The shear-induced migration of dense suspensions with continuously distributed (polydisperse) particle sizes is investigated in planar channel flows for the first time. A coupled lattice Boltzmann–discrete element method numerical framework is employed and validated against benchmark experimental results of bulk shear-induced migration and segregation by particle size. Distinct dependence on the particle size distribution is shown in the flowing (non-plugged) regime (where the bulk solid volume fraction, $\bar{\phi}$ , $\leq 0.3$ ) resulting from a dual dependence on the particle self-diffusivity and local rheology imposed by the particle pressure gradient. Close agreement between statistically equivalent bidisperse and polydisperse suspensions suggests that the bulk migration, and by extension the shear-induced diffusivity, is completely characterised by the first three statistical moments of the particle size distribution. For both bidisperse and polydisperse suspensions in the plugging regime, $\bar {\phi }\geq 0.4$ , the smallest particles preferentially form the plugs, causing the largest particles to segregate to the channel walls. This effect is accentuated as $\bar {\phi }$ increases and has not been reported in the literature hitherto. It is proposed that smaller particles preferentially form the plugs due to their higher shear-rate fluctuations, which completely dominate particle motion near the plug where the mean shear rate vanishes. Finally, increasing inertia causes a greater bulk migration towards the channel walls, but increased mid-plane migration for the largest particles due to the dependence of the particle self-diffusivity on the particle Reynolds number. As $\bar {\phi }$ increases shear-induced migration dominates and these inertial effects disappear, as does dependence on the particle size distribution.

Optimal proppant placement is critical to maintaining productivity from stress-sensitive reservoirs, in which gas conductivity depends on the connectivity of the network of secondary fractures to the wellbore. In a colloquial sense, this research places micro-proppants in induced and natural fractures, shows how they are excluded from reaching far into the reservoir, and describes which sizes of proppants this occurs for. Micromechanical modelling of a hydraulic fracturing fluid, in which the hydrodynamics between the fluid and solid phases are fully resolved, is achieved via the lattice Boltzmann method (LBM) for fluids coupled with the discrete element method (DEM) for particles. It is shown that proppant transport along the primary hydraulic fracture channel is strongly inhibited by leak-off into the secondary fracture system. This leak-off is strongly affected by the migration of particles across the fracture width, which in turn is a function of reservoir and treatment properties. A novel numerical approach is proposed for predicting proppant transport through the secondary fracture system, with far-reaching applications to porous media particulate transport.

Low permeability, naturally fractured reservoirs such as coal seam gas (CSG, coalbed methane or CBM) and shale gas reservoirs generally require well stimulation to achieve economic production rates. Coupling hydraulic fracturing and micro-proppant or graded particle injections (GPI) can be a means to maximise hydrocarbon recovery from these tight, naturally fractured reservoirs, by maintaining or improving cleat or natural fracture conductivity. This paper presents a summary of the National Energy Resources Australia (NERA) project "Converting tight contingent CSG resources: Application of graded particle injection in CSG stimulation" - which assessed the application of micro-proppants, providing guidance on key considerations for GPI application to CSG reservoirs. Over the last decade, laboratory research and modelling have shown the benefits of the application of GPI to keep pre-existing natural fractures and induced fractures open during production of coal reservoirs with pressure dependent permeability (PDP). Laboratory studies, within this study, provide further insight on potential mechanisms and key factors, including proppant size and optimum concentration, which contribute to the success of a micro-proppant placement. Accompanying numerical modelling studies will be presented that describe the likely fluidized behaviour of micro-proppants (e.g., straining models, electrostatic effects, and ‘screen out’ prediction). This paper outlines the necessary reservoir characterization, treatment considerations, and key numerical modelling inputs necessary for the design, execution, and evaluation of GPI treatments, whether performed standalone or in conjunction with hydraulic fracturing treatments. It also provides insight on the practical application of GPI efficiently into fracturing operations, minimizing natural and hydraulic fracturing damage effects, thereby maximizing potential production enhancement for coals, shales and other tight, naturally fractured reservoirs exhibiting pressure-dependent permeability effects.

No abstract

In fluid-driven particulate flows through channels, jamming occurs when static bridges of particles form as the channel becomes narrow relative to the particle size. For micro-sized particles, however, the significance of electrostatics relative to hydrodynamics must be considered. The present work develops a numerical framework based on the inclusion of DLVO theory in fully-resolved lattice Boltzmann method-discrete element method (LBM-DEM) simulations. Strong dependence of jamming on the ionic strength of the fluid medium is demonstrated. Further, continuous functions are fit to the probabilistic jamming data, representing a novel approach to predicting the onset of jamming compared to existing empirical models.