Micromechanics of proppant agglomeration during settling in hydraulic fractures

Abstract

The micromechanics of proppant settling in quiescent fluid in a rough and relatively narrow rock fracture is investigated. The study focuses on particleparticle and particle-wall interactions in a dense-phase particle settling. The study used a coupled discrete element method and computational fluid dynamics (DEM-CFD) code because DEM-CFD is the most suitable computational method for modeling the frequent interactions of dense assembly of rigid particles and enables modeling of two-way solid-fluid interactions. Due to frequent particleparticle interactions of grains submerged in fluids, the particle interaction model in DEM is improved by the incorporation of the effects of lubrication due to a layer of fluid surrounding particles. Results of the numerical study are compared to previous experimental and theoretical relationships. The findings of the study highlight conditions that lead to proppant aggregation due to the fluid viscosity and fracture width in relation to particle diameter ratio. In the light of the DEM-CFD results, it was found that published relationships are inadequate in describing the settling rates for proppant in a rough and narrow hydraulic fracture and high fluid viscosity. Micromechanical particle interactions during settling and erratic upward and fluid counter-flow may cause proppant trajectories that are not always in the direction of gravity in a rough fracture resulting in clogging of the fracture or forming faster settling particle agglomerates. The maximum packing density 0.3-0.4 (3.9-5.9 lbs/gal) of proppant in a narrow and rough hydraulic fracture was obtained in this study, which is lower than the usually assumed one of 0.5 (9.8 lbs/gal) for any given fracture roughness.Keywords DEM-CFD Á Lubrication Á Fluid-solid coupling Á Dense-phase flow Á Proppant settling Á Hydraulic fracture List of symbols

The Effective Transmissivity of a Plane‐Walled Fracture With Circular Cylindrical Obstacles

“…Unfortunately, the final distributions of proppant grains are virtually impossible to image nowadays, and direct numerical simulations of suspension flows in narrow apertures have just begun to emerge (Shiozawa & McClure, 2016;Tomac & Gutierrez, 2015). In general, the processes of proppant transport and settlement are complex and dependent on a large number of parameters including proppant density, fracking fluid properties, pumping rate, and fracture geometry.…”

Early-flowback tracer signals for fracture characterization in an EGS developed in deep crystalline and sedimentary formations: a parametric study

“…Thus the fact that dispersivity has a stronger effect on SW tracer signals than the other parameters, in conjunction with relatively narrow ranges for parameters like proppant-packing porosity, GPF thickness or WF aperture, may allow to at least roughly estimate dispersivity without knowing the exact values of other parameters. A possible workaround to parameter interplay is the use of semi-empirical correlations between hydrogeological parameters to constrain the value of either dispersivity as a function of fracture aperture and porosity (Horne and Rodriguez, 1981;Fossum and Horne, 1982;Pruess and Bodvarsson, 1984), or porosity as a function of permeability (Guimerà and Carrera, 1997), or proppantpacking dispersivity as a function of proppant treatment protocols (Tomac and Gutierrez, 2015). Alternatively, it may be possible to make a priori assumptions on the magnitude of the Peclet number (Pe) of the overall transport scenario (with Pe defined as the ratio between tracer-invaded fracture radius and longitudinal dispersivity).…”

Numerical simulation of particulate suspension transport and permeability impairment in an actual rough fracture under normal stresses

“…Considerable research based on numerical simulation methods has been conducted. The numerical simulations can be divided into several types, such as the particle quasi‐fluid method, CFD‐DEM method (including LBM‐DEM, LES‐DEM, and DNS‐IB), particle tracking method, and Brownian dynamics simulation . The existing studies demonstrated that the numerical simulation is an effective means to study the particulate suspensions transport and permeability damage to the porous media.…”

“…The numerical simulations can be divided into several types, such as the particle quasi-fluid method, CFD-DEM method (including LBM-DEM, LES-DEM, and DNS-IB), particle tracking method, and Brownian dynamics simulation. [12][13][14][15][16][17][18] The existing studies demonstrated that the numerical simulation is an effective means to study the particulate suspensions transport and permeability damage to the porous media. The particle size-exclusion mechanism in a smooth fracture with a constrict has been extensively studied, and the general conclusion of these studies is that mechanical bridging is more prone to occur when the particle size is larger than one-third of the fracture width.…”