Grid discretization effects in the presence of realistic geological heterogeneities

Grid discretization has an effect on reservoir characterization. This effect on reservoir description is exaggerated in the presence of realistic reservoir heterogeneity. The discretization and scaling effects in geological modeling, in turn, are also reflected in reservoir flow simulation. This study describes procedures to understand and quantitatively assess grid discretization effects. Different scaling techniques are applied and the results are compared. Also described is the uncertainty in reservoir flow modeling for different types of fluids flowing in the reservoir. In this study, different ranking techniques as a fast way to assess the uncertainty in reservoir modeling are also presented.; The results of this study show that the magnitude of the grid discretization effects depends on the porosity-permeability properties, and the nature of fluids flowing in the reservoir. Recovery factors and breakthrough times increase as the number of cells increases. For the homogeneous reservoir case, the recovery factors for the coarsest grid used in this study, were 4–5% lower compared to the finest grid. The breakthrough times were 23–30% lower for the coarsest grid compared to the finest grid. For the reservoir with downscaled porosity-permeability, recovery factors were 6–8% lower and breakthrough times were 25–36% lower for the coarsest grid than for the finest grid. For the reservoir with upscaled porosity-permeability, based on the mean values of 100 realizations, the recovery factors were 3–7% lower and the breakthrough times were 5–9% lower for the coarsest grid than for the finest grid. Uncertainty in geological modeling contributes more to the total uncertainty than does the grid discretization. Recovery factors and breakthrough times obtained from a flow simulation using a very coarse grid can be used as fair to good ranking parameters to assess the uncertainty in the reservoir flow simulation.