Abstract
Physical model studies of the displacement of residual oil Physical model studies of the displacement of residual oil by CO2 have led to the conclusion that the driving mechanism for the process is that of a gas drive of the swollen crude. In this final phase of the study, the fluid/fluid displacement process was gravity-stabilized; mixing of the CO2 with the residual crude oil was minimized. As a result, the performance of CO2 was only marginally better than that of relatively insoluble nitrogen.
Introduction
The work described herein concludes the scaled physical model study of the displacement of residual crude by CO2 that had been sponsored by the U.S. DOE and the U. of Southern California. The overall goal of the work was to elucidate the mechanism of the process and to gain insight into its ultimate efficiency in recovering the crude oil remaining in a reservoir after completing an efficient water flood. The conclusions drawn from the work are limited to the displacement and recovery of residual, nonmobile crude oil. The general applicability of these conclusions to the recovery of higher saturations of mobile crude oil by CO2 is still moot. In the first phase of this work the role of CO2 was simulated by using fluids that were both liquid and completely miscible with the residual crude oil in scaled models of a linear reservoir. The results were consistent with expectations based on the basic laws for fluid flow in porous media.1. After a water flood, the water is the first phase to be displaced by the injected oil-miscible fluid. This can be accounted for readily by the water being the only mobile phase in the reservoir after water flooding has been concluded.2. The first appearance of the injected oil-miscible fluid in the effluent and the volumetric efficiency of the fluid displacement process are functions of the density and viscosity of the injected oil-miscible fluid. These observations can be related to the injected oil-miscible fluids being less dense and less viscous than the continuous, mobile water: therefore the injected fluids both segregate as a result of gravitational instability and finger through the water as a result of viscous instability.3. The residual crude oil was found to be produced along with the injected oil-miscible fluid. It was then inferred that the mechanism of the recovery of the residual crude required solution of the CO2 in the crude, swelling (saturation increase), and, ultimately, displacement of the swollen crude by the injected fluid. The scaled model studies, in which the role of CO2 was simulated by low-viscosity miscible fluids, was followed by high-pressure studies in which CO2 itself was used. However, the geometry was still restricted to that of a linear low-permeability reservoir having a circular cross section. The low permeability of the prototype, 25 md, was chosen since the earlier work already prototype, 25 md, was chosen since the earlier work already had indicated the highly pernicious effect of gravity segregation on the performance of CO2 in displacing and recovering residual crude oil. The circularity and linearity of the reservoir were dictated by the cost of a more realistic three-dimensional model. However, it was easy to show that the linearity and circularity of the prototype led to, if anything, a more optimistic performance being deduced for the recovery process. The results of the high-pressure study duplicated those in which the role of CO2 had been simulated. In addition, however, specific effects of pressure, temperature, and slug size were revealed.1. The efficiency of the displacement process in terms of oil produced per 1,000 scf of CO2 injected (bbl/10 scf) increases rapidly with pressure, but the effect of pressure becomes smaller and smaller as the pressure pressure becomes smaller and smaller as the pressure approaches 1,000 to 1,200 psi [6895 to 8274 kPa] (at temperatures in the range of 100 to 150 degrees F 138 to 66 degrees C]). The reduced effect of pressure on efficiency occurs when the density also becomes less sensitive to further increases in pressure-i.e., when the behavior of the density suggests the CO2 is liquid rather than gaseous.2. Increasing temperature reduces the efficiency of the process and the experiments showed that the negative process and the experiments showed that the negative effect of temperature could be offset by an increase in pressure. The efficiency of the CO2 was the same as long pressure. The efficiency of the CO2 was the same as long as the pressure and temperature of the experiment resulted in an identical density of the CO2.3. The efficiency increases with residual oil saturation (ROS).
SPEJ
P. 593
Data Size Requirement for Forecasting Daily Crude Oil Price with Neural Networks