Waterflooding and CO2 Injection in the Naturally Fractured Spraberry Trend Area

Abstract

Naturally fractured reservoirs are prevalent throughout the world. The influence of fractures or " cracks" in the rock have increasingly become recognized as important to understand to properly develop such reservoirs. In turn, the methods to detect, characterize, and simulate fracture systems have become increasingly sophisticated. The naturally fractured Spraberry Trend Area in west Texas was discovered in 1949 and continues to produce 60 thousand barrels a day from more than 9,000 wells. The producing extent encompasses n enormous area found in eight counties. Spraberry reservoirs are producing in a 2,500 square miles range as shown in Figure 1. Spraberry reservoirs originally contained some 10 billion barrels of oil. In the 50+ years of production history, 790 MMbbl have been produced, representing a recovery factor less than 10%. This paper describes some of the historical background of Spraberry waterflooding and reports recent results of water and CO2 injection in the E.T. O'Daniel Unit. The Spraberry Trend Area has proven to be elusive to engineers since discovery. Half a century later, the reservoir has maintained its status as one of the more complicated naturally fractured reservoirs to understand or predict response to water injection. Essentially, displacement and sweep efficiency are not well understood. To this day, there is no engineering consensus regarding the impact of water injection in Spraberry reservoirs. Most operators do not believe water injection is effective; therefore, a relatively small volume of water has been injected. Originally, water and gas injection in fractured reservoirs were considered haphazard. The presence of fractures would lead to rapid channeling and excessive production of injected fluid. However, the imbibition mechanism was identified as a possible drive mechanism if injected water within the fractures contacts tight matrix rock where a majority of the oil is stored(1). Waterfloods were initiated in the 1950s to test this idea in the Spraberry Trend. Two small pilots and limited water injection were pursued, yet not much enthusiasm resulted once initial and longtem results were analyzed. Atlantic(2) conducted an early water injection test in the 1950s based on the results of an imbibition experimental effort(1). Offset wells across a lease line showed incremental oil recovery as a result of injection in three wells. However, the center test well showed no response, so the test was deemed a failure. This test corroborated the N50 ºE orientation for the natural fractures often cited in the literature. Humble Oil Co. embarked upon another 80- acre pilot(3) beginning in 1955 in the Midkiff area of Spraberry (Figure 1). Unlike the Atlantic pilot, production in the center well of a five-spot pattern increased from 70 to 250 barrels of oil per day. Humble measured the permeability anisotropy (the ratio of permeability along the fracture orientation versus perpendicular) to be 144:1. The response of this single well in the Humble pilot dictated future development of waterflood patterns(4). A large water injection project in the Driver Unit(5, 6) in the early 1960s failed to reproduce the results of the pilot conducted by Humble in the Midkiff Unit.