Raimondi, P., SPE-AIME, Gulf Research and Development Co. Gallagher, B.J., SPE-AIME, Gulf Energy and Minerals Co.-U.S. Ehrlich, R., SPE-AIME, Gulf Research and Development Co. Messmer, J.H., SPE-AIME, Gulf Research and Development Co. Bennett, G.S., SPE-AIME, Gulf Research and Development Co. A field trial of caustic waterflooding in a Queen sand lens in the North Ward-Estes field is described. Comparison of results with projections for a conventional waterflood, while uncertain, indicate that about 25 percent more oil was produced. Operational problems are discussed. Introduction The North Ward-Estes field, Ward and Winkler Counties, Tex., has been under successful water drive for many years. However, a considerable amount of oil is left behind even in the water-swept zones. Laboratory tests have indicated that the oil displacement efficiency could be improved significantly by adding a high-concentration slug of sodium hydroxide (NaOH or caustic) to the flood water. In the swept zone, the laboratory-measured improvement was about 125 bbl/acre-ft. This is attributed to generation of surfactants by the reaction of caustic with organic acids in the oil and the effect of these surfactants on multiphase flow, primarily a wettability shift in the water-wet direction. These mechanisms have been discussed in other papers. The field test is being conducted in the Queen sand, East Flat area of the North Ward-Estes field. This area was discovered in 1969 and development drilling was completed during 1970 with the drilling of 20 additional wells. The oil production rate peaked at 85,000 bbl/ month during mid-1970, followed by a rapid decline in rate resulting from the highly undersaturated nature of the oil. To offset this decline, full-scale water injection was begun in late 1970. Production response to water injection was almost immediate and a peak rate of 40,000 bbl/month was reached in early 1971. To test the laboratory results, a 5-acre inverted fivespot pattern was developed by drilling four water-injection wells, WI-79 through WI-82. around an existing producing well, Well 874 (Figs. 1 and 2) in an area of the field that was presumed to have not yet been affected by the ongoing waterflood. The objective was to determine the effect of a caustic slug injected early in the life of a waterflood. This paper describes the test area, the laboratory evaluation, the design and conduct of the test, the production behavior and an analysis of the recovery. The production behavior and an analysis of the recovery. The test was still in progress when this paper was written. Reservoir Description Geologically, the Queen sand is a channel-type deposit. These deposits exhibit strongly directional permeabilities, generally northwest-southeast. Normal to permeabilities, generally northwest-southeast. Normal to this, the permeability is reduced to one-fourth or less of the dominant permeability. The East Flat Queen sand is rather fine-grained argillaceous, becoming increasingly anhydritic and dolomitic away from the central portion until reservoir development ceases to occur. Permeability and porosity pinchouts control the reservoir limits. However, northwest of Well 838 (Fig. 1) and along the southwest boundary, little geologic control is available and some water influx may have occurred. Fig. 3 shows a typical log section. The presence of gypsum and dolomite above and below the sand is evident from visual and other observations. X-ray diffraction analysis of producing sands gave 80 percent quartz and 10 percent feldspar as well as varying amounts of illite and montmorillonite clays. JPT P. 1359

Summary The McElroy field in west Texas is an example of a carbonate reservoir that contains significant quantities of gypsum. Large amounts of reservoir data have been obtained in this field by conventional core analyses and by sidewall neutron porosity (SNP) logs. These data do not account for the effect of gypsum in porosity determination. In an effort to characterize this heterogeneous reservoir more thoroughly, we used special low-temperature core analysis and developed a new method of log analysis to correct old core and log porosity data for gypsum content. These data then were used for a more accurate determination of OOIP. Introduction The McElroy field is one of the major reservoirs of west Texas. It is located in Crane and Upton Counties (Fig. 1). The reservoir encompasses about 30,000 acres (121.5×106 m2) and contains about 2,000 wells. Production is from the Grayburg-San Andres dolomite formation of the Permian period at an average depth of about 3,000 ft (900 m). The field was discovered in 1926 and has been in various stages of development since that time.1 Most of the porosity data acquired have been in the form of conventional core analyses and SNP logs. These data are known to be affected significantly by the presence of gypsum. Since this reservoir rock has been found to contain significant amounts of gypsum, direct use of these data would lead to significant errors in porosity estimation. In early 1959, a method of core analysis was developed that prevented the dehydration of gypsum.2 As new infill wells were drilled, selected wells were cored, and the cores were analyzed by this technique. To utilize as much data as possible in this very large reservoir. Many of the more recently drilled wells were logged with SNP logs.3 To utilize these data, a method of correcting these logs was developed. The corrections were based on information obtained from special log analysis techniques in nearby infill wells. Porosity data from logs and cores in the new wells and corrected porosity data from logs and cores in the old wells were used in conjunction with capillary pressure measurements to determine OOIP by the volumetric method. Theory and Definitions Core Analysis Gypsum (CaS04·2H20) is an evaporite mineral found in many carbonate and some