Tertiary Recovery Techniques Research Articles

Laboratory Investigation of Weyburn CO2 Miscible Flooding

Abstract Weyburn reservoir, located in southeast Saskatchewan and operated by PanCanadian Petroleum Ltd., has reached its economic limit of production by waterflooding. The reservoir is a target for tertiary CO2 miscible flooding to enhance oil recovery and extend its production life. A comprehensive study has been conducted by the Saskatchewan Research Council and PanCanadian to assess the suitability of the process for Weyburn. This paper presents a technical evaluation of CO2 near-miscible injection for Weyburn reservoir. This is based on laboratory studies conducted with three Weyburn oils collected from different regions of the reservoir. These studies included:measurement of minimum miscibility pressure for the reconstituted reservoir fluids for pure and impure CO2,determination of PVT properties for reservoir fluid-CO2 mixtures, andassessment of recovery behavior from uniquely designed laboratory core-flood tests representing the permeability contrast of the reservoir. The studies indicated that the CO2 minimum miscibility pressure for the Weyburn reservoir oils varied from about 11.5 to 14.5 MPa, indicating suitability for CO2 miscible flooding. PVT data generated for the three Weyburn reservoir fluid-CO2 mixtures showed that viscosity reduction and oil swelling by CO2 also contributed to oil recovery. A slight manipulation of the measured PVT properties of the mixtures made it possible to obtain single property curves for the three Weyburn oils. This feature can be used to estimate the PVT behavior for any Weyburn oil from the reservoir. Core-flood studies showed that CO2 injected into the Vuggy zone could rise to the upper Marly zone and help enhance oil recovery. A separate simulation and modelling study conducted by PanCanadian Petroleum Ltd. satisfactorily matched the experimental PVT properties, MMP values and recovery mechanism suggested by the core-flood behaviour. Introduction The total oil reserves of Saskatchewan are estimated at over 3 billion m3 of initial oil-in-place (IOIP). The light and medium oil reserves contribute approximately 1.6 billion m3 IOIP(1). Over 10﹪ of the light and medium reserves, or 178 million m3 IOIP, are held in Weyburn reservoir. The Weyburn field is located approximately 130 km southeast of Regina in Saskatchewan and covers over 180 km2 of production area. The field was discovered in 1955 and PanCanadian Petroleum (PCP) Ltd. is the major operator. It has over 627 producing wells and 162 water injection wells on approximately 24 ha spacing(2). The oil production from the Weyburn field comes from Midale beds of the Mississippian Charles Formation in the Williston Basin at a depth of 1,310 to 1,500 m. The field was produced by primary depletion for about nine years until 1964. Waterflood was started thereafter using a nine-spot pattern. In 1985, to optimize the waterflood performance, both horizontal and vertical infill drilling programs were initiated. The combined oil recovery by primary and secondary production was approximately 28﹪ IOIP by 1994. Waterflooding has almost reached its economic limit for the Weyburn reservoir. The ultimate conventional recovery is expected to be 31﹪ IOIP, leaving a large target for tertiary recovery techniques(3). A similar situation exists for the majority of Saskatchewan's light and medium reservoirs(1,4).

Technical Feasibility of CO2 Flooding in Weyburn Reservoir - A Laboratory Investigation

