Summary This paper summarizes the response from foam injection in a pilot test conducted at the East Vacuum Grayburg/San Andres Unit (EVGSAU), including results of the reservoir characterization effort and an extensive monitoring program. The paper specifically presents (1) the positive injection-well response (documented with injectivity data and profile logs), (2) the positive production-well response, and (3) the economics of the foam test. Introduction A pilot pattern (Fig. 1) in the EVGSAU, operated by Phillips Petroleum Co., was selected for a comprehensive evaluation of the use of foam for improving the effectiveness of a CO2 flood. This 4-year field trial was jointly funded by the EVGSAU working-interest owners, the U.S. Dept. of Energy (DOE), and the State of New Mexico. The Petroleum Recovery Research Center (PRRC), a division of the New Mexico Inst. of Mining & Technology, provided laboratory and research support. Although the overall CO2 project performance at EVGSAU has been very encouraging, certain wells/patterns have shown anomalously high CO2 production. This has resulted in isolated cases of poor pattern sweep efficiency, inefficient CO2 utilization, and increased recycling costs and compression requirements. We suspected that these problems resulted from channeling of injected fluids through high-permeability zones, most likely exacerbated by dissolution of anhydrite in these zones. The goal of the field trial was to investigate foam for conformance control to aid in suppressing this rapid CO2 breakthrough. Foam may improve injection conformance in two directions:in a horizontal direction by increasing areal sweep efficiency andin a vertical direction by diverting fluids to other, underinjected zones. Specifically, the prime directive of the trial was to prove that a foam could be generated and that it could change the mobility of CO2 in the reservoir. Proving or even determining economics or optimizing the size of the foam slug, while important, was not the original goal. A geologic study delineated the major flow units and identified high-permeability layers that were channeling CO2 to a producing well in the pilot pattern. Production and injection logs were run in the producer and the injector and confirmed the channel. Laboratory work was performed at three different laboratories to determine which surfactant was most compatible with EVGSAU reservoir rock and fluids. Refs. 3 through 6 give details of that laboratory work, and Ref. 7 summarizes the work.

SPE Members Abstract The East Vacuum Grayburg San Andres Unit (EVGSAU) operated by Phillips Petroleum, is the site selected for a comprehensive evaluation of the use of foam for improving the sweep efficiency of a CO2 flood. The four-year project is jointly funded by the EVGSAU Working Interest Owners (WIO), the U.S. Department of Energy (DOE), and the State of New Mexico. The Petroleum Recovery Research Center (PRRC), a division of the New Mexico Institute of Mining and Technology (NMIMT), is providing laboratory and research support for the project. A Joint Project Advisory Team (JPAT) composed of technical representatives from numerous major oil companies, PRRC, and DOE provides input, review and guidance for the project. This paper is the second in a series of papers detailing various aspects of the CO2 Foam Field Verification Pilot test at EVGSAU. An earlier paper summarized the project plans and detailed the laboratory work leading to the selection of a surfactant for the field trial. This paper presents:an overview of the operating plan for the project,details of the foam injection schedule and design criteria, anda discussion of the data collection program and performance criteria to be used in evaluating successful application of foam for mobility control in the EVGSAU CO2 project. Specific items discussed in the foam injection design include the determination of surfactant volume and concentration, selection of the surfactant-alternating-gas (SAG) injection sequence for foam generation, field facilities, operations during foam injection, and contingency plans. An extensive data collection program for the project is discussed including production testing, injection well pressure and rate monitoring, injection profiles, production well logging, observation well logging program, and both gas and water phase tracer programs. Introduction The EVGSAU, located about 15 miles northwest of Hobbs in Lea County is the site of the first full-scale miscible carbon dioxide injection project in the state of New Mexico. CO2 injection at EVGSAU began in September, 1985. The CO2 project area covers 5000 acres developed using an 80-acre inverted nine-spot flood pattern. The total CO2 injection rate is about 30 MMcf/D. A water-alternating-gas (WAG) ratio of 2:1 (time basis) is used in the project, resulting in about one-third of the project area being on CO2 injection at any one time. In any given area, a WAG cycle consists of about four months of CO2 injection followed by eight months of water injection. This results in approximately 1.5 to 2% hydrocarbon pore volume (HCPV) CO2 injection and 3 to 4% HCPV brine injection per WAG cycle. The project is currently in the seventh WAG cycle. The tertiary oil response at EVGSAU to date has been very favorable. As shown in Figure 1, the waterflood decline established prior to CO2 injection has been arrested, and oil production has held approximately constant near the current 9000 BOPD Unit total for the past six years. P. 115^

