Summary. The minimum miscibility pressure (MMP) for a gas/oil pair can be measured within 1 hour with the rising-bubble apparatus (RBA). Development of miscibility between a gas bubble and an oil can be observed visually. The measurements of the MMP with the RBA compare favorably with those based on slim-tube experiments and predictions from phase-behavior studies. Introduction For a gas flood in an oil reservoir, MMP is the lowest possible operating pressure at which the gas can miscibly displace oil. The MMP of a gas/oil pair is traditionally determined by flooding an oil-saturated slim tube with a gas at four or five different pressures; the MMP is found from the pressure dependence of oil recovery. Stalkup describes slim tubes in detail. We designed and built the RBA as a reliable, fast alternative to a slim tube for measuring MMP. With the RBA, direct visual observations of miscibility development can be made. In contrast to the slim tube, pressure dependence of oil recovery is not used to indicate MMP with the RBA. This is not a great loss, however, because we do not believe that oil recovery and its dependence on pressure as measured in a slim tube correspond to what might occur pressure as measured in a slim tube correspond to what might occur in an oil reservoir. Coreflooding, combined with simulation and PVT studies, is probably the best way to determine the sensitivity PVT studies, is probably the best way to determine the sensitivity of field-scale oil recovery to flooding pressure. RBA Design and Operation. The heart of the RBA is a flat glass tube mounted vertically in a high-pressure sight gauge in a temperature-controlled bath. The rectangular internal cross section of the glass tube is 0.04 × 0.20 in. [1 × 5 mm]. The visible portion of the tube is about 8 in. [20 cm] long. The sight gauge is backlighted for visual observation and photography of rising bubbles in the oil. With the flat side of the tube perpendicular to the direction of the incident light, gas bubbles are visible even in opaque crudes. A hollow needle for injecting gas bubbles into the glass tube is mounted at the bottom of the sight gauge. For further details, see Ref. 2. The RBA as now designed can operate at up to 300 deg. F [420 K]. For pressures up to 5,000 psi [34 MPa], a single-window sight gauge is pressures up to 5,000 psi [34 MPa], a single-window sight gauge is used; for pressures up to 10,000 psi [69 MPa], a multiple-window sight gauge is used. In preparation for an experiment, the sight gauge and glass tube are filled with distilled water. Enough oil is then injected into the glass tube to displace all but a short column of water in the tube's lower end (Fig. 1). Next, a small bubble of gas of the desired composition is launched into the water. The buoyant force on the bubble causes it to rise through the column of water. then through the water/oil interface. As the bubble rises through the oil, its shape and motion are observed and photographed with a motor-driven 35-mm camera. Between 5 and 30 seconds are needed for the bubble to rise the length of the oil column. After two or three bubbles have risen through the column of oil, the "used" oil is replaced with fresh oil. For a gas/oil pair, rising-bubble experiments are repeated over a range of pressures. From the pressure dependence of the behavior of the rising bubbles, MMP is inferred. Multiple-Contact Miscibility Process. We believe that the mass-transfer process that occurs as the gas bubble rises through the oil in the glass tube is similar to the multiple-contact process described for gas displacements of oil in a slim tube. This multiple-contact miscibility process is frequently shown in a pseudoternary diagram (Fig. 2). Of course, the phase behavior is pressure-dependent. With increasing pressure, the two-phase region shrinks in size and the critical tie-line shifts. As a rising bubble contacts fresh oil in the glass tube, an overall Composition alpha forms in the two-phase region, with equilibrium phases of Compositions g1 and l1. Because the bubble of Composition g1 is buoyant, it rises to contact fresh oil. An overall Composition alpha 2 with Equilibrium Compositions g2 and 2, results. As this process proceeds, the composition of the bubble creeps around the two-phase region until it becomes miscible with the oil. If the compositions of the gas bubble and oil are on the same side of the critical tie-line, miscibility cannot be generated. But if the gas and oil compositions lie on opposite sides of the critical tie-line (as in Fig. 2), multiple-contact miscibility is possible. At the MMP, the critical tie-line extends through the crude-oil composition. For pressures above the MMP with some gas/oil pairs, it is possible for the two-phase region to be so small that the gas and oil are first-contact miscible. Interpretation of RBA Experiments Direct visual observations and photographs obtained over a range of pressures from RBA experiments are used to determine the MMP of a gas/oil pair at a constant temperature. MMP is inferred from the pressure dependence of the behavior of the rising bubbles. Bubble behavior varies significantly over a range of pressures and can be divided into three distinct patterns. 