Abstract Miscible or near-miscible flooding with CO2 or hydrocarbon gases will significantly increase, and extend the production life of, Saskatchewan light and medium oil reserves. These pools are approaching their economic limit of production under conventional technology (primary and secondary recovery methods). Responding to industry needs, a comprehensive study by Saskatchewan Research Council is assessing the suitability of the CO2 flooding process for reservoirs in southeast Saskatchewan. This paper presents the details of a laboratory study to evaluate the applicability of CO2 flooding for a southeast Saskatchewan reservoir, Weyburn. The evaluation is based on the measurement of CO2 minimum miscibility pressure for the Weyburn reservoir fluid, of PVT properties for reservoir fluid CO2 mixtures, and of oil recovery behaviour from core flood tests conducted in the near-miscible region. Laboratory studies showed that the tertiary CO2 injection at over 12 MPa into the waterflo'oded reservoir is expected to produce miscible flooding conditions. At this pressure of injection, over 60 mol % CO2 was dissolved in the reservoir fluid. This reduced the viscosity of the fluid nearly fourfold to 0.70 mPa.s and swelled the oil about 45%. Coreflood studies indicated that the microscopic displacement efficiency of the process increased with injection pressure in the near-miscible region. It is believed that the geology of the Weybum reservoir will prove advantageous for near-miscible operation. However, this aspect needs to be further investigated. Introduction Most of Saskatchewan's light and medium oil (LMO) reservoirs have reached their economic limit of production under conventional technology. Primary and secondary methods together recover about 21% of the initial oil-in-place (IOlP). Although the remaining light and medium oil-in-place in Saskatchewan is 1,373 million cubic metres(1) the remaining reserves producible under primary and secondary recovery are only 59 million cubic metres. This latter number would rise by 20 million cubic metres if medium oil in southwest Saskatchewan, not considered here, were included(1). The development of tertiary recovery techniques is, therefore, essential to increase the production life of these reservoirs and to maintain current production. Application of the miscible or near miscible oil displacement process using CO2, hydrocarbon or other gases as injection fluids is expected to increase the LMO reserves threefold and extend the production life of these pools by two decades(2). Miscible (or near-miscible) flooding with carbon dioxide or hydrocarbon solvents is considered to be one of the most effective enhanced oil recovery (EOR) processes applicable to LMO reservoirs. A comparison of effectiveness of CO2 and hydrocarbon gases as miscible solvents has been discussed(3). CO2 floods minimize gravity segregation compared with the hydrocarbon solvents. In addition, CO2 generally costs less than hydrocarbon miscible solvents in the United States(3). There is a limited supply of CO2 in Saskatchewan and Alberta for EOR purposes. However, CO2 can be imported from the U.S. to be implemented in miscible displacement projects in Canada. Current industry interest in CO2 miscible flooding is high, as evidenced by the growing level of activity in field testing(4) and increasing production from this method worldwide.

A Laboratory Evaluation of Suitable Operating Strategies For Enhanced Heavy Oil Recovery By Gas Injection

Abstract A majority of Saskatchewan's heavy oil reservoirs are thin and marginal, unsuited for thermal recovery methods. The immiscible CO2 or gas displacement process appears to be a very promising enhanced oil recovery technique for these heavy oil reservoirs. This process, if proven applicable, will access 90% of the initial-oil-in-place. Since an inexpensive source or CO2 is not available in Saskatchewan/Alberta, alternative gases containing CO2 can be used. This paper discusses results of a laboratory study carried out by Saskatchewan Research Council (SRC) to assess the suitability of a flue gas containing 85 mol% N2 and 15 mol% CO2, and to evaluate various operating strategies for heavy oil recovery. The test heavy oil (14 º API gravity) was collected from Senlac reservoir located in the Lloydminster area. The study investigated the effect on oil recovery of drive type (secondary vs. tertiary), injection mode (slug test vs. water- alternating-gas [WAG] flood), oil type (Senlac live oil vs. dead oil), and injection gas (flue gas vs. CO2). Results of one-dimensional linear coreflood tests (conducted at 28 ºC, 2.5–2.7 MPa) with Senlac oil-flue gas led to four major conclusions. (1) The secondary gas slug flood with live oil was the most suitable injection strategy for heavy oil recovery from among three different injection strategies (secondary slug, tertiary slug, and tertiary WAG) investigated because this flood had the highest displacement efficiency; (2) The tertiary WAG process recovered more oil than the tertiary slug process because the former predominantly improved mobility control; (3) Ultimate oil recovery was higher in runs conducted using live oilthan in those using dead oil. This is attributed to the relatively more favourable mobility of live oil and slightly higher operating pressure (2.7 MPa instead of 2.5 MPa) in live oil runs; (4) Flue gas appeared to be an effective flooding agent, since the oil recoveries were only 2 to 4 percentage points lower than those obtained with CO2, Comparable oil recoveries in flue gas runs were believed to be a combined result of two competing mechanisms -a free gas mechanism provided by N2 and a solubilization mechanism provided by CO2 in contrast to the solubilization mechanism (provided by CO2) which predominates in CO2 floods. Introduction In Canada, Alberta and Saskatchewan account for nearly all of the proven heavy oil resource. In Saskatchewan alone, l.5 billion m3 of proven and 3.4 billion m3 of probable heavy oil reserves exist,(2) over 62% of the total resource(1).. These reservoirs are, however, characterized by thin pay, shaly sand, heterogeneity, low roductivity, and bottomwater(3, 4). Of the province's proven initial heavy oil-in-place, 97% is contained in reservoirs less than 10 m pay zone, and 55% in reservoirs less than 5 m thick(5, 6). Primary and secondary methods combined recover only about 7% of the initial-oil-in-place (IOIP) of the proven resource(2). In comparison, such conventional techniques can access, on average, about 22% of the proven 0.43 billion m3 of medium oil. The remaining heavy and medium oil resources of Saskatchewan offer potential for exploitation by tertiary recovery techniques.

