A Numerical Simulation Comparison of Five-Spot vs. Line Drive in Micellar/Polymer Flooding

The work presented in this paper was undertaken to study the effect of pattern configuration on oil recovery by the Maraflood oil-recovery process. The patterns studied are the five-spot and the 4 × 1 line drive. These patterns are obtained by placing infill wells in an existing 10-acre (40 469-m2) waterflooded five-spot pattern to obtain the 2.5-acre (10 117-m2) patterns. The number of infill wells is the same for both the new five-spot and new line-drive configurations and is about three times the number of existing wells. Both patterns have been used successfully in field applications by Marathon before this study. For instance, a line-drive pattern was used in Project 119-R and a five-spot pattern was used in Project 219-R. This work shows that the line drive produces more tertiary oil than the five-spot under otherwise identical reservoir conditions. Breakthrough times and oil rates for line-drive production wells are nearly the same. Meanwhile, five-spot production wells have vastly differing oil breakthrough times and oil rates. Both of the latter effects result from a nonuniform distribution of waterflood residual oil saturation in the field. Our study also shows that if producing wells in each line-drive row are connected by a perfect vertical fracture and if the same is true of the injection wells, the line-drive efficiency will improve very little. Introduction The Maraflood oil-recovery process is a viable enhanced oil-recovery technique. An appraisal of this process and other surfactant-enhanced oil-recovery schemes was reported by Gogarty. Three significant field tests of the Maraflood process were reported by Earlougher et al. In addition, a large-scale field application of this process was presented recently by Howell et al. in field applications of the Maraflood process, both line-drive and five-spot configurations have been used. In our field experience, an existing five-spot waterflood pattern is convened to another five-spot or 4 × 1 line-drive configuration by adding infill wells. The new five-spot or line-drive pattern has an area-per-well spacing of one-fourth of the original waterflood spacing. In practice, the number of infill wells required for both cases is somewhat greater than three times the number of existing wells. As the total number of wells increases, this ratio approaches the theoretical limit of three. In addition to the preceding arrangements of infill wells, many others are possible. In some arrangements, fewer infill wells are required than in our five-spot and 4 × 1 line drive. In such cases, the area per well increases, which generally causes these problems:required injectivity per injection well increases and may not be attainable because of the high viscosity of the injected fluids andthe breakthrough time is delayed. As an example, consider the case where no infill wells are drilled. In addition to the two problems just listed, the micellar/polymer flooding scheme will sweep only those regions that already have been swept well by the waterflood. The regions left unswept by the waterflood also will be left essentially unswept by the micellar/polymer flood. This means that a substantial amount of oil is left in place. Therefore, these types of undesired patterns were not considered in this study. Patterns with more infill wells than those in this study were not considered because of current economic limitations. Because of the likelihood of economic and technical merits, we also considered the placement of long vertical fractures to connect existing waterflood wells in place of infill wells. The fractures were arranged to form a more effective line drive. We emphasize that the patterns studied in this paper are those usually used in micellar/polymer flooding applications. Muskat has reported breakthrough waterflood sweep efficiencies of 72% and 88% for five-spot and 4 × 1 line drive patterns when the mobility ratio is unity. Muskat's results are for ideal plug flow displacement of red water by blue water in a perfectly homogeneous reservoir. SPEJ P. 69^