The modern hydraulic fracturing techniques employed in unconventional resource extraction usually require multiple fracturing stages. Each stage typically involves the simultaneous creation of several fractures through a horizontal well. During this process, a significant volume of proppant—often in the order of millions of pounds per well—is injected alongside a fluid to maintain conductivity after closure. Despite the crucial role that proppant plays in enhancing well productivity, basic models are often utilized to estimate its placement within fractures. Such models may indicate propped or effective fracture lengths that are 100% to 300% longer than the lengths derived from production data.
One prevalent assumption in proppant transport models is that the average velocity of proppants due to pressure-driven flow aligns with the average velocity of the carrier fluid. Meanwhile, the calculation for settling velocity typically follows Stokes' law. However, to accurately predict proppant placement within fractures, it is essential to consider numerous effects that these assumptions overlook.
This study investigates the dynamics of particles moving between fracture walls in a fluid medium through simulations employing a coupled computational fluid dynamics and discrete element method (CFD-DEM). This approach meticulously addresses both particle and fluid behaviors. Simulation results reveal that the proppant concentration and the ratio of the proppant diameter to the fracture width are pivotal in determining the relative velocities of proppant and fluid interaction.
In smaller fractures, proppant settling velocity was evaluated to comprehensively understand how various independent variables, including concentration, affect proppant dynamics. Surprisingly, the simulations indicated that higher concentrations lead to an increased average settling velocity. This observation stands in stark contrast to much of the existing literature, which generally suggests that increased concentration results in reduced settling velocity. Notably, this discrepancy arises from the absence of displacement-driven counter-current fluid flow. These findings underscore the complexities surrounding proppant settling processes in hydraulic fractures.
From the direct numerical simulations conducted, a proppant transport model was formulated, which also incorporates existing particle settling velocity correlations. This model has been integrated into a sophisticated three-dimensional hydraulic fracturing simulator. The simulator unites the geomechanics of fractures with fluid dynamics and proppant transport, allowing for a precise determination of fracture geometry and proppant distribution. By varying two essential engineering fracture design parameters—namely, injection rate and proppant diameter—it is possible to comprehend their relative significance and optimize treatment tailored to specific applications.
Adjacent natural fractures found in unconventional reservoirs can markedly enhance well productivity. As proppant is conveyed through a hydraulic fracture, the presence of a widened natural fracture may create a fluid-accepting branch, leading to proppant entry. Using the CFD-DEM framework, researchers determined the proportion of proppant entering a branch at steady state, described by a dimensionless "particle transport coefficient" normalized by the amount of fluid reaching that branch. Factors such as Reynolds number at the inlet, branch aperture, the orientation angle between the main slot and branch, and particle size and concentration all impact this transport coefficient.
In certain scenarios, a distinct physical phenomenon can hinder particle transport into a branch—the development of a stable particle bridge that blocks further entry of proppants. This bridge formation was observed in various simulation cases. The complete set of equations used in a three-dimensional formulation for rectangular displacement discontinuity elements has enabled researchers to assess the width distribution of hydraulic fractures and dilated natural fractures. These widths have been analyzed across different combinations of stress anisotropy, net pressure, and hydraulic fracture dimensions.
It was revealed that, in most cases examined, natural fracture dilation is nonexistent. For those scenarios where dilation is noted, the efficiency of proppant transport associated with the natural fracture width is significantly less than one, and in numerous instances, it is zero due to size exclusion phenomena. Interestingly, neither the location nor the orientation of the natural fractures substantially influences their width. However, both their length and the elastic moduli of the rock notably affect width variations.
As hydraulic fracturing gains momentum, understanding the efficient delivery and placement of proppant in fractures has become increasingly crucial. Fracturing connected with weak formation bedding planes leads to unique fracture propagation patterns. While these patterns can impede vertical fracture growth, they may also encourage horizontal propagation along pre-existing bedding planes. Bedding planes introduce additional paths for hydrocarbon flow during recovery, which facilitates proppant movement along the horizontal sections of fractures. The effective ability to deliver proppants to these bedding planes is intricately linked to effective fracture height and overall well performance.
This paper delves into the aspects of horizontal movement and delivery of proppants, along with their distribution across both vertical and horizontal fractures, through experimental evaluations.
Proppant size emerges as a pivotal determinant when assessing uplift and settling velocities. Smaller proppants can travel deeper into fractures due to their slower settling rates, which are less influenced by gravitational forces. A fine mesh is highly recommended for achieving effective proppant delivery to horizontal bedding planes. Conversely, larger proppants necessitate a critical bed height to infiltrate the bedding plane situated above the vertical fractures. Increased injection rates enhance proppant delivery to bedding planes in inverted T-shaped geometries, as this creates a more expansive area for particle settling.
However, excessively vigorous turbulent flows generated by certain inlet configurations can prevent proppant settling and migration into the bedding planes. Continuous mixing and agitation further disrupt settling processes, restricting additional proppant coverage in these areas. While the inlet design exhibited negligible influence in T-shaped fractures, introducing more injection points significantly improved proppant placement within bedding planes. Interestingly, directly injecting into bedding planes did not substantially alter outcomes in T-shaped scenarios. Ultimately, only a marginal percentage of proppant is capable of infiltrating bedding planes across various T-shaped and inverted T-shaped fracture geometries.
For more information regarding AnYiCheng, and to learn more about hydraulic fracturing proppant, do not hesitate to contact us for expert consultations!
Comments
0