Search

Summary Effective proppant placement has been one of the key objectives of hydraulic fracturing. Different proppant and fracture fluid characteristics and placement methodologies have been historically applied based on learnings from standard proppant transport studies with parallel plate slots. The standard test setup represents a simplified planar fracture with constant width and confined height, incorporating only basic flow characteristics, and thus, is inadequate to capture unique phenomena of proppant transport in unconventional reservoirs. In this study, proppant transport laboratory tests were conducted on a large-scale (10×20 ft) tortuous slot flow system. This novel setup incorporates many significant unconventional fracture features, including lateral and vertical tortuosity, variable width, leakoff, fluid dynamics replicating upward fracture growth, and so on. Proppant transport behavior was investigated with multiple parameters such as proppant size, density, and concentration; fracture fluid type and viscosity; pumping sequence; pump rate; and fracture properties (width, leakoff location and rate, fracture tortuosity profile, and flow directions). The detailed parametric and integrated study of test results includes analysis of proppant dune evolution, dune shape, particle-size distribution across dune, propped area, fluid, and proppant collected from leakoff and exit ports. Multiple unique phenomena occurring at tortuous interfaces were observed, including the generation of isolated pockets of proppant pack, restriction of upward movement owing to proppant bridging, and creation of discontinuous and sparsely distributed proppant pillars above the dune. The test results demonstrated a larger proppant dune angle in front of the dune peak during injection and a subsequent falloff of proppant pack with a higher percentage of smaller mesh proppant backfilling the area at and near the inlet (analogous to the wellbore). Self-segregation of proppant in slickwater as per mesh size resulted in higher percentage of smaller mesh (larger size) proppant settled near the injection point, and a higher percentage of larger mesh (smaller size) proppant placed farther in the system. These observations and novel learnings highlight that it is critical to account for tortuous fracture pathway, leakoff effects, and flow directions (both lateral and upward) to better understand proppant transport behaviors in unconventional fractures. A partially proppant-filled fracture area is recognized in unconventional fracture in addition to general classification of propped and unpropped fracture area. Using proppant with large mesh size distribution range or pumping smaller mesh proppant first in slickwater helps achieve dual benefits of higher near-wellbore conductivity and improved far-field transport. This study demonstrates and physically verifies unique proppant transport behaviors in unconventional hydraulic fractures. It also provides novel learnings that will help the industry to optimize hydraulic fracture design through the selection of optimum proppant and fluid properties with enhanced pumping strategies for overall well productivity improvement in an unconventional reservoir.

Summary The objective of this study is to understand how microseismic events are generated during hydraulic fracturing, as well as the role of geomechanical conditions (i.e., stress and mechanical stratigraphy) in this process. In the industry, microseismic event clouds have been generally used as an “outer boundary” of the “stimulated reservoir volume” (SRV). However, by comparing with other surveillance data (low-frequency distributed acoustic sensing, or LF-DAS, strain) in the Hydraulic Fracturing Test Site (HFTS) 2 experiment, we show that this assumption is fundamentally flawed. The HFTS 2 data have three unique observations that have not been commonly observed in other datasets: (1) Due to the influence of offset pad depletion, microseismic data show that hydraulic fractures from the child well can propagate over 3,000 ft into the depleted low-stress zone. (2) By comparing microseismic and horizontal fiber LF-DAS strain data, we observe that the microseismic event cloud does not necessarily reflect the created hydraulic fracture volume. Particularly, the extent of microseismic event clouds near heel stages is much shorter than what is shown with LF-DAS strain data. (3) Microseismic event magnitudes are larger in the depleted regions. Through geomechanical analysis, we demonstrate that the “bedding-plane-slip” model is likely the mechanism for microseismic generation during hydraulic fracturing. This model successfully explains the above field observations from the HFTS 2 experiment. We also provide a quantitative relationship connecting the microseismic event magnitude with fracture width increment and layer mechanical property contrast.

