Waterflood Performance of a Shaly Sand Reservoir, Smiley-Dewar Field, Saskatchewan

Abstract

The recovery from a waterflood in the Viking sand could be predicted within reasonable limits because fairly complete reservoir descriptions were available. Lack of data on residual oil saturations for the full range of sand conditions, however, made it necessary to apply a reduction factor to the floodable pore volume to make the results of theoretical waterflood calculations acceptable. Introduction The producing interval in the Smiley-Dewar field in western Saskatchewan is the Viking sand. It is of Cretaceous age and lies at a depth of about 2,400 ft. The oil pool includes an area of about 6,000 acres developed on 40-acre spacing, and the pay interval with an average thickness of about 13 ft is characterized by shale content of varying degrees, low permeability, and high water saturation. A 2,000-acre extension to the south, proved by step-out drilling after 1970, is not included in these discussions. The field began producing in the latter part of 1953. Individual wells declined steadily in rate, mainly as a result of increasing GOR's which reached an average of about 3,200 cu ft/bbl before the beginning of water injection in Feb. 1964. Cumulative oil production from the Viking sand, excluding the Viking chert zone in the lower part of the formation, was about 7.2 million bbl when injection began. Recovery by natural depletion of the solution-gas-drive reservoir was estimated at 10 million bbl of 33 degrees API crude. Detailed reservoir description indicated this recovery to be only 16 percent of the stock-tank oil originally in place. It became obvious in the early life of the field that some enhanced recovery method would be required to complete the development program. In early reviews of performance, it was established that initial costs would have to be kept at a minimum and that if the sand was floodable a conventional water injection project on an inverted nine-spot pattern would best serve the needs of the field. The initial investigations of waterflooding feasibility included waterflooding tests on sand samples, an evaluation of water supply, injectivity tests on wells, and a waterflood pilot test. Although theoretical waterflood calculations were based on a detailed reservoir description, the recovery predictions were too optimistic and were considered unacceptable as a prudent or realistic basis for analyzing the economic feasibility of the project. To our knowledge, case histories of waterfloods in similar sands were not available. But by analysis of available data, the ultimate recovery of the waterflood project was estimated to be 23 million bbl of stock-tank oil, or approximately 36 percent of oil originally in place if produced to a water cut of 95 percent. It was agreed among the four operators that the waterflood development and operation would be carried out under a cooperative plan. Field-wide water injection was begun in Feb. 1964 when the average producing rate was about 1,600 BOPD at an average GOR of 3,200 cu ft/bbl before oil wells were converted to injectors. Cumulative oil production at that time was 7.2 million bbl. Peak production at that time was 7.2 million bbl. Peak rate reached 2,970 BOPD in 1967. Cumulative oil production as of Dec. 31, 1972, was 14.3 million bbl production as of Dec. 31, 1972, was 14.3 million bbl when the rate was 1,861 BOPD with 2,366 BWPD. Projection of the existing producing trend suggests Projection of the existing producing trend suggests that the above calculated oil recovery can be attained only if waterflood operations continue for another 40 years. JPT P. 1375