADS WorldCat Crossref Hall F , Wang Y . , 2012 Seismic response of fractures by numerical simulation , Geophys. J. Int. , vol. 189 (pg. 591 - 601 ) 10.1111/j.1365-246X.2012.05360.x Google Scholar Crossref Search ADS WorldCat Crossref Hampton J , Frash L , Gutierrez M . , 2013 Investigation of laboratory hydraulic fracture source mechanisms using acoustic emission The 47th US Rock Mechanics/ Geomechanics Symp San Francisco, CA 23–26 June 2013 Kresse O , Cohen C , Weng X , Wu R , Gu H . , 2011 Numerical modeling of hydraulic fracturing in naturally fractured formations 45th US Rock Mechanics / Geomechanics Symp. San Francisco, CA June 26–29 2011 Kresse O , Weng X , Gu H , Wu R . , 2013 Numerical modeling of hydraulic fractures interaction in complex naturally fractured formations , Rock Mech. Rock Eng. , vol. 46 (pg. 555 - 568 ) 10.1007/s00603-012-0359-2 Google Scholar Crossref Search ADS WorldCat Crossref Liu Z , Chen M , Zhang G . , 2014 Analysis of the influence of a natural fracture network on hydraulic fracture propagation in Carbonate Formations , Rock Mech. Rock Eng. , vol. 47 (pg. 575 - 587 ) 10.1007/s00603-013-0414-7 Google Scholar Crossref Search ADS WorldCat Crossref Norman L , Jessen F W . , 1963 The effects of existing fractures in rocks on the extension of hydraulic fractures , J. Pet. Technol. , vol. 15 (pg. 203 - 209 ) 10.2118/419-PA Google Scholar Crossref Search ADS WorldCat Crossref Potluri N , Zhu D , Hill A D . , 2005 Effect of natural fractures on hydraulic fracture fracture propagation SPE European Formation Damage Conf. Sheveningen, The Netherlands 25–27 May paper #94568 Rao Y , Wang Y . , 2009 Fracture effects in seismic attenuation images reconstructed by waveform tomography , Geophysics , vol. 74 (pg. R25 - 34 ) 10.1190/1.3129264 Google Scholar Crossref Search ADS WorldCat Crossref Renshaw C E , Pollard D D . , 1995 An experimentally verified criterion for propagation across unbounded frictional interfaces in brittle, linear elastic materials , Int. J. Rock Mech. Min. Sci. Geomech. Abstr. , vol. 32 (pg. 237 - 249 ) 10.1016/0148-9062(94)00037-4 Google Scholar Crossref Search ADS WorldCat Crossref Tan Y , Chai C , Engelder T . , 2014 Use of S-wave attenuation from perforation shots to map the growth of the stimulated reservoir volume in the Marcellus gas shale , Leading Edge , vol. 33 (pg. 1090 - 1096 ) 10.1190/tle33101090.1 Google Scholar Crossref Search ADS WorldCat Crossref Warpinski N R , Teufel L W . , 1987 Influence of geologic discontinuities on hydraulic fracture propagation , JPT , vol. 39 (pg. 209 - 220 ) 10.2118/13224-PA Google Scholar Crossref Search ADS WorldCat Crossref Weng X , Kresse O , Cohen C . , 2011 Modeling of hydraulic fracture network propagation in a naturally fractured formation , SPE Production & Operations , vol. 26 (pg. 368 - 80 ) Google Scholar Crossref Search ADS WorldCat Wang W , Olson J E , Prodanovic M . , 2013 Natural and hydraulic fracture interaction study based on semi-circular bending experiments SPE Unconventional Resources Technology Conf. Denver, CO 12–14 August paper #168714 Wu R , Kresse O . , 2012 Modeling of interaction of hydraulic fractures in complex fracture networks SPE Hydraulic Fracturing Technology Conf. The Woodlands, TX 6–8 February paper #152052 Zhou J , Chen M , Jin Y , Zhang G . , 2008 Analysis of fracture propagation behavior and fracture geometry using a tri-axial fracturing system in naturally fractured reservoirs , Int. J. Rock Mech. Min. Sci. , vol. 45 (pg. 1143 - 1152 ) 10.1016/j.ijrmms.2008.01.001 Google Scholar Crossref Search ADS WorldCat Crossref © 2015 Sinopec Geophysical Research Institute
TI - Experimental study of step-displacement hydraulic fracturing on naturally fractured shale outcrops
JF - Journal of Geophysics and Engineering
DO - 10.1088/1742-2132/12/4/714
DA - 2015-08-01
UR - https://www.deepdyve.com/lp/oxford-university-press/experimental-study-of-step-displacement-hydraulic-fracturing-on-RsU1MS1uIl
SP - 714
VL - 12
IS - 4
DP - DeepDyve
ER -