Summary Understanding fracture height growth can be of great significance to optimizing field development and improving recovery. The Hydraulic Fracturing Test Site 2 (HFTS-2) has provided a unique opportunity and an advanced data set to allow us to observe and understand fracture geometries rigorously. Low-frequency distributed acoustic sensing (LF-DAS) data from a vertical well in HFTS-2 showed three key observations: (i) excessive upward height growth (>1,000 ft) and limited downward growth of the hydraulic fractures during pumping, (ii) considerable additional upward fracture height growth (~300 ft) after well shut in, and (iii) very complex LF-DAS strain rate patterns for a small fiber-to-stage offset. Advanced geomechanical modeling was performed to simulate the hydraulic fracture propagation and the resulting strain responses in the vertical direction. The modeling results demonstrated asymmetric upward and downward fracture height growths as observed in HFTS-2 with a similar upward height growth rate. Simulated waterfall plots of vertical strain rate showed distinct patterns for different fiber-to-fracture distances. The upward-growing fracture tip can be clearly identified by the interfaces between compressing and extending zones. It was also found that the complex strain rate patterns observed in HFTS-2 for small fiber-stage offsets were not caused by the mechanical layering but possibly result from the simultaneous propagation of multiple hydraulic fractures at different rates. The simulation results improved the understating of the HFTS-2 LF-DAS data, and the simulated strain rate patterns could also serve as templates for fracture height interpretation from LF-DAS data in future.

Summary Multistage plug and perforation (plug-n-perf) fracturing is commonly used for horizontal well completion in unconventional reservoirs. Uniform distribution of proppant across all clusters in each stage has proved to be challenging with low viscosity slickwater owing to its limited transport capability. Computational fluid dynamics (CFD) has been used to model proppant transport in wellbore to improve perforation and fracturing design for achieving uniform proppant placement. However, traditional CFD modeling of a full-scale stage is computationally expensive, which limits its applicability in the completion design optimization. A new approach was developed in this paper to efficiently predict proppant placement along a multicluster stage based on a machine learning (ML) model trained with extensive CFD modeling results. Its high computational efficiency permits quick sensitivity analyses to optimize perforation and fracturing designs. The new approach was validated against full-stage CFD modeling results as well as post-treatment field diagnostics. Sensitivity analyses show that proppant inertia effect is a key factor affecting proppant placement in heel clusters with higher slurry flow rates, allowing more proppant carried to the toe owing to its higher density in comparison with fluid. Proppant settling allows bottom perforations to accept more proppant than top perforations. This gravitational effect is not negligible near the heel at high flow rates and becomes more dominant near toe clusters where the flow rate is reduced. Near-uniform proppant placement is achievable via perforation design optimization by taking advantage of these two key mechanisms controlling proppant transport in horizontal wellbores. It is demonstrated that in-line perforating designs with all perforations having the same orientation in each cluster or the entire stage, especially with perforations at the bottom or on the side of the wellbore, improve the proppant placement uniformity. However, it is recommended that the optimum perforation design should be identified case by case depending on specific input parameters. The ML-based model developed in this study has overcome some of the limitations from existing models in the literature and is able to provide quick and yet reliable solutions to proppant placement prediction and design optimization.

Cryogenic fracturing with liquid nitrogen (LN2) offers the benefits of reducing the water consumption and adverse environmental impacts induced by water-based fracturing, as well as potentially enhancing the fracture complexity. We performed a series of laboratory experiments to explore the key mechanisms governing the breakdown pressures of shale during cryogenic fracturing. In this study, cylindrical shale samples were pre-conditioned by exposing a borehole to low-temperature LN2 for a certain time period, and then, the samples were fractured using gaseous N2 under triaxial stress and a high reservoir temperature. The effects of various key parameters on the breakdown pressure were investigated, including the duration of the low-temperature LN2 treatment, the confining pressure, the reservoir temperature, and the direction of the shale bedding relative to the borehole axis. The results demonstrate that the injection of low-temperature LN2 as a pre-fracturing fluid into a borehole can significantly reduce the breakdown pressure of the shale during subsequent nitrogen fracturing. This reduction in breakdown pressure can be further intensified by increasing the duration of the LN2 pre-conditioning. Without LN2 pre-conditioning, the breakdown pressure initially increases and then decreases with increasing reservoir temperature. When LN2 pre-conditioning is applied, the breakdown pressure keeps decreasing with increasing reservoir temperature. As the confining pressure increased, the breakdown pressure increased linearly in the tests with and without LN2 pre-conditioning. The experimental results demonstrate that LN2 pre-conditioning before N2 fracturing is a promising waterless fracturing technique that reduces the breakdown pressure and enhances the fracture complexity.

