Numerical experiments in the simulation of enhanced oil recovery from a porous formation

A numerical reservoir simulation model for the study of enhanced oil recovery (EOR) from a porous formation has been presented. The resistance to oil movement arises from viscous forces in the fluid phase as well as surface tension. Viscous forces can be lowered by hot water injection into the formation or by raising the formation temperature. These methods have been numerically analyzed in the present study. The role of the operating parameters such as the injection pressure and temperature on oil recovery has been reported. Displacement of oil by water is clearly brought out by the saturation and the temperature profiles. The numerical solution of the EOR problem experiences growth of errors during long time integration, particularly on large regions. Possible reasons are scatter in the constitutive relationship data, inexact outflow boundary condition and round-off errors in the calculation of the matrix inverse. The nature of these errors has been addressed in the present work. To solve the computationally intensive field-scale problems, two domain decomposition algorithms namely, Schwarz's and Uzawa's algorithms have been evaluated. Results show that oil recovery can be improved when the formation temperature is higher, or the injection temperature and pressure are raised. Adverse results can however be obtained when the injection temperature exceeds a critical value. Optimum conditions prevail when the speed of the oil-water interface is matched with that of the thermal front. As a computational too[, the domain decomposition algorithms are conditionally seen to improve the numerical performance of the oil recovery codes. (C) 2001 editions scientifiques et medicales Elsevier SAS.