A Laboratory Miscible Displacement Study For the Recovery of Saskatchewan's Crude Oil

Abstract Most of Saskatchewan's light and medium oil (LMO) reservoirs have reached their economic limits of production under waterflooding, leaving 70 to 80 per cent of the initial oil-in-place (IOIP) in the reservoirs. The miscible displacement process using carbon dioxide, hydrocarbon, and other gases as solvents appears to be one of the most promising methods for these reservoirs. A laboratory study was conducted to evaluate the applicability of various solvents for the recovery of oil from a southeast Saskatchewan reservoir. The physical, chemical, and phase behavior (PVT) properties of wellhead and reservoir fluids were determined. Slim tube displacement tests were conducted between the reconstituted Pinto reservoir fluid and three solvents: carbon dioxide, ethane, and wellhead gas from the Steelman gas plant. Tests were carried out at various pressures and the reservoir temperature of 56 °C. The minimum miscibility pressures (MMPs) determined for the above three systems demonstrated that the MMP for ethane was much lower than that for carbon dioxide, which was lower than that for Steelman gas. The MMPs for ethane and carbon dioxide were below the reservoir fracture pressure. Introduction Most of Saskatchewan's light and medium oil (LMO) reservoirs have reached their economic limit of production under current technology. Primary and secondary methods together recover about 21 % of the initial oil-in-place (IOIP). Although the remaining light and medium oil-in-place in Saskatchewan is 1135 million cubic metres(1), the remaining producible reserves under primary and secondary recovery are only 65 million cubic metres. Medium oil in southwest Saskatchewan has not been considered here, but if it is included then up to 17 million cubic metres of additional oil (through primary and secondary recovery) may be recovered(1). The development of tertiary recovery techniques, therefore, is essential to increase the production life of these reservoirs and to maintain current production. Application of the miscible oil displacement process using CO2, hydrocarbon, or other gases as injection fluids is expected to increase the LMO reserves threefold and extend the production life of these pools by two decades(2). Miscible flooding with carbon dioxide or hydrocarbon solvents is considered to be one of the most effective enhanced oil recovery (EOR) processes applicable to LMO reservoirs. Current industry interest in CO2 miscible flooding is high, as evidenced by the level of activity in field testing(3,4) and CO2 source development in the United States. CO2 has a viscosity similar to hydrocarbon miscible solvents. Both types of solvent affect the volumetric sweep out because of unfavorable viscosity ratio. However, CO2 density is similar to that of oil. Therefore, CO2 floods minimize gravity segregation compared with the hydrocarbon solvents. In addition, the supply cost of CO2 is generally more favorable than hydrocarbon miscible solvents in the United States(3). However, there is no cheap source of CO2 in Canada yet. It should be noted that, since the world oil price collapse in 1986, the number of EOR projects in the United States has decreased markedly from 512 in 1986 to 366 in 1988, whereas the number of CO2 miscible projects has increased from 38 in 1986 to 49 in 1988, a 29% rise during these two years(4).

Yates Field, West Texas: Geology of a Permian Basin Giant

Yates field, which produces oil largely from Late Permian San Andres dolomites, was discovered in 1926 and logged its billionth barrel in 1985. It is remarkable for the size of its original oil in place (four billion barrels estimated) and for its early ability to produce hundreds to thousands of barrels per hour from individual wells at depths of less than 2,000 ft. The high quality of the reservoir has been related to the following geologic characteristics: (1) a broad anticline with significant closure located at the apex of regional structure which formed a trap for large volumes of primary and secondary oil; (2) a vast integrated system of porosity (ranging from intergrain/intercrystal pores to joints, fractures, and small caves) that imparts high storage capacity and transmissivity to the reservoir dolomites; (3) a thick, efficient seal of anhydrite capping the reservoir sequence; and (4) high flowing pressures during much of the history of the field. The western area of the Yates field reservoir is dominated by peritidal to lagoonal mudstones and wackestones; the eastern area, by shallow subtidal packstones and grainstones. These depositional facies reflect both aggradation and eastward progradation of upper San Andres carbonates. Effective engineering management of the more » field has followed from a recognition of the quite different reservoir qualities in the two areas of the field and from adapting secondary and tertiary recovery techniques to constraints imposed by the geologic factors listed above. « less