Non-aqueous or gaseous stimulants are alternative working fluids to water for hydraulic fracturing in shale reservoirs, which offer advantages including conserving water, avoiding clay swelling and decreasing formation damage. Hence, it is crucial to understand fluid-driven fracture propagation and morphology in shale formations. In this research, we conduct fracturing experiments on shale samples with water, liquid carbon dioxide, and supercritical carbon dioxide to explore the effect of fluid characteristics and in situ stress on fracture propagation and morphology. Moreover, a numerical model that couples rock property heterogeneity, micro-scale damage and fluid flow was built to compare with experimental observations. Our results indicate that the competing roles between fluid viscosity and in situ stress determine fluid-driven fracture propagation and morphology during the fracturing process. From the macroscopic aspect, fluid-driven fractures propagate to the direction of maximum horizontal stress direction. From the microscopic aspect, low viscosity fluid easily penetrates into pore throats and creates branches and secondary fractures, which may deflect the main fracture and eventually form the fracture networks. Our results provide a new understanding of fluid-driven fracture propagation, which is beneficial to fracturing fluid selection and fracturing strategy optimization for shale gas hydraulic fracturing operations.

Hydraulic fracturing is a key method in the stimulation of shale gas reservoirs. Shale gas formations often contain natural fractures which are fluid-pressure sensitive and dilate in response to the inflation of the fracture, increasing fluid loss and slowing down and potentially prematurely arresting fracture propagation. Models typically assume 1-D single-porosity/permeability (Carter) leakoff perpendicular to the hydraulic fracture. However, the leakoff process in naturally fractured formations is considerably more complex. In this study, we present an hydraulic fracturing model based on the PKN-formalism which accommodates leakoff into a pressure-sensitive dual porosity medium. Proppant transport is accommodated by introducing empirical constitutive equations to determine the proppant distribution during the hydraulic fracturing treatment. The model is solved numerically and is validated against known small and large time asymptotic solutions. The model is capable of providing a rapid estimation of the morphology of hydraulic fractures in naturally fractured formations and the corresponding proppant distribution. The simulation results illustrate that the leakoff into a dual porosity medium, where fracture permeability is a strong function of applied fluid pressure, results in a reduced length of the propagating fracture due to the fugitive fluid leakoff from the fracture into the surrounding formation and that this in turn results in a reduced maximum width during the treatment. The ability to infuse proppants in fluid-driven fractures penetrating large distances from the injection wellbore is further limited by premature screen-out. This may compromise the ultimate efficiency of the final hydraulic fracture regarding gas recovery. Reduced propagation and premature screen-out are limited by low permeability and large spacing of the natural fractures. The presence of an existing network of natural fractures, including those adjacent to the hydraulic fracture that may become propped, aids in the recovery of the resource by reducing diffusion lengths of the hydrocarbon to the main fracture.

