Abstract
Results of an experimental study of oil recovery by a steam slug driven by a cold waterflood in a linear porous medium are described. The model included simulation of heat losses to the adjacent formations. Steam displacements were conducted, using a number of hydrocarbons and various steam-slug sizes, with the core initially containing a residual oil or irreducible water saturation.
It was found that the steam-slug displacement is more efficient in the case of light oils than for the heavier ones. The injection of cold water following steam resulted in almost total condensation of the steam present in the porous medium, with the process degenerating into a hot waterflood. The oil process degenerating into a hot waterflood. The oil recovery efficiency of the process depends on whether an oil bank is formed during the steam-injection phase and whether the oil responds favorably to a hot phase and whether the oil responds favorably to a hot waterflood
Introduction
Steam injection has been shown to be an effective oil recovery method both by field and laboratory tests. However, the method has the inherent disadvantages of a high cost of operation and excessive heat losses. The modification discussed here consists in the injection of cold water after a slug of steam, which helps to offset the above disadvantages partly at the expense of oil recovery. The injected water serves to propel the oil bank formed ahead of the steam-invaded zone and transports the heat contained in the steam-swept zone farther downstream, thus leading to more complete utilization of the heat injected.
EXPERIMENTAL APPARATUS AND PROCEDURE
Fig. 1 depicts a schematic diagram of the apparatus employed. It consisted of a 4-ft-long core composed of a steel tube having a rectangular cross-section (see Table 1 for dimensions and other information) packed with glass beads (mesh size 200 to 270, corresponding to 0.0021 to 0.0029 in.) and fitted with 15 iron-constantan thermocouples and eight pressure gauges. The two ends of the core were fitted with sintered bronze plates to ensure strictly linear fluid flow. In order to simulate the underlying formations, the core was placed upon a sand-filled wooden box having a depth placed upon a sand-filled wooden box having a depth of 2.5 ft and a length and width equal to those of the core. An identical box was placed in contact with the top surface of the core to simulate the overlying formations. The sand packs simulated infinitely thick formations, since the temperatures at the upper and lower extremities remained undisturbed during a run. The sides of the two boxes were fitted with thermometers and insulated, together with the exposed surface of the core; the top and bottom surfaces of the core were in contact with sand. An electrical system was designed for temperature measurement at the 15 points; the core inlet and outlet were fitted with thermocouples. A technique was devised for pressure measurement virtually without disturbing the flow.
A positive-displacement pump, in conjunction with a coil immersed in a high-temperature oil bath, was used for conducting hot waterfloods as well as for preparing the core for a run (Fig. 1). Steam, having a quality of 95 percent was supplied by an electric boiler capable of delivering up to 69 lb/hr at pressures up m 250 psig. The core effluent was passed though a suitable condenser provided with passed though a suitable condenser provided with a backpressure regulator used to control the steam injection rate. The average steam (as condensate) injection rate for a run was estimated by dividing the total effluent volume minus the volume of the water needed to fill up the core at the end of steam injection, by the steam injection time.
The properties of the fluids used are listed in Table 1. The hydrocarbon mixtures were chosen to study the steam distillation effects. Drakeol 15 and 33 at 80 deg. F are high-boiling mineral oils having viscosities of 515 and 100.0 cp, respectively. Viscosity-temperature behavior for the hydrocarbons used is shown in Fig. 2.
The core was saturated with distilled water and then saturated with the oil to be tested by displacement (terminal WOR 1:100). If desired, the core was waterflooded prior to steam injection (terminal WOR 100:1).
SPEJ
P. 351
Optimization Study of Integrated Scenarios on Cyclic Steam Stimulation (CSS) Using CMG STARS Simulator
