(575d) Compositional Streamline Simulator with Gravity and Compressibility Effects

Field scale compositional simulations are routinely performed to study the performance of miscible/immiscible gas flooding and water-alternating-gas (WAG) injections in the oil recovery processes. The computational time is very high for these simulations in conventional finite difference reservoir simulators where the time-step size is constrained by the CFL conditions. To minimize computational time large grid blocks are used to represent the reservoir which leads to high numerical dispersion and inaccurate results. To overcome these problems, streamline methods are being developed in which fluid is transported along the streamlines and larger time steps and high speed up factors can be achieved as compared to finite difference simulations.

In this work a 3-dimensional four-phase (water, oil, gas, and second liquid phase) compositional streamline simulator is developed to study gas injections and WAG processes in a quarter five-spot pattern. 3-D multi-component material balance equations are decomposed into 1-D equations along the streamlines using the streamline time of flight as the spatial coordinate. Pressure field is solved in the conventional finite difference manner and streamlines are traced from injector to producer. Gravity effects are added using operator splitting technique to account for the gravity segregation due to density differences which is very prominent in gas displacements. This simulator can also handle the formation and flow of the fourth phase (second liquid phase) which is observed in CO2 injections into certain types of oils under specific conditions. Compressibility effects are included in the component transport equation by appropriately adding a source term that accounts for the changes in pressure field and volume changes on mixing. Different forms of higher order TVD schemes are implemented to construct the accurate numerical solution along the streamlines by reducing the impact of numerical dispersion.

WAG and gasflooding are simulated in a quarter 5-spot pattern with a vertical injection well and a horizontal production well. The injectant composition is varied. Different types of permeability heterogeneities are generated by changing the correlation length and variance of the permeability field. Effect of length of the horizontal production well, solvent composition, slug size, WAG ratio, correlation length and permeability variance on sweep and recovery is computed. Simulation results and computation time are compared with those of a standard finite-difference simulator.