The residual opening of fluid-driven fractures is conditioned by proppant distribution and has a significant impact on fracture conductivity - a key parameter to determine fluid production rate and well performance. A 2D model follows the evolution of the residual aperture profile and conductivity of fractures partially/fully filled with proppant packs. The model accommodates the mechanical response of proppant packs in response to closure of arbitrarily rough fractures and the evolution of proppant embedment. The numerical model is validated against existing models and an analytic solution. Proppant may accumulate in a bank at the fracture base during slick water fracturing, and as hydraulic pressure is released, an arched zone forms at the top of the proppant bank as a result of partial closure of the overlaying unpropped fracture. The width and height of the arched zone decreases as the fluid pressure declines, and is further reduced where low concentrations of proppant fill the fracture or where the formation is highly compressible. This high-conductivity arch represents a preferential flow channel and significantly influences the distribution of fluid transport and overall fracture transmissivity. However, elevated compacting stresses and evolving proppant embedment at the top of the settled proppant bed reduce the aperture and diminish the effectiveness of this highly-conductive zone, with time. Two-dimensional analyses are performed on the fractures created by channel fracturing, showing that the open channels formed between proppant pillars dramatically improve fracture transmissivity if they are maintained throughout the lifetime of the fracture. However, for a fixed proppant pillar height, a large proppant pillar spacing results in the premature closure of the flow channels, while a small spacing narrows the existing channels. Such a model provides a rational means to design optimal distribution of the proppant pillars using deformation moduli of the host to control pillar deformation and flexural spans of the fracture wall.

Relative to water-based fluids, non-aqueous fracturing fluids have the potential to increase production, reduce water requirements, and to minimize environmental impacts. Since the viscosity of supercritical CO2 is one-tenth that of water, its density is close to that of water, and is capable of promoting sorptive rejection of methane, it is an attractive substitute for water in the extraction of shale gas and coalbed methane. The following defines a geomechanical model accommodating the interaction of fluid flow, adsorption-induced swelling stress, solid deformation and damage to quantify rock-gas interactions during supercritical CO2 fracturing for shale gas production. The architecture of the shale is accommodated that includes both pore- and micro-crack-based porosity. According to the microcrack model representing shales with low porosity, both analytical and numerical results show that the effective stress coefficient is much smaller than unity. We analyze the potential advantages of fracturing using supercritical CO2 including enhanced fracturing and fracture propagation, increased desorption of methane adsorbed in organic-rich portions of the shale and the potential for partial carbon sequestration. Rock-gas interactions include both the linear poroelastic response and the chemo-mechanical interaction due to sorption. Simulation results demonstrate that supercritical CO2 fracturing indeed has a lower fracture initiation pressure and a significantly lower breakdown pressure, as observed in experiments, and that fractures with greater complexity than those developed with liquid CO2 and water fracturing result. With increasing dynamic viscosity of the fracturing fluids, the predicted breakdown pressure also increases, consistent with experimental observations.

Conventional water-based fracturing treatments may not work well for many shale gas reservoirs. This is due to the fact that shale gas formations are much more sensitive to water because of the significant capillary effects and the potentially high contents of swelling clay, each of which may result in the impairment of productivity. As an alternative to water-based fluids, gaseous stimulants not only avoid this potential impairment in productivity, but also conserve water as a resource and may sequester greenhouse gases underground. However, experimental observations have shown that different fracturing fluids yield variations in the induced fracture. During the hydraulic fracturing process, fracturing fluids will penetrate into the borehole wall, and the evolution of the fracture(s) then results from the coupled phenomena of fluid flow, solid deformation and damage. To represent this, coupled models of rock damage mechanics and fluid flow for both slightly compressible fluids and CO2 are presented. We investigate the fracturing processes driven by pressurization of three kinds of fluids: water, viscous oil and supercritical CO2. Simulation results indicate that SC-CO2-based fracturing indeed has a lower breakdown pressure, as observed in experiments, and may develop fractures with greater complexity than those developed with water-based and oil-based fracturing. We explore the relation between the breakdown pressure to both the dynamic viscosity and the interfacial tension of the fracturing fluids. Modeling demonstrates an increase in the breakdown pressure with an increase both in the dynamic viscosity and in the interfacial tension, consistent with experimental observations.

