Steam-Foam Mechanistic Field Trial in the Midway-Sunset Field

Abstract

Summary A one-pattern, steam-foam mechanistic field trial was conducted in Section 26C of the Midway-Sunset field (upper Monarch sand). The test objectives were (1) to understand the mechanisms of steam diversion caused by foam under reservoir conditions, (2) to establish whether foam can exist in-depth away from the injection well, and (3) to measure incremental oil that can be attributed to foam. Surfactant was injected with steam and nitrogen continuously, and bottomhole injection pressure (BIHP) increased from 100 to 300 psig, indicating good foam generation. Better steam distribution across the injector's perforations occurred when foam was generated. Improvements in both vertical and areal sweep efficiency of steam were observed. Substantial temperature and gas saturation increases coincided with surfactant breakthrough and local reservoir pressure increases at observation wells. Complementary laboratory corefloods showed that foam generation could occur at low-pressure gradients, which are typical of in-depth conditions. Both laboratory and field data were interpreted as evidence that the in-depth presence of foam was the result of local generation wherever surfactant, steam, and nitrogen were present, rather than propagation of a foam bank: generated near the injector. Some oil-production increase was also observed during the test; however, an accurate quantitative estimate of incremental oil owing to foam was difficult to establish. Introduction Foam can be used to reduce injected-gas mobility in both noncondensable gas- and steamdrives if early gas breakthrough occurs at producing wells because of either gravity override or viscous fingering. Mechanisms of gas mobility reduction with foam have been studied both theoretically and experimentally. Specifically, such events as lamella generation,1 coalescence,2,3 and bubble trapping that establish local foam texture and hence mobility have been observed and understood in porous media under flow conditions. A critical pressure gradient, which must be exceeded for the onset of foam generation, has been measured in sandstone rocks.4 These studies, although rigorous, are not sufficient to understand the impact of foam on gas mobility under reservoir conditions fully. Many operators have field-tested foam technology with varying degrees of success. Many of these projects were conducted in steamdrives to establish the impact of foam on both the steam-swept PV and ultimate oil recovery.5–7 In the most-successful tests, surfactant was injected either continuously or semicontinuously with steam and nitrogen for at least 1 year to ensure in-depth penetration of the surfactant bank:. In these cases, increases of the steam chest attributed to foam were observed with strategically located observation wells. However, neither direct evidence of an in-depth foam bank: nor a mechanistic model of steam mobility reduction with foam could be inferred from these tests. The amount of incremental oil resulting from foam varied widely among field tests, with a surfactant requirement from ˜ 1 lbm/incremental bbl in the most successful case7 to > 10 lbm/incremental bbl in Kern River tests.6 To understand foam behavior under reservoir conditions better, we conducted a steam-foam mechanistic field trial in the Midway-Sunset field. The test objectives were to establish whether and how foam propagated away from the injection well, the impact of foam on the steam-swept PV, and whether incremental oil was produced. Prefoam Study Preparatory work was completed before foam injection. This included site selection to meet test objectives; geologic study; drilling and completion of three observation wells; surface equipment design; acquisition of baseline temperature, gas saturation, and oil production data; and surfactant selection in the laboratory. Pattern Selection. A multidisciplinary team of engineers, development geologists, formation evaluation specialists, production engineers, and researchers determined the optimum pattern location and monitored pilot performance. All fields operated by Chevron U.S.A. Inc. in the San Joaquin Valley were considered. The following four qualitative criteria were used to select the location.The site should have minimum reservoir heterogeneity to ensure that foam had a chance to alter steam paths in the reservoir and not be limited by permeability barriers or pinchouts.The site should contain the most bypassed oil reserves, ensuring that enough oil was still available for an incremental production response if foam altered steam flow.The site should have a steady producing baseline to prove that any production response was the result of foam and not of some injection or facility change. In addition, we felt that reservoir conditions would be more stable if injection and production parameters had been constant for a long time before the field trial.The site should have the lowest consistent steam quality to provide the highest water saturation near the injector for easy foam generation and propagation. The southern portion of Section 26C of the Midway-Sunset field was selected because it had been on continuous steamdrive at a consistent generator quality of 40% to 50% for ˜ 5 years. The team determined that Pattern 68BW (Fig. 1) was the best choice for the following reasons.Reservoir heterogeneity was least severe and most similar to that of a previously conducted foam pilot.8 More semicontinuous diatomite streaks were present in other patterns that might confine steam to specific channels, and oil desaturation in those channels might be significant.Estimated remaining oil reserves were highest. Reserves had been at least partially replaced by gