Baker Hughes introduced its REAL Connect service designed to help improve hydrocarbon recovery in hydraulic-fracturing applications. The service's diverter systems redirect fracturing-fluid flow to the untreated perforation sets within a stage to increase fracture-network complexity and provide more-uniform coverage. Once the stimulation treatment is complete and the diverter materials dissolve, ultralightweight proppant material remains in the near-wellbore region for long-lasting production flow paths. The service helped an operator in De Soto Parish, Louisiana, rejuvenate several mature wells from their existing inventory of unconventional wells that were reaching the low end of the production curve. Baker Hughes designed a refracturing program that used the REAL Connect service to temporarily isolate existing fracture networks in order to redirect fluids to untreated zones and stimulate untapped portions of the reservoir.
This is the flexural-slip story, which "speaks" to the deformation PROCESS that causes fracturing to occur Then the core data arrives... Oh, no! It's a fractured reservoir This is the main point! The workflow is that the "far field" stress is resolved onto each plane, and then the one(s) with the smaller eff normal stresses are Those predictions are contradicted by simulations involving interactions (see later). Simple elasticity is not sufficient Why? Need more than 2x this volume to hold "lost" f
During hydraulic fracturing, natural fractures and bedding planes can intersect with growing hydraulic fractures and form complex fracture networks. This can result in the flow of fluid and proppant in convoluted fracture pathways with highly variable fracture width and height. Existing models of hydraulic fracturing assume a planar fracture geometry and are unable to simulate proppant placement in such complex networks. In this work, we investigate proppant transport in growing fracture networks using a fully three-dimensional, geomechanical fracture flow, network model with the ability to simulate proppant transport.
A three-dimensional hydraulic fracturing simulator developed using the displacement discontinuity method is coupled with a network model for proppant transport. The simulator captures the effect of proppant concentration, fracture width, and fluid rheology on proppant transport. The equations for the fracture network geomechanics, the fluid flow, and the proppant transport are solved in a coupled manner. This provides an accurate estimation of both the fluid pressure and the proppant distribution as the fracture network grows. The geometry of each fracture segment affects the flow distribution in the network. Simulations are then conducted to study the redistribution of proppant as it settles in the fracture network during shut-in to get the final proppant distribution in the network.
It is observed that changes in the in-situ stress due to heterogeneity and the stress-shadow induced near the intersection of a hydraulic fracture and a natural fracture may reduce the fracture width and suppress the ability of the proppant to move into the natural fracture. In low permeability formations, due to low leak-off rates, the proppant almost always forms a proppant bank at the bottom of the fracture during shut-in. For planar fractures, proppant settling may disconnect the conductive proppant bank from the wellbore, isolating the productive propped fracture from the wellbore. This problem is exaggerated in the case of fracture networks, where every intersection point between fractures can potentially act as a bottleneck for the flow of produced hydrocarbons. The increase in the surface area due to hydraulically connected natural fractures increases fluid leak-off, reduces the average width of the fracture network, increases proppant concentration, and increases the likelihood of proppant bridging.
This work allows us to improve our understanding of proppant placement in three-dimensional, mechanically interacting, complex fracture networks. By coupling geomechanics with proppant transport in fracture networks, it is now possible to study the impact of the stress shadow on proppant placement in natural fractures. The results will assist in improving hydraulic fracture design for naturally fractured reservoirs.
A modeling exercise was performed investigating hydraulic fracture interaction with pre-existing fractures, based on a benchmark modeling exercise lead by ARMA (American Rock Mechanics Association). The modeled scenarios are based on interaction with two existing faults under different geomechanical conditions. Simulations were performed with a coupled hydraulic-geomechanical-seismological, discrete-element model. The results show that the hydraulic fracture aperture is restricted at the intersection with the faults, to a degree depending on the slip induced on the fault. The aperture restriction was found to also limit the extent of the proppant distribution. In scenarios with relatively more slip, the fault activation was found to occur above and below the injection layer and resulted in some hydraulic fracture height growth relative to the depth contained fracture occurring without fault activation. As expected, the fault slip and intensity of associated microseismicity is related to the geomechanical prepotency for slip. Fractures oriented at 30° to the hydraulic fracture and maximum principal stress direction, resulted in the largest microseismic magnitude and significant hydraulic fracture restriction at the fault intersection. With increasing angle, the magnitude decreased and the hydraulic fracture had a larger half-length. Increasing the differential stress resulted in increased magnitudes and decreased fracture half-length, as did the case of a weaker fault. The study demonstrates through a numerical-physics simulation how a hydraulic fracture system behalves as it interacts with a pre-existing fracture under various geomechanical conditions.