Restimulation of previously hydraulically-fractured wells can restore productivity to near original levels. Understanding the stress state resulting from the original hydraulic fracturing and subsequent depletion is vital for a successful refracuring treatment. The stress obliquity in the vicinity of the wellbore, due to production from a previously introduced hydraulic fracture, promotes a new concept – that of altered-stress refracuring which allows fractures to propagate into previously unstimulated or understimulated areas and therefore enhancing recovery. In this study, a coupled poromechanical model is used to define stress redistribution and to define optimal refrac timing as defined by maximizing the size of the stress reversal region. Key factors include the time dependency of the stress reorientation, the threshold stress ratio σ h max / σ h min and the influences of permeability anisotropy/heterogeneity, pressure drawdown and rock-fluid properties. The results show that stress reorientation develops immediately as the reservoir begins to produce. This stress reversal region extends to a maximum extent before retreating as the direction of the maximum principal stress gradually returns to the initial state. The optimal refrac timing and the size of the stress reversal region are positively correlated with pressure drawdown and Biot coefficient, negatively correlated with stress ratio σ h max / σ h min ratio and Poisson’s ratio and ambiguously correlated with permeability anisotropy. Permeability magnitude and porosity have no influence on the size of the resulting zone but are negatively and positively correlated to the timing, respectively. Permeability heterogeneity has no influence on the size nor the timing. Coupled fluid flow and damage-mechanics simulations follow fracture propagation under the effect of stress redistribution during refracturing treatments. These results define the evolving path of secondary refracture as it extends perpendicular to the initial hydrofracture and ultimately turns parallel to the hydrofracture as it extends beyond the stress-reversal region. This discrete model confirms the broader findings of the continuum model.

Hydraulic fracturing operations have been widely used to enhance reservoir permeability during the extraction of oil, shale gas and tight sandstone gas. Recently, non-aqueous fracturing fluids, such as supercritical carbon dioxide and nitrogen, have been proposed as the potential fracturing fluid candidates due to their advantages of decreasing formation damage, conserving water resources and avoiding injection-induced seismicity. However, the breakdown process of the non-aqueous fracturing process and the breakdown mechanism remain mysteries. In this research, we report a series of hydraulic fracturing experiments with low-permeable shale samples and high-permeable sandstone samples by water, supercritical carbon dioxide, and nitrogen gas under different injection rates. Moreover, we use a coupled fluid-rock interaction model to visualize the distribution of fluid pressure near the injection borehole during hydraulic fracturing before the sample breakdown. The results indicate that the fluid infiltration decreases the breakdown pressure by increasing the pore pressure and decreasing the effective stress, especially for high-permeable sandstone rock and low-viscosity fracturing fluid. Also, breakdown pressure increases with increasing fluid injection rate, and the conventional Hubbert-Willis and Haimson-Fairhust equations can still effectively determine the upper and lower limits in predicting the breakdown pressure. Our results suggest that lower viscosity fluid with a lower injection rate leads to a lower breakdown pressure, which is suitable to be implemented for the reservoirs with high intrinsic permeability.

A model accommodating multi-scale pores containing kerogen within an inorganic matrix is used to explore the complex multi-mechanistic transport mechanisms of shale gas reservoirs. These include the complex evolution of pressure, diffusion and flow within both kerogen and inorganic components and their interaction with effective stresses. A general poromechanical model is proposed considering desorption and molecular diffusion in the kerogen, viscous flow in the inorganic matrix and fracture system, and composite deformation of the triple porosity assemblage. The model is verified by history matching against field data for gas production rate. The simulation results indicate that the pattern of gas flow is sequential during gas depletion – pressure first declines in the fracture, followed by the inorganic phase and then in the kerogen. The evolution of permeability is pressure dependent and the evolution of pressure is closely related to the intrinsic gas diffusion coefficient in the kerogen, inorganic matrix intrinsic permeability and fracture intrinsic permeability. A series of sensitivity analyses are completed to define key parameters affecting gas production. The study shows that dominant influence of the fracture network in acting as the main permeable conduit. The intrinsic permeability and porosity of the fracture have a positive correlation with gas production, while fracture spacing has a negative correlation to gas production. Kerogen also plays a critical role in gas production for shale reservoirs with higher total organic carbon. The enhancement of inorganic matrix permeability and gas diffusion coefficient in kerogen could efficiently guarantee a long-term gas production with a higher rate.

