Phase Behavior of Several CO2/ West Texas-Reservoir-Oil Systems

Abstract

Summary. The design of miscible CO2 recovery methods and the evaluation of laboratory CO2 coreflood and pilot field studies require knowledge of the phase behavior encountered in such processes and the ability to make phase behavior encountered in such processes and the ability to make reliable predictions. Because of the complexity of CO2/hydrocarbon phase behavior, experimental measurements are necessary as a basis from which to develop an understanding. In this paper, measured phase equilibria and volumetric properties are reported for several west-Texas-reservoir- oil/CO2 systems. Reservoir oils studied were from the San Andres, Grayburg, and Devonian Chert formations. Both static (single-contact) and multiple-contact measurements have been conducted in a visual fluid property cell. Static data cover a wide range of CO2 compositions and property cell. Static data cover a wide range of CO2 compositions and provide a general understanding of CO2/reservoir-oil phase behavior. provide a general understanding of CO2/reservoir-oil phase behavior. Multiple-contact measurements in which the CO2-rich phase is repeatedly contacted with recombined reservoir oil (oil cycling) simulate phase behavior occurring at the flood front. Multiple-contact measurements in which the oil-rich liquid phase is repeatedly contacted with pure CO2 (CO2 cycling) simulate the phase behavior exhibited by residual oil near a CO2 injection well. The multiple-contact data cover narrow compositional paths encountered in displacement processes and serve as a basis for equation-of- state (EOS) evaluation. Phase behavior trends common to all systems are discussed. Introduction Current interest in miscible EOR methods has led to the use of compositional simulators to understand and predict the performance of such processes. Fluid property considerations are highly important processes. Fluid property considerations are highly important from the following two standpoints. 1. An essential part of such a simulator is a means of predicting the complex phase equilibria likely to be encountered in such EOR processes. While reliable EOS have been developed to calculate phase behavior in these processes, most often parameter adjustments are required to describe the processes, most often parameter adjustments are required to describe the CO2/reservoir-oil systems properly. These adjustments require experimental data on systems of interest. 2. The evaluation of the physics required in such a simulator depends on an understanding of the fluid properties that will be encountered. The relative importance of viscous fingering, gravity override, physical dispersion, and low-interfacial-tension effects must be assessed through interpretation of laboratory CO2 coreflood and pilot field studies. This requires an understanding of the phase equilibria encountered in such processes and the ability to phase equilibria encountered in such processes and the ability to make reliable predictions. For these reasons, an experimental study of the phase behavior exhibited by CO2/west-Texas-reservoir-oil systems was initiated. The resultant data serve as criteria for the evaluation of predictive tools and to advance the general understanding of fluid properties encountered in these systems. With the exception of compositional analysis, which are tabulated in this paper, fluid properties measured in this work are presented graphically; however, values can generally be read from the figures to the accuracy with which the properties were measured. properties were measured. Measurement Procedure Both static (single-contact) and multiple-contact phase-equilibrium measurements have been conducted for mixtures of CO2 with various west Texas reservoir oils. Static tests were performed to define phase boundaries and quality (volume percent oil-rich phase) lines accurately, to determine equilibrium phase compositions, and to measure phase densities and viscosities. Multiple-contact measurements provided phase splits, equilibrium phase compositions, phase densities, and phase viscosities as bulk hydrocarbon phase densities, and phase viscosities as bulk hydrocarbon compositions were altered through the oil- and CO2-cycling processes. Following the laboratory procedures developed by Jacoby and Yarborough, each reservoir oil was prepared by recombination of primary separator liquid and gas to agree with the compositions and bubblepoint pressures determined for the previously studied oils from these fields. Chromatographic analyses of the recombined reservoir oils are shown in Table 1. Breakdowns of the C7+ fractions are given in terms of narrow boiling-point cuts, each of which includes all of the components with boiling points greater than the previous fraction and extending through the designated normal paraffin component. The detailed C7+ analyses were conducted with a temperature-programmed HP-5890 gas chromatograph incorporating a 49-ft [15-m] methyl silicone-coated Megabore column. Specific gravities and molecular weights are also listed for the lumped C7+ heavy ends. Also included are weathered oil samples used to assess the effects of using "dead" oils in the phase equilibrium studies. Recombined oil was weathered by placing an open beaker of the oil on a stirred heating plate within an exhaust hood for 24 hours at the specified temperature. Static Phase-Equilibrium Measurements. Static phase-equilibrium measurements were conducted in the following manner. Each recombined reservoir oil was charged to the variable-volume windowed PVT cell in which the phase-equilibrium measurements were performed. CO2 was then added incrementally to obtain the desired mixture compositions. Pressure was continually increased until a single-phase condition was observed. The fluids were continually agitated during these steps to ensure homogeneity of the single-phase fluid. A constant-temperature pressure/volume traverse was then performed to establish the saturation pressure of the mixture and to check for the presence of three phases (oil-rich liquid, CO2-rich liquid, and vapor) at high bulk CO2 compositions. Sufficient time for mixing and equilibration was allowed between each pressure reduction to ensure stable conditions. Mixtures of the pressure reduction to ensure stable conditions. Mixtures of the reservoir oil and CO2 led to the precipitation of a finely dispersed asphaltic phase. This phase did not settle out appreciably with time (up to 48 hours) and in some cases obscured visual observation of the cell contents. Its presence was disregarded with respect to our conclusions concerning the attainment of equilibrium. (Its presence was also discounted with regard to phase volumes.) If three phases were observed, the upper and lower pressure bounds were located. These steps were repeated until the three-phase locus was established in overall composition space at a given temperature. Phase compositions were determined by chromatographic analysis using dimple valves to trap fluid samples from the equilibrium cell and to introduce them directly into the carrier gas stream of the temperature-programmed chromatograph. SPERE P. 537