sandstone reservoirs. In laboratory measurements, gypsum has been observed to convert to anhydrite and water at temperatures above about 140°F (60°C). Assuming that the conversion from gypsum to anhydrite takes place at approximately this same temperature in the reservoir and that a typical geothermal gradient exists, gypsum general1y is not expected to exist below a depth of about 6,000 ft (1800 m). Standard core analysis can cause serious errors in the measurement of porosity if gypsum is present in the cores. Vacuum distillation or retort procedures yield porosity values that are too high by as much as 48 % of the bulk volume of gypsum present. Heating the core samples for cleaning or measurement of fluid content is a necessary step in most forms of standard core analysis. The heat applied is high enough for the gypsum to lose its crystallization water and to form anhydrite. It is the loss of this crystallization water that causes the error in the measurement of porosity. This process was described in detail by Hurd and Fitch2 in 1959. Core Analysis Gypsum (CaS04·2H20) is an evaporite mineral found in many carbonate and some sandstone reservoirs. In laboratory measurements, gypsum has been observed to convert to anhydrite and water at temperatures above about 140°F (60°C). Assuming that the conversion from gypsum to anhydrite takes place at approximately this same temperature in the reservoir and that a typical geothermal gradient exists, gypsum general1y is not expected to exist below a depth of about 6,000 ft (1800 m). Standard core analysis can cause serious errors in the measurement of porosity if gypsum is present in the cores. Vacuum distillation or retort procedures yield porosity values that are too high by as much as 48 % of the bulk volume of gypsum present. Heating the core samples for cleaning or measurement of fluid content is a necessary step in most forms of standard core analysis. The heat applied is high enough for the gypsum to lose its crystallization water and to form anhydrite. It is the loss of this crystallization water that causes the error in the measurement of porosity. This process was described in detail by Hurd and Fitch2 in 1959.

Summary This paper describes the results, operational problems encountered, and the ongoing technical evaluation of an in-situ combustion pilot project in the North Ward-Estes field in Ward County, TX (Fig. 1). This sandstone reservoir was discovered in 1929 and has been under waterflood since 1955. Even though the reservoir characteristics were marginal in comparison with the screening criteria for in-situ combustion, laboratory tests indicated that this process could be viable. Introduction The North Ward-Estes field is an anticlinal structure of lagoonal deposition 18 miles (28.9 km) long and 4 miles (6.4 km) wide, trending north to south along the western edge of the central basin platform. The Yates formation is located at a depth of 2,400 to 2,800 ft (732 to 853 m). This Permian age reservoir consists of seven sand members trapped by dense dolomites and is limited a really by shales and evaporites. Approximately 10% of the rock volume consists of clay: montmorillonite, 80 to 85%; illite, 5 to 10%; chlorite, 1 to 5%. Some minor evidence of fracturing has been indicated, and a northwest-southeast directional permeability is also present. The G.W. O'Brien Well 4 was drilled in April 1929 by Gulf Oil Exploration and Production Co. After this discovery well, approximately 2,000 wells producing from the Yates and Queen formations were drilled over an area of about 30,000 acres (121 × 10(6) m2). By 1955 most of the primary oil was depleted by solution gas drive. At that time, a waterflood was initiated that still remains in operation. Production and injection have been from open hole multi sand completions, which do not lend themselves to reliable calculations of oil saturation. Approximately 350 million bbl (55 × 10(6) m3) of oil have been produced from this reservoir since discovery. A dry forward in-situ combustion pilot was initiated on April 11, 1978. This project, known as the Section 10 combustion project, was implemented in the J-2 sand, which is the lowest of seven Yates sand members (Fig. 2). The rock and fluid properties are given in Table 1. Comparison of project parameters with various screening criteria indicated that the reservoir oil saturation was marginal (Table 2). All other parameters were within the minimum criteria. Laboratory experiments using produced fluids and preserved core demonstrated the applicability of an in-situ combustion process in this watered-out reservoir. We initiated a pilot project to substantiate these tests and to answer four major objectives:Can ignition be obtained and can combustion be propagated in a light-gravity crude?Can flue gas be isolated from other producing sands?Can produced flue gas be disposed of readily?Will the oil-production rate be increased? This pilot was implemented in two phases. We conducted Phase 1 in an inverted 10-acre (40 469-m2) five- spot pattern. High O2 and low CO2 concentrations in gas samples from offset producing wells indicated marginal bum quality. Rapid flue gas migration and the evidence of a low-quality bum caused this phase to be terminated in April 1979. In Nov. 1978, we began Phase 2 in an inverted 40-acre (161 875-m2) nine-spot pattern west of the Phase 1 pilot in a location having a thicker sand section. The initial O2 and CO2 concentrations in the produced flue gas indicated a better burn quality in this area. An observation well was dulled and cored behind the calculated location of the combustion zone to determine whether ignition was obtained and whether combustion propagation was possible in this formation. Another observation well was drilled and cored in an attempt to locate the leading edge of the combustion zone and to determine whether an oil bank was developing. JPT P. 2244^