Refracturing can be an appealing technique to mitigate flow rate decline in wells where the original treatment failed to adequately stimulate the formation. To optimize refracturing performance, it is crucial to understand the stress redistribution due to the poroelastic effect, which determines candidate selection, timing, and effectiveness of refracturing. In reservoirs containing natural fractures, stress redistribution can be complicated. Very few papers in the literature discuss refracturing in naturally fractured reservoirs. The objective of this work is to predict stress redistribution due to depletion and optimize the timing and locations for refracturing in reservoirs with complex hydraulic and natural fractures.
In this study, pressure and stress distribution due to depletion in a highly naturally fractured reservoir are predicted using our coupled fluid flow and geomechanics model with Embedding Discrete Fracture Model (EDFM). The model was developed based on a well-known fixed-stress split, which is unconditionally stable. EDFM was coupled to the model to gain the capability of simulating complex fracture geometries and high-density natural fracture system using structured grids. The model was validated against classical Mandel's problems to ensure accuracy. In addition, the effects of natural fractures density, hydraulic fracture spacing, differential in-situ stress, and reservoir permeability have been studied.
Synthetic cases with multiple natural fractures were created to study the effects of natural fractures on stress evolution. The results show that there is a significant difference in stress redistribution due to production when comparing a highly naturally fractured reservoir with a reservoir without natural fractures.. The stress distribution has to be considered when determining where to initiate the new hydraulic fractures. The critical time to perform refracturing is also recommended at different scenarios as orientation and magnitude of principal stresses change as reservoir pressure declines over the time.
Beyond the critical timing, the child fractures may not be able to propagate towards intact areas at all and may damage parent fractures as a result of the reversal of maximum horizontal stress. This difference indicates that effect of natural fractures cannot be neglected in highly fractured reservoirs when performing refracturing. A change in density of natural fractures directly affects the size and shape of depleted areas resulting in alteration of stress redistribution both inside and outside SRV region. Other parameters, such as. hydraulic fracture spacing, differential in-situ stress, and reservoir permeability should also be taken into consideration when studying refracturing as they affect magnitude and redistribution of principal stresses and yield different optimum locations and critical timing in highly fractured reservoirs.
This paper predicts the stress evolution induced by depletion in highly fractured reservoirs and considers the effects of heterogeneous natural fracture distribution and density on stress redistribution. The results can be use to determine the optimum refracturing locations as well as critical timings to perform refracturing, which provides critical insights for refracturing optimization in highly fractured reservoirs.
Complex fracture networks are formed when hydraulic fractures grow in naturally fractured reservoirs. Current planar fracture models are inadequate for capturing the effect of natural fractures on fracture propagation and addressing the important question of optimum fracture spacing and well spacing. Stress interference due to three-dimensional fracture networks can result in intricate fracture geometries, which are usually neglected by fracture models. In this paper, we present a three-dimensional hydraulic fracturing simulator that models the deformation and stress fields induced by both the dilation and shear failure of all existing and propagating hydraulic or natural fractures. It is shown that the simulator allows us to capture the complex fracture geometries, and microseismic signatures often observed in heterogeneous and naturally fractured rocks.
Fracture geomechanics is modeled in a computationally efficient manner using a fully three-dimensional displacement discontinuity method. The simulator captures the physics of fracture growth, fracture turning, fluid distribution in fracture networks, and the intersection of hydraulic fractures with pre-existing natural fractures. The model captures the interaction between multiple branches of a hydraulic fracture (stress shadow effect). The model also simulates the shear failure of hydraulically disconnected natural fractures to simulate microseismic activity and can account for the effect of shear failure and slippage along bed boundaries and along natural fractures on hydraulic fracture propagation.
The effect of pre-existing natural fracture density and orientation on the geometry of the fracture network generated is systematically studied. It is shown that natural fractures play an important role in determining the propagation direction of hydraulic fractures and this effect is quantified. At high natural fracture density, the propagation direction of a hydraulic fracture is dominated by the orientation of natural fractures rather than the far field stress magnitude and direction. The density of the natural fractures also affects the complexity of the final created fracture geometry.
