Grain-Scale Modeling of Lost Circulation Materials and Application to Reservoir Simulation

Abstract

We present a grain-scale model of granular particle transport around a single fracture to simulate the fracture clogging mechanism of lost-circulation materials (LCM). These simulations also provide an estimate of the permeability reduction due to fracture clogging. We use 2D Navier-Stokes equations in conjunction with an Immersed Boundary Method (IBM) to model the particle-fluid-fracture system. The fracture aperture is kept constant at 0.75 mm, and three particle sizes of 0.18 mm, 0.22 mm, and 0.33 mm are simulated to examine the effect of LCM particle diameter on clogging capacity. The Hertz-Mindlin model is used to simulate normal and frictional contact forces between particles. Simulations with holding pressures in the range of 0.15 MPa to 1.5 MPa show the chief fracture-clogging mechanism of the LCM to be via formation of unstable bridges at the fracture entry point. Our simulations show the effective permeabilities of the clogged fracture to be 870 mD, 536 mD, and 77 mD, for particle sizes of 0.18 mm, 0.22 mm, and 0.33 mm, respectively. Our simulations also show that increasing the LCM particle concentration in the fluid can improve fracture clogging. To assess the impact of LCM on fluid flow at the reservoir scale during geothermal drilling, we use a modified version of the fully compositional reservoir simulator TOUGH+RealGasBrine (T+RGB). First, we added the ability to represent drilling muds in the simulator, including the flow and thermal properties. Then, we used parametric functions to apply the permeability reductions computed by the grain-scale model to regions of the reservoir infiltrated by LCM-containing drilling muds. We created several hypothetical geological models representative of vertical wells penetrating geothermal systems, with fluid-entry zones represented as horizontal zones of fractured rock within lower-permeability porous media layers. For each geological model, we compare fluid flow into the formation, fluid loss, and the pressure reduction in the wellbore for scenarios with and without LCM of various particle sizes. The reservoir-scale simulations show that the predicted permeability reductions are sufficient and effective for reducing fluid flow into high-permeability fluid-entry zones and maintaining wellbore pressures, and thus preserving circulation within the wellbore.