Gas-Oil Relative Permeability Ratio Correlation From Laboratory Data

Abstract

Abstract Gas-oil relative permeability ratio is an important relationship in oil reservoir predictive calculations. A correlation has been developed from 107 gas-flood krg/kro tests on Venezuelan core samples. The correlating parameter is based on restored-state water saturation tests and is applicable to both consolidated and poorly consolidated sandstone reservoirs. Data of the correlation show that there are no distinguishable differences between the mass-data groupings for the two classifications. A procedure is recommended for running sufficient relative permeability analyses to compute a geometric mean of the sample group. The geometric mean is more representative of the total core, and probably the entire reservoir. For example, while only one in four of the krg/Kro test curves agreed closely with the resultant correlation of this report, the geometric mean curves of the 16 suites (three samples or more) showed good agreement in three cases out of four. Introduction The gas-oil relative permeability, ratio is an important, fundamental relationship in most oil reservoir predictive calculations. Predictive calculations are made to estimate future reservoir production characteristics and ultimate oil recovery. The Krg/Kro relationship is specifically needed to relate the surface gas-oil ratio to the reservoir oil and gas saturation, and to calculate the relative movement of these phases within the reservoir whenever some of the more complex driving mechanisms are present. Laboratory Krg/Kro tests are not generally run as a routine analysis. Consequently, Krg/Kro data often are not available when needed because the cost of laboratory work could not be justified or the need for such data had not been properly anticipated. When laboratory Krg/Kro data are available, they are often very difficult to interpret. For example, wide divergence is sometimes shown in a family of Krg/Kro tests representative of the producing horizon in a single well. With these considerations in mind, a study was made to determine if a relationship might exist between the Krg/Kro curve and some other simple laboratory test criteria. The most probable Krg/Kro curve correlation for Venezuela described in this paper is the result of the investigation. The presented correlation defines the most probable gas-flood Krg/Kro curve through the medium of air-water capillary displacement and centrifuge water saturation tests. The laboratory procedures of these tests are relatively simple and inexpensive: test data should be widely available from routine analysis. DATA AVAILABLE, LABORATORY METHODS The report correlation utilized 107 gas-flood Krg/Kro tests run on sandstone cores of Venezuelan reservoirs. Table 1 is a general tabulation of data pertinent to the tests, while Table 2 summarizes the data. The tests include 96 from Western Venezuela and 11 from Eastern Venezuela. Eighty-two of the 107 test samples were sandstones that varied from poorly consolidated to unconsolidated; 25 were consolidated. The average sample porosity was 26.7 per cent and the average permeability was 1,121 md; these values typify the better sandstone reservoirs of Venezuela. The Welge gas-flood technique. based on fundamental Buckley-Leverett frontal displacement theory, was introducted in about 1952 and is widely accepted in the industry. The laboratory procedure is relatively simple, rapid, and can be performed on small core samples. While there have been some minor variations in sample preparation and laboratory procedure in the tests used for the correlation, these tests can be generally summarized as follows. The core sample was first solvent-extracted and dried. Connate-water saturation was restored by the oil-flushing or evaporation-blowdown methods. At the beginning of flood the hydrocarbon pore volume was completely saturated with the test oil phase. Unsteady-state gas-ail displacement then began with the injection of nitrogen or helium, while the displaced oil and gas phases were incrementally metered at the out-flow face. From the test data. the Krg/Kro curve was calculated by the Welge method. The individual oil and gas relative permeabilities were also calculated. CORRELATING PROCEDURES In attempting to establish a basis of correlation, we found that broad mid-range sections of 105 of the 107 Krg/Kro test curves could be closely duplicated by a straight line. Only two curves did not show a degree of linearity in this region. Correlation-curve definition parameters were subsequently developed from this observation of consistent mid-range linearity. Possible correlating variables were limited to the physical properties measured on core samples thatwere widely available as common test data andcould be easily and cheaply obtained through future laboratory work. JPT P. 1111ˆ