History of the Welch Field San Andres Pilot Water Flood

A successful San Andres water flood pilot has been in operation in the Welch field for five years. Because of the unusual amount of data available, it should be possible to use its performance to test water flood predictions. One method of prediction is suggested and results are compared with actual performance. Data are given on the dilution of injected water by connate water. Since many other San Andres floods are being proposed, this paper should be useful to engineers who will be concerned with predicting their performance. Introduction The prediction of waterflood performance is always attended with a certain amount of hazard. For those floods on which history is available, reservoir data often are missing; conversely, when reservoir data are available adequate history is missing. Therefore, it seldom is possible to relate performance to reservoir parameters. All the basic data necessary for waterflood prediction have been collected on this pilot which has been in operation for five years. Although restricted by proration it is 40 per cent depleted. This is one case then where it should be possible to relate reservoir data to actual performance, and the purpose of this paper is to present one method of doing this. History and Development The Welch field pilot was initiated in 1955. It consists of one 40-acre five-spot which was formed by converting four producers to injection and drilling a producer in the center. This is not a large pilot, but a large pilot was not needed. The main purpose of the pilot, to test injectivities and obtain reservoir data, was accomplished in about six months. Efforts toward unitization were then begun, and the West Welch Unit was effected March 1, 1960. The pilot area since has been expanded, with the new injectors taking water in Nov., 1960. In the interim the pilot was produced, but at restricted rates. Five years of production history are now available. There are two San Andres reservoirs in the Welch field. The pilot produces from the upper one, which begins 300-ft below the top of the San Andres and which has a gross producing interval of 100 ft. The second reservoir is separated from the upper by a dense zone 30-ft thick. In the area of the pilot, the lower zone produces water. It is possible to frac into this lower zone, and the small amounts of water produced from all wells are attributed to communication with the lower zone through native fractures. Wells were completed by setting casing on top of the first porosity and drilling-out to the top of the dense zone. After clean-up and acidizing, production was 100 B/D on pump but declined rapidly to 10 to 15 B/D. Most of the development took place between 1945 and 1950. No logs were taken until 1955, and only a few cores were taken during 1949–50. These were analyzed by conventional methods and, although they showed more oil in place than would be suspected from the performance, permeabilities all were less than 1 md. This information stimulated interest in secondary recovery, but injectivities calculated from the core analysis were on the order of only 40 B/D/well. To test injectivities and obtain more reservoir information, the decision was made to install the pilot test herein reported. Supply water was obtained by drilling a Santa Rosa well, a new producer was then drilled in the center of a 640-acre lease, and the four surrounding wells were converted to input. These injectors were eight years' old, had an average cumulative production of 54,300 bbl and were producing 8 B/D/well when converted. Reservoir and Performance Dab The producing well was diamond-cored through the total section. Whole-core analysis is shown in Table 2. A build-up curve taken on completion showed an average permeability of 4.9 md, which compares with 6.3 md from core analysis and which was tenfold greater than previous plug-core analysis. Thus, injectivities should be in the 300 to 600-B/D range. Reservoir pressure was 377 psig in spite of the fact that surrounding offsets had been producing for eight years. A reservoir sample gave a saturation pressure of 265 psig and showed that no gas saturation had developed at this point. Later calculations estimated the gas saturation in the 40-acre area at only 2 per cent. Reservoir fluid and rock properties are shown in Table 3. JPT P. 745^