Summary This paper presents results of hydraulic fracture azimuth and in-situ stress measurements for two wells in the west Texas Canyon Sands formation. The paper gives information for designing infill drilling patterns and improving fracture treatment designs. Eight techniques were applied to measure fracture azimuth; the results from each are discussed. In-situ stress was measured in 12 intervals, and the results were used to calibrate long-spaced digital sonic logs. This paper also discusses the application of calibrated stress to fracture treatment design. Introduction Since 1984, the Gas Research Inst.'s (GRI) Tight Gas Sands Program has sponsored a number of field-based experiments to verify the effectiveness of recently developed technologies in geologic basins known to contain significant amounts of potentially recoverable gas in tight-gas-sand reservoirs. The strategy of this research is to conduct an integrated program emphasizing resolution of three main tight-sand R&D needs:improved geology and resource parameter quantification;enhanced stimulation techniques, including modeling of hydraulic fracture treatments; andimproved fracture diagnostics for determining fracture azimuth and dimensions. This strategy was extended to the Canyon Sands in Sutton County, TX, through data acquisition and analysis on cooperative wells with Phillips Petroleum Co. and Enron Oil & Gas Co. Fig. 1 shows the wellsites. The project was initiated because of the significant tight gas resource in the Canyon Sands and because of the relative effect that advanced technology could have on improving extraction of this resource. The overall GRI data collection and analysis effort on both wells supported geologic evaluation, log interpretation development, and reservoir analysis of the Canyon Sands. However, the objective of this paper is to present the results obtained in predicting or measuring hydraulic fracture azimuth and in measuring the in-situ stress of various layers in the Canyon Sands. This information will be useful for (1) more effective infill drilling patterns of hydraulically fractured wells and (2) improved design and modeling of hydraulic fracture treatments in the Canyon Sands interval. Background The two wells that have been analyzed extensively in this project are within 4 miles of each other (see Fig. 1). The Phillips Petroleum Ward Well C-11 (the Ward well) is in the Sonora field, and the Enron Oil & Gas Sawyer 144A Well 5 (the Sawyer well) is in the Sawyer field. Drilling and openhole data acquisition - including extensive coring, logging, and openhole stress testing - began in March 1990 on the Ward well and in April 1990 on the Sawyer well. Cased-hole data acquisition - including cased-hole stress testing, pre- and postfracturing well testing, and fracture treatment monitoring - were performed during the remainder of the year. The Canyon Sands interval in each well is about 1,250 ft thick and occurs in the approximate depth interval from 5,250 to 6,500 ft. When the two wells were completed, the gross Canyon Sand interval was informally subdivided into upper, middle, and lower intervals. The lower interval is characterized by discrete sandstone reservoirs bounded by thicker shales. The middle interval is characterized by thick sandstone reservoirs and a few thin shale beds. Like the lower interval, the upper Canyon Sands interval is characterized by shale-bounded sandstone reservoirs. Approximate well-to-well correlations of these subdivisions are possible, but individual sandstone reservoirs within these intervals are usually discontinuous. Preliminary interpretations based on examination of Canyon Sands core indicate a deep-water submarine fan depositional setting.** Techniques of Predicting Hydraulic Fracture Azimuth A combination of eight techniques was applied to estimate hydraulic fracture azimuth in the Canyon Sands. These included (1) oriented overcoring and direct observation of hydraulic fractures created during openhole stress tests, (2) use of wireline logging tools [Circumferential Acoustic Scanning Tool (CAST) and Formation MicroScanner (FMS)] to image the openhole stress test fracture in the borehole wall, (3) correlation of coring-induced, drilling-induced, and natural fractures to in-situ stress fields, (4) acoustic velocity anisotropy in core, (5) an elastic strain recovery, (6) oriented caliper logging to determine preferential direction of borehole breakouts, (7) postfracture oriented gamma ray logging, and (8) continuous microseismic radiation surveys in the treatment well. Not all techniques were attempted on both wells, and not all were effective in providing useful fracture azimuth information. The results acquired from each of these techniques are presented in the following sections. Overcoring of Open-Hole Stress Tests. Two openhole stress tests were conducted in the Canyon Sands, and the hydraulically induced fractures were subsequently overcored. The basic technique is (1) to set an openhole packer to isolate a zone between the packer and the bottom of the borehole, (2) to pump fluid down the drillpipe and induce a fracture in the formation beneath the bottom of the borehole and in the borehole wall, and (3) to cut an oriented core and attempt to intercept the fracture induced beneath the borehole (i.e., over core). Daneshy et al.1 described this technique more completely. The sequence of operations described above was successfully implemented in the Ward well at 6,050 ft. The core cut after the openhole stress test was 4 in. in diameter (in a 7 7/8-in. hole) and 24.5 ft long. No hole deviation problems were noted, and the borehole at the stress test interval was within 1/2° of vertical. The core was oriented with an electronic survey instrument. Core-orientation quality-assurance measures were implemented rigorously at the wellsite, and the resulting orientation data quality is valid. Overcoring of Open-Hole Stress Tests. Two openhole stress tests were conducted in the Canyon Sands, and the hydraulically induced fractures were subsequently overcored. The basic technique is (1) to set an openhole packer to isolate a zone between the packer and the bottom of the borehole, (2) to pump fluid down the drillpipe and induce a fracture in the formation beneath the bottom of the borehole and in the borehole wall, and (3) to cut an oriented core and attempt to intercept the fracture induced beneath the borehole (i.e., over core). Daneshy et al.1 described this technique more completely. The sequence of operations described above was successfully implemented in the Ward well at 6,050 ft. The core cut after the openhole stress test was 4 in. in diameter (in a 7 7/8-in. hole) and 24.5 ft long. No hole deviation problems were noted, and the borehole at the stress test interval was within 1/2° of vertical. The core was oriented with an electronic survey instrument. Core-orientation quality-assurance measures were implemented rigorously at the wellsite, and the resulting orientation data quality is valid.