1. Far below MMP, a bubble retains its initial near-spherical shape as it rises through the column of oil, although the size of the bubble decreases as gas transfers to the oil phase. As the pressure approaches MMP, a bubble still remains nearly spherical on top, but the bottom interface of the bubble changes from spherical to flat or "wavy." 2. At or slightly above MMP, tail-like features quickly develop on the bottom of a rising bubble, which remains spherical on top. Then, starting at the bottom of the bubble, the gas/oil interface vanishes, and the contents of the bubble rapidly disperse in the oil. This type of behavior suggests a multiple-contact miscibility process, not a first-contact process, because the bubble did not process, not a first-contact process, because the bubble did not immediately disperse when it first contacted the oil at the water/oil interface. During this multiple-contact process, the volume of the bubble is almost constant (a 10 to 20 % shrinkage is common) until the interface starts to deteriorate. 3. At pressures far above MMP, a bubble will disperse more rapidly than at pressures just above MMP; with some oils, CO2 bubbles disperse immediately after reaching the water/oil contact in the glass tube (first-contact miscibility). SPERE P. 522

SPE Members Abstract This paper describes an immiscible CO2 project conducted in Halfmoon field, Wyoming. Laboratory results indicated that CO2 could improve recovery of the asphaltic, 17 degrees API gravity crude. The primary project incentive was that a gas source existed in the field. Incremental oil was produced in the field, but the project was not economic at present oil prices. Introduction Halfmoon field is located in Park County, northwestern Wyoming, in the Big Horn basin. The field is developed on 10 acre spacing, and has 27 active producers and 1 water injector. Reservoir properties are summarized in Table 1 and well locations are shown in Figure 1. The Phosphoria producing interval is a limestone/dolomite of Permian age, with open natural fractures on the crest of an anticlinal structure. The Tensleep producing interval is a sandstone of Pennsylvanian age. The structure is also an anticline, somewhat more fractured than the Phosphoria. The reservoir drive mechanism is weak, natural water drive from the deeper Madison formation, except for the northern portion of the Tensleep which experiences stronger water support. The water is believed to entrain production gas, composed of 94 mole percent CO2. Immiscible gas injection was considered because response to the existing water injector suggested the field was a poor waterflood candidate. The major reason for proposing a CO2 project was that a field CO2 source was available. The project was conceived as a field-wide CO2 huff 'n' puff operation because cyclic economics could be evaluated quickly. Several concerns were identified in the beginning stages of project design. CO2 injection rate would be limited by production supply. Natural fractures would cause conformance problems. The distribution of remaining oil was unknown, and mobilization of altered oil might be inefficient due to oil heaviness and the weak reservoir drive. The influence of rock type on process response was unknown. Finally, asphaltene precipitation was possible. A laboratory evaluation was undertaken to alleviate project concerns. In particular, scoping corefloods would address the influence of ROS, rock type, and reservoir drive. The tendency for asphaltenes to precipitate would be determined. Coreflood and modeling results would assist deliberations on whether cyclic CO2 injection should evolve into a flood drive. Field testing would gain operational experience and define the ultimate potential of using CO2 to improve oil recovery. LABORATORY EVALUATION Literature Review Changes in the properties of Halfmoon crude from CO2 contact were estimated from the literature. For the range of current reservoir pressures of about 500 to 900 psia, oil swelling was 5 to 10%, and oil viscosity reduction was 3 to 9 fold. Oil viscosity reduction was expected to be the primary mechanism of enhanced oil recovery. P. 155^

This paper examines the feasibility of cyclic natural gas injection for the enhanced recovery of light oil from waterflooded fields. Approximately 40 percent of waterflood residual oil was recovered in corefloods using two huff 'n' puff cycles at immiscible conditions. Gross gas utilizations were as low as 3 MSCF/STB. Response to cyclic injection contrasted favorably with immiscible WAG displacement. Coreflood results and numerical simulation indicated that incremental oil was not highly sensitive to remaining oil saturation. Predicted field recovery could be improved by managing offset wells. Response to cyclic gas injection was mostly dependent upon the amount of gas injected, suggesting that there will be a maximum economic slug size in field applications. Results indicated that repressurization and gas relative permeability hysteresis are the major oil recovery mechanisms. 78-1