This dissertation examines -by means of the discrete element method- the mechanisms that govern hydraulic fracturing in poorly consolidated formations. The motivation to take on this project was the widely reported discrepancies between values predicted by conventional hydraulic fracturing simulators and the values encountered in the field, in formations of this type. The proposed numerical framework integrates dissimilar methodologies for the simulation of the solid and fluid components. The solid material -the rock- was represented as a collection of discrete particles that interact with each other by linear, spring-like contacts; this method is known as the Discrete Element Method (DEM). Meanwhile, the fluid was modeled by finite-difference discretization of the equations of conservation, applied to fluid flow in porous media (i.e. Navier-Stokes equations in porous media). A coupling mechanism conveys information about the interactions between the fluid and solid components.

The results of this study suggest that conventional simulation models ignore mechanisms that control hydraulic fracture propagation in poorly consolidated formations. Principally, the assumption of linear elastic behavior and the normal displacement of the newly cracked surfaces are not always the dominant features observed in DEM simulations. Instead, minor adjustments of dislodged particles yield zones of high concentration of stresses and posterior extension of the fracture. Plots of injection vs. volumetric strain exhibit a multi-linear and sometimes non-linear shape. Moreover, particle readjustments occur in tangential directions (i.e. shear cracking) on a very regular basis. The importance of shear cracking is commonly neglected in the better-known hydraulic fracturing models.