- https://www.onepetro.org/conference-paper/SPE-178930-MS
- doi:10.2118/178930-MS
Abstract — The infiltration of fines into a reservoir during drilling or water reinjection and the accompanying production decline or loss of injectivity, are long-standing problems in the petroleum industry. Here, we propose a methodology that combines two different pore-scale models to better quantify formation damage. Further we validate the proposed model with core flood experiments. We first perform an experimental study of suspension flow into sintered glass bead plugs and measure changes in porosity and permeability. Specifically, glass bead suspensions of known mono- and poly- disperse combinations of particle sizes, concentrations, and flowrates are flooded through the core plugs, while keeping the total invaded particle volume constant. The resulting changes in porosity and permeability are quantified using a CT scanner and pressure transducers, respectively. We then establish two different numerical models (each focusing on a different mechanism/length scale) to predict permeability reduction. The first model takes a pore-scale approach that models straining of larger particles through the pore structures extracted from X-ray tomographic images of rock and grain pack samples from first principles. The detailed pore structure output from the first model is used as an input in the second model, which is a network model. This pore network model simulates permeability impairment caused by both large and small particles deposition in porous media. Forces exerted on small particles include hydraulic drag, gravity, buoyancy, electric double layer, and Van der Waals. Particle trajectories in a converging- diverging pore throat are calculated dynamically. We incorporate surface roughness and particle-surface interaction to determine particle detachment and attachment. Pore throat structure and hydraulic conductivities are updated dynamically to account for the effect of previously deposited particles. We finally compared the experimental results to simulation predictions and found that the combined pore- scale model is capable of predicting the porosity of the invaded core only in the deeper regions of the core.