Dynamic gas fracturing is a well stimulation technique that is able to create multiple fractures emanating from a wellbore. It operates by pressurizing at rise-times and peak pressures intermediate between conventional hydraulic fracturing and explosive fracturing. Two consecutive processes operate during this fracturing process: (i) generation and propagation of a dynamic stress wave that overpowers the static stress field and creates multiple radial fractures around borehole, followed by (ii) quasi-static pressurization and further extension of those starter-fractures by the expanding gas. Dynamic analysis is first performed to follow the evolution of the stress wave propagating from the borehole. The radial (r) distribution of peak tensile hoop-stress diminishes as 1 / r α with the power exponent ( α ) asymptoting to 2 as the loading rate decreases. This rapid attenuation generally limits the length of the body-wave-generated radial fractures to several borehole radii. The gas-loading of the borehole wall is followed by permeation of the gas pressure into the newly created radial fractures. Linear elastic fracture mechanics (LEFM) perturbation analysis shows that a regular distribution of multiple radial starter-cracks will preferentially propagate the longer cracks at the expense of the shorter cracks – that will arrest and snap-shut. This system naturally selects of the order of six dominant fractures that may grow in unison until the in situ stress field reasserts control as the fractures lengthen. This restricts the maximum number of the dominant fractures to of the order of six at the conclusion of treatment - a common observation in both in situ and laboratory experiments.

Gases show promise as an alternative to water-based fracturing fluids because they are non-damaging to water-sensitive formations, show some potential to create complex fracture networks, flow back to the well rapidly after treatment, and deliver some environmental benefits. However, the ability of gases to transport proppant has been questioned due to their relatively low viscosity and density. The fracturing then proppant-carrying capacity of various gases is investigated to determine the form and function of the emplaced proppant pack. First, fracture propagation and proppant transport driven by several commonly-used pure gases (CO2, liquified petroleum gas, ethane, and N2) is simulated and compared against slickwater fracturing – generally identifying inferior reach and functionality. Several methods are then investigated to improve the proppant-carrying capacity of the pure gases. Results show that, compared with slickwater, gases create shorter and narrower fractures and carry proppant shorter distances due to their lower viscosity and faster leak-off. Among the gases examined in this study, liquified petroleum gas and CO2 return the deepest proppant penetration along the fracture, followed by ethane, and with N2 unable to carry proppant into the fracture due to the resulting narrow fracture. However, elevating injection rate of gases could improve their fracture-inducing potential and proppant-transport capability to a level competitive with that of slickwater. A near-uniform proppant distribution may be achieved by using a gelled gas, with an approximately two order-of-magnitude enhancement in viscosity, or a foam-based fluid with a high quality. The fracture length may also be extended by limiting leak-off due to the increased viscosity. Moreover, ultra-light-weight proppants (ULWPs) perform better with gases than commonly-used sands in terms of uniformity of distribution, due to decelerated proppant settling. Well performance is improved significantly by fracturing with gelled gases or foams instead of pure gases or by pumping ULWPs instead of normal sands.

Acoustic emission (AE) in coal is anisotropic. In this paper, we investigate the microstructure-related scale effect on the anisotropic AE feature in coal at unconfined loading process. A series of coal specimens were processed with diameters of 25 mm, 38 mm, 50 mm, and 75 mm (height to diameter ratio of 2) and anisotropic angles of 0°, 15°, 30°, 45°, 60°, and 90°. The cumulative AE counts and energy dissipation increase with the specimen size, while the energy dissipation per AE count behaves in the opposite way. This may result from the increasing amount of both preexisting discontinuities and cracks (volume/number) needed for specimen failure and the lower energy dissipation AE counts generated by them. The effect of microstructures on the anisotropies of AE weakens with the increasing specimen size. The TRFD and its anisotropy reduce as the specimen size increases, and the reduction of fractal dimension is most pronounced at the anisotropic angle of 45°. The correlation between TRFD and cumulative AE energy in the specimens with different sizes are separately consistent with the negative exponential equation proposed by Xie and Pariseau. With the specimen size gain, the reduction of the TRFD weakens with the increasing amount of cumulative absolute AE energy.