Production of CO2 From a Reservoir - A New Concept

Summary Very large reservoirs of approximately 985 CO2 have been discovered in southwestern Colorado. These reservoirs have pressures up to 2,700 psi (18.70 Mpa) and temperatures to 171°F (77°C). Wells are capable of flowing more than 10 MMcf/D (0.28×106 m3/d) CO2. The CO2 flows to the surface at pressures and temperatures below the critical and, therefore, in two-phase flow. For convenient handling, the CO2 must be heated to form a single-phase gas before it is transported, dehydrated, and compressed. This paper discusses a patented procedure for downhole pressuring followed by the production and processing of CO2 under continuous supercritical conditions. It also describes a field test of the procedure and the test results. The process is dependent on the properties of CO2 under continuous high pressure and even at high temperature, which approach those of a liquid. Thus, the CO2 can be produced by conventional pumping operations. The process appears applicable to production from any CO2 reservoir containing a fluid above its critical pressure. The results of the field test were very close to those anticipated by calculation. Introduction For at least the past 15 years, interest in CO2 as a tertiary recovery technique has grown. To date, most of the CO2 has been salvaged off-gas from sweetening operations in gasoline plants. This gas generally contained 10% or more extraneous gases, including nitrogen, hydrogen sulfide, and methane. The principal advantage of CO2 in tertiary recovery is its miscibility, under certain conditions of pressure and temperature, with the reservoir liquids. The first reservoir where large volumes of CO2 have been injected is the Canyon Reef reservoir of the SACROC unit in Scurry County, TX. The procedure and results of that operation are well documented.1–4 It was established that CO2 of less than approximately 87.5% pure would not be miscible with the reservoir liquids under the Canyon Reef conditions. The purity required varies from field to field because of differences in temperature, pressure, and reservoir liquids characteristics. Many fields in west Texas are reported candidates for CO2 flooding; however, some are at lower pressures than the Canyon Reef, and some can be expected to require higher purity of CO2 than 87.5%. Higher purity CO2 is miscible with reservoir fluids at considerably lower pressure. Fig. 1 shows the approximate conditions for miscibility of 87.5% pure CO2 with produced oil from various reservoirs as they existed in 1966. Reservoir pressures have decreased in some of these reservoirs because of fluid withdrawals; others have increased because of pressure maintenance. Therefore, it is fortunate that approximately 98% pure CO2 has become available in large volumes in the McElmo field in southwestern Colorado and in other areas.5 The peculiarities of CO2, particularly the characteristic that pure CO2 exists as a two-phase fluid at ambient temperature and pressures up to 1,080 psi (7.44 MPa), make it difficult to handle at the flowing conditions of the McElmo field. Consequently, the process of pumping the supercritical CO2 with a multistage centrifugal pump was conceived.6 The procedure has been tested in the field and has proved successful.

