Impact of Fracture Conductivity and Matrix Relative Permeability on Load Recovery

Abstract

Abstract The cleanup of fracturing treatments in tight-gas sands has been a subject of focus because of the disappointing performance of many treatments. A fit-for-purpose fracture-cleanup simulator was used to explore factors that might impact cleanup. The simulator includes physics of flow and chemical physics of overall complexity that cannot be effectively studied in the laboratory. Numerical simulations were conducted to study the impact of dimensionless fracture conductivity (Fcd), matrix relative-permeability quality for gas and water, and shut-in times on load recovery after a treatment. Simulation results included the impact of these various factors on gas production rates, total load-recovery volumes, and fracturing-fluid dilution profiles. Three sets of relative permeabilities (rel-perms), four Fcds, and various shut-in times were studied. Numerical simulations demonstrated the complex interactions of fracture conductivity, matrix rel-perm quality, and shut-in on load recovery and gas productivity. The results gave insight into preferred procedures and practices for cleaning up fracturing treatments in tight-gas sands. Gas-production rates and load recoveries were strongly affected by conductivity and matrix rel-perm quality. Rel-perm sets with a crossover at 10% of rel-perm exhibited good load recovery at high Fcd and poor recovery at low Fcd, with gas-production rates dominated by Fcd alone. A crossover at 0.1% of rel-perm exhibited slow recovery of fluid and slow improvement in gas rates, with peak rates not occurring for months. Extended shut-in times reduced load recoveries for high-quality rel-perms the most. Fracturing-fluid dilution profiles were observed to be a function of both Fcd and rel-perm quality. This work demonstrates that an arbitrary choice of matrix rel-perms for water and gas can result in prediction of either good or poor load recovery, depending on the quality of rel-perms chosen. In fact, variation in rel-perm quality from reservoir to reservoir might help explain contradictory observations on best practices regarding shut-in times.