Syfan, Frank E. (Rhino Chemical International) | Holcomb, David L. (Pentagon Technical Services) | Lowrey, Terry A. (Rhino Chemical International) | Nickerson, Randy L. (Caza Petroleum) | Sam, Anthony B. (Caza Petroleum) | Ahmad, Yusra (Nissan Chemical America Corporation)
During 2016 - 2017, the stimulation of several Wolfcamp and Bone Springs targeted wells in the northern Delaware Basin using fracturing treatments and a new method employing relatively small pre-pad pill volumes of nanoparticle dispersions ahead of each stage of treatment have been successfully performed. The liquid nanoparticle dispersion pre-pad pills used in each stage consist of highly surface modified, neutral wet nanoparticles dispersed in small water volumes which penetrate the reservoir's natural fracture and secondary induced fracture network via accelerated diffusion into the reservoir beyond the primary induced fracture network, producing a Brownian motion activated, mechanical advantage process known as disjoining pressure. These neutral wet, solid nanoparticles aid in delivering improved efficacy in the recovery of hydrocarbons via counter-current imbibition by fragmenting the disjoined oil into smaller oil droplets, enabling a more efficient backflow to the propped fracture network and back to the wellbore. The nanoparticle dispersion pre-pad pills are used to introduce the nanoparticle dispersion farther into the reservoir's naturally fractured/permeable matrix network and employ the mechanisms described above to enhance and sustain the mobility of hydrocarbons (overcome confining capillary pressures) back to the propped facture network and subsequently to the wellbore.
Advantages of this method and mechanical process are the ability to access the reservoir beyond where proppants can be placed thereby improving the effective stimulated reservoir volume. The evaluation of results in these zones has shown that utilizing the nanoparticle pre-pad pills has resulted in significantly improved performance (20-30 percent higher than the best available technology) compared to over 15 offset wells when normalized with respect to stimulation and production techniques and procedures. The results include significantly higher oil cuts for a longer sustained time periods as well as providing earlier oil to surface after treatment flowback and continued flowing production before requiring artificial lift implementation. While the number of offset wells is greater than nanoparticle dispersion treated wells, the results clearly demonstrate that there is a positive impact as well as persistent improvement factor (EUR) from using the pre-pad nanoparticle dispersion pills ahead of each stage. The liquid nanoparticle dispersion pre-pad pills used in each stage consist of highly surface modified, neutral wet nanoparticles dispersed in small water volumes which penetrates via accelerated diffusion into the reservoir beyond the induced fracture network, producing a Brownian motion activated, mechanical process known as disjoining pressure. These neutral wet, solid nanoparticles aid in delivering improved efficacy in the recovery of hydrocarbons via counter-current imbibition by fragmenting the disjoined oil into smaller oil droplets, enabling a more efficient backflow to the propped fracture network and back to the wellbore.
Lee, Hunjoo P. (Center for Petroleum and Geosystems Engineering, The University of Texas at Austin) | Razavi, Omid (Center for Petroleum and Geosystems Engineering, The University of Texas at Austin) | Olson, Jon E. (Center for Petroleum and Geosystems Engineering, The University of Texas at Austin)
The propagation path of hydraulic fracture is significantly affected by the properties of cemented natural fractures (i.e., veins), such as, stiffness, tensile and shear strength, and flow characteristics. Moreover, the magnitude of stress anisotropy can control the results of hydraulic fracture crossing or diverting into the vein. To study the effect of vein properties and the differential stress on the interaction of hydraulic fractures with mineral-filled veins, we employed a three-dimensional (3D) Discrete Element Method (DEM) model which couples the fluid flow and the bonded-particle assembly. The hydraulic fracture interaction with the vein was investigated by monitoring the bond breakages and measuring the fracture diversion distance along the vein before kinking back into the rock matrix. Numerical results are in good agreement with the fracture diversion results from the published experimental and numerical SCB tests of Marcellus shale with calcite-filled veins. The propagating hydraulic fracture is more likely to divert into the veins with smaller approach angle. Increasing the differential stress leads to fracture crossing than diversion into the vein. For veins with greater stiffness than the host rock, microcracks were generated in the vein before the hydraulic fracture intersected it. For tight formations with stiffer veins, the vein permeability did not have much influence on the fracture diversion result since the damage induced by the stress concentration ahead of the hydraulic fracture tip dominated the fluid flow in the vein.