SPE Members Abstract The East Vacuum Grayburg/San Andres Unit (EVGSAU), operated by Phillips Petroleum Company, is the site selected for a comprehensive evaluation of the use of foam for improving the effectiveness of a CO2 flood. The four-year project is jointly funded by the EVGSAU Working Interest Owners (WIO), the U.S. Department of Energy (DOE), and the State of New Mexico. The Petroleum Recovery Research Center (PRRC), a division of the New Mexico Institute of Mining and Technology (NMIMT), is providing laboratory and research support for the project. A Joint Project Advisory Team (JPAT) composed of technical representatives from numerous major oil companies provides input, review, and guidance for the project. The EVGSAU, located about 15 miles northwest of Hobbs in Lea County, is the site of the first full-scale miscible carbon dioxide injection project in the state of New Mexico. The 5000 acre CO. project is divided into three water-alternating-gas (WAG) areas where CO2 injection was initiated in September of 1985. A 2:1 WAG ratio was chosen so that while CO2 is injected into one area, water is injected into the other two areas of approximately equal pore volumes. After each fourth month of operation, CO2 injection is rotated into another WAG area. While tertiary oil response at the EVGSAU is very favorable, some wells are showing excessive CO2 breakthrough, thereby increasing CO2 recycling and compression costs. This project includes a field demonstration of the use of foam to reduce the mobility of the injected CO2, reduce excessive CO2 production, improve the volumetric sweep efficiency of the injected CO2, and increase the incremental oil recovery from the tertiary project. Thus, a suitable pattern in the EVGSAU was selected, based on the criterion that the production there be typical of other patterns without a distinctly better or worse record of CO2 breakthrough than in the rest of the field. An observation well was drilled in the pattern; location of this well is approximately 150 ft from the pattern injection well, The observation well was cored and logged to improve reservoir characterization in the pattern area, as well as to provide reservoir cores for laboratory tests with suitable foam-generating surfactants. In order to use the borehole as a logging monitor well, the bottom 800 ft was cased with fiberglass. The objective of this four-year project is to conduct reservoir studies, laboratory tests, simulation runs, and field tests to evaluate the use of foam for mobility control or fluid diversion in a CO2 flood. A geological characterization of the pilot area and surrounding patterns has been assembled for the history matching and reservoir simulation studies that are in progress. The foam-flood mechanistic model developed at the PRRC is being incorporated into the field-scale reservoir simulator. This paper summarizes the project plans, the baseline field testing, and the laboratory test results that pertain to surfactant selection. This overview provides a background for subsequent papers that will report on various aspects of the project. Background The use of CO2 as a displacement fluid during enhanced recovery processes has increased in recent years, and work involving the selection and development of mobility control additives for use in CO2 flooding has gained importance. Several organizations have been working on processes to improve the efficiency of CO2 displacements that consist of the injection of a mixture of dense CO2 with an aqueous solution of a suitable surfactant. This mixture generates lamellae (bubble films) in the pore space of the rock which allows the mixture to move through the rock with a mobility that is significantly lower than that of CO2 alone. The CO2-foam that is generated can also reduce the nonuniformities of the displacement front that are otherwise induced by now through the heterogeneities of the rock. Thus, the use of CO2-foam as a displacement fluid can give two benefits over the use of CO2 alone: it can reduce or suppress the formation of fingers caused by the instability of the displacement front, and it can reduce the severity of channels or preferential now that would otherwise occur because of heterogeneity of the reservoir rock. For several years, laboratory work has been conducted at the Petroleum Recovery Research Center (PRRC), a division of New Mexico Institute of Mining and Technology (NMIMT), on the use of surfactants to generate foam for increasing the efficiency of CO2 floods. P. 201^