A Recipe for Residual Oil Saturation Determination

Before initiation of a CO2 tertiary recovery project, a residual oil saturation study was conducted in the deep, hot, pilot reservoir. The paper discusses the many technical problems that had to be resolved before and during the study. The problems confronted are outlined so that the procedure can be used much like a cookbook in designing future studies in similar reservoirs. Introduction With the advent of increased oil prices, tertiary recovery techniques are becoming more economically attractive. The necessity for residual oil saturation measurements will become more common since these numbers are mandatory in defining the profitability of any tertiary project. Routine procedures to measure this saturation in situ become necessary. However, we are dealing with such a delicate system that the details of planning such an operation become very involved.In 1978 Shell Oil Co., in conjunction with the U.S. DOE, conducted a residual oil saturation study in a deep, hot, high-pressured U.S. gulf coast reservoir. The work was conducted before initiation of a CO2 tertiary recovery pilot.Primary discussion in this paper centers around planning and results of a "log-inject-log" (LIL) operation used as a prime method to determine the residual oil saturation. Several independent methods were used to calculate the residual oil saturation in the subject well in an interval between 12,910 and 12,920 ft (3935 and 3938 m). In general, these numbers were in good agreement and indicated a residual oil saturation between 22 and 24%.A free-gas saturation was measured in the residual oil saturated interval. This saturation may well be typical of U.S. gulf coast reservoirs that are below the original bubble point at the time tertiary recovery is initiated. As a result, residual oil saturation typically may be less than those values determined by "flood pot" or countercurrent imbibition tests run in the laboratory. Likewise, residual saturations conventionally measured by openhole resistivity and porosity logs may be residual hydrocarbon and not residual oil saturations.The test was conducted in the "S" Sand Reservoir B (SRB), a 12,800-ft (3901-m) deep U.S. gulf coast reservoir in Weeks Island field, Iberia Parish, LA. The project setting is depicted in Fig. 1. The SRB initially contained two oil columns with more than 3 million bbl (476,700 m3) of in-place oil. Virtually all of this was in the pilot area. Before initiation of the tertiary project, the SRB had been flooded with fresh water. The remaining oil column had been produced to an estimated thickness of 23 ft (7 m). The 900 acre-ft (1 110 140 m3) of the SRB we propose to displace should contain a minimum of 258,000 bbl (40 996 m3) of water-drive residual oil. If the process is proved successful in the SRB, other major watered-out reservoirs in Weeks Island field have an estimated tertiary potential of 26 million bbl (4 131 400 m3) of oil which could be recovered by CO2 displacement. JPT P. 1999＾

Petroleum Developments in Europe in 1979

In 1979 the rise of oil prices and the threat to oil supplies from traditional Middle East sources caused an increase in development of known hydrocarbon resources in Europe. Oil production increased by 30% over the 1978 figure. The major share of the increase came from the North Sea, the Spanish Gulf of Valencia, and northern Italy as the discoveries of the 1970s were brought on stream. Vigorous appraisal and development drilling programs and new discoveries--Statoil/Shell's 31/2-1A gas discovery in the Norwegian North Sea may prove to be another giant--will ensure that production increases in these areas will continue into the 1980s. Tertiary recovery and new fracturing techniques enabled increased output from some existing fields. Exploration activity moved into high-cost and high-risk areas. Increased geologic knowledge obtained through recent offshore drilling, together with the application of new seismic techniques, has enabled the definition of deeper and more subtle traps. There was successful wildcat drilling of new plays in the producing offshore areas of the North Sea and the Italian Adriatic and also in many of the traditional onshore hydrocarbon producing areas of Europe, West Germany, Austria, France, Netherlands, Yugoslavia, Italy, Spain, and the United Kingdom. Several of these onshore areas suffered a drop in production levels in 1979. Seismic exploration and licensing awards were made in undrilled remote, often deep-water, and even hostile environments as well as in the less well-explored offshore areas where earlier drilling had given promising results. Results of wildcat drilling were not always released but BP's 26/28-1 discovery in the deep-water Porcupine Trough appears to be a major find.

Interfacial Tension Required for Significant Displacement of Residual Oil

Abstract A static analysis of displacement in a single irregular pore partially filled with oil is used to investigate the effects of interfacial tension and wettability on the tertiary recovery of residual oil by a low interfacial tension waterflood. There are several results. For the most efficient displacement of residual oil, either the porous structure should be water-wet or intermediate-wet. There is a critical value for the interfacial tension above which the residual oil cannot be displaced, but instead will assume a static configuration. Although the computation describes the static configuration of an oil globule in a pore, it suggests the underlying mechanism for the episodic motion or jump of a globule as it is displaces. The discussion of displacement in a single pore is extended by an estimate for the value of the capillary number, required for recovery of residual oil from those pores whose neck radii are larger than the mean pore neck radius. A comparison is offered with data taken from the literature. Introduction Petroleum is found in the microscopic pores of sedimentary rocks such as sandstones and limestones. Not all pores will be filled with petroleum. Some pores contain water or brine that petroleum. Some pores contain water or brine that is saturated with the minerals in the local rock structure. In the primary stage of production, oil and brine are driven into a well from the surrounding rock by the relatively large difference between the initial field pressure and the pressure in the well. Perhaps 10 to 20% of the oil originally in place is recovered in this manner. In the secondary stage of production, water or steam is pumped into a selected pattern of wells in a field, forcing a portion of the oil into other production wells. The void volume in a permeable production wells. The void volume in a permeable rock may be thought of as many intersecting pores of varying diameters. Consider two parallel pore spaces of unequal permeabilities. A blob of oil (residual oil) will be trapped in that pore space, through which the oil is displaced by the water more slowly. In this manner, 10 to 40% of the oil initially in place will be recovered, but 40 to 80% of the oil originally in place is left behind in the field. Several tertiary recovery techniques have been proposed. We focus our attention here on the proposed. We focus our attention here on the recovery of residual oil by a low interfacial tension waterflood. One can distinguish between at least two different types of such waterfloods. One type uses a small volume (3 to 20% PV) of a relatively concentrated surfactant solution. The concentration of the surfactant solution is high enough to ensure that it is miscible with the crude oil in all proportions. As a small slug of this surfactant solution moves through the porous structure mixing with and displacing the crude oil, surfactant is lost by adsorption on the rock. The solution is diluted further with the connate water present in the structure. As the concentration of present in the structure. As the concentration of surfactant falls, the water/crude oil/surfactant mixture can move from a single-phase region to a multiphase region on its phase diagram. We should expect that a miscible displacement conducted with a small slug of surfactant solution will revert to an immiscible displacement at some point in the reservoir. In the type of waterflood with which we are concerned here, a large volume (15 to 60% PV or more) of a dilute surfactant solution is used. The crude oil is nearly insoluble in the surfactant solution, and we speak of an immiscible displacement. It has been estimated that in carefully selected, well-designed, well-performing operations, an additional 30% of the oil originally in place might be recovered by a low interfacial tension waterflood. The crude oil/water interfacial tension can be reduced by orders of magnitude when either a mixture of surfactants or an alkaline solution is added to the waterflood. SPEJ P. 83

Sandstone and Carbonate Two- and Three-Phase Relative Permeability Characteristics

Abstract Three-phase relative permeability characteristics applicable to various oil displacement processes in the reservoir such as combustion and alternate gas-water injection were determined on both outcrop and reservoir core samples. Steady-state and nonsteady-state tests were performed on a variety of sandstone and carbonate core samples having different wetting properties. Some of the tests were performed on preserved samples. Some of the three-phase tests were performed on samples that contained two flowing phases and a third nonflowing phase, either gas or oil. These were classed as three-phase flow tests because the third phase played an important role in the flow behavior which was determined. The three-phase relative permeability test results are directly compared with the results of two-phase gas-oil and water-oil test. Wetting-phase relative permeability was found to be primarily dependent on its own saturation, i.e., relative permeability to the wetting phase during three-phase flow was in agreement with and could be predicted from the tow-phase data. Nonwetting-phase relative permeability-saturation relationships were found to be more complex and to depend in some cases on the saturation history of both nonwetting phases and on the saturation ratio of the second nonwetting phase and the wetting phases. Trapping of a given nonwetting phase or mutual flow interference between the two nonwetting phases when both are flowing accounts for most of the low relative permeabilities observed for three-phase flow tests. However, in special cases nonwetting-phase relative permeabilities at a given saturation are higher than those given by two-phase flow data. Despite these complexities some types of three-phase flow behavior can be predicted from two-phase flow data. Through its effect on the spatial distribution of the phases, wettability is shown to be a controlling factor in determining three-phase relative permeability characteristics. however, despite the importance of wettability the present data shown that for both water-wet and oil-wet systems oil recovery can be improved by several different injection processes, but the additional oil recovery is accompanied by lower fluid mobility. Introduction The increasing emphasis on optimizing recovery and the rapid and extensive development and use of mathematical modes for predicting reservoir performance are together creating a widespread need for reliable basic data on rock flow behavior. The two-phase imbibition or drainage flow relationships common to conventional oil recovery processes (depletion, gas or water injection, gravity drainage) are not applicable to some of the newer secondary and tertiary recovery techniques. This is because the reservoir displacement process may differ from that easily simulated in laboratory relative permeability studies. in some situations, data are needed fro a three-phase system where almost any combination of two fluids or even all three fluids may be flowing. In other, however, only two flowing phases are present, but the saturation history of the system is unique. Leverett and Lewis were the first to collect experimental relative permeability data on a three-phase system. Corey et al. were similarly leaders in efforts to define three-phase flow relationships using empirical approaches. Space does not permit a critical review of these earlier works. For those interested, a recent article by Saraf and Fatt provides a brief discussion of the experimental techniques used by earlier investigators. Suffice it to say that both experimental and empirical approaches have been used, but the applicability of both has been limited because in only one case have three-phase relative permeability data been obtained on reservoir rock material. SPEJ P. 75ˆ