Predicting fluid flow behavior in naturally fractured reservoirs is a challenging area in petroleum engineering. Two classes of models used to describe flow and transport phenomena in fracture reservoirs are discrete and continuum (i.e. dual porosity) models. The discrete model is appealing from a modeling point of view, but the huge computational demand and burden in porting the fractures into the computational grid are its shortcomings. On the other hand, the diagonal representation of permeability, which is customarily used in a dual porosity model, is valid only for the cases where fractures are parallel to one of the principal axes. This assumption cannot adequately describe flow characteristics where there is variation in fracture spacing, length, and orientation. To overcome this shortcoming, the principle of the full permeability tensor in the discrete fracture network can be incorporated into the dual porosity model. Hence, the dual porosity model can retain the real fracture system characteristics.
Expelling oil from matrix blocks into fractures based on water imbibition leaves behind significant amounts of oil in the matrix blocks in the form of residual oil. Chemical flooding is an excellent means to expel more oil from the matrix into fractures. A fully implicit parallel, compositional chemical dual porosity model in conjunction with full permeability tensor for the fracture system has been developed. The model is capable of simulating large-scale chemical flooding processes. The matrix blocks are discretized into both rectangular rings and vertical layers to offer a better resolution of transient flow.
The developed model was successfully verified against the UTCHEM simulator. Results show excellent agreements for a variety of flooding processes. Full permeability tensor results were also verified with a discrete fracture simulator.
This study leads to a conclusion that the full permeability tensor representation is essential to accurately simulate fluid flow in heterogeneous and anisotropic fracture systems.