This paper describes research in modeling unstable displacement processes (especially related to the application of CO2 flooding) with the ultimate goal of predicting recoveries in unstable field scale displacements. The research has focused on understanding and quantifying the effects of process parameters, such as reservoir heterogeneity, phase behavior, fluid properties, and solvent/water injection ratio, on the stability of displacements. Numerical simulation using very fine grids was used to obtain these results. These fine grid simulations were compared to simulations using the more standard models based on the mixing parameter concept of Todd and Longstaff and the theory of Koval.
This work produced the following observations:
In secondary miscible displacements in heterogeneous reservoirs, the value of the mixing parameter can be strongly dependent on mobility ratio.
In displacements which are not first contact miscible, the shockfront mobility ratio is generally less than the fluid viscosity ratio. This effect tends to stabilize the displacement front. The mixing parameter model can, in some cases, be used to model such processes, and the value of the mixing parameter can be estimated from knowledge of the mobility ratio across the displacement front.
In general, for displacements which are not first contact miscible, the mixing parameter model is inadequate due to its inability to account for thermodynamic nonequilibrium between zones of oil contacted and uncontacted by solvent.
In tertiary displacements the frontal mobility ratios, as computed from a one-dimensional solution, are in general not sufficient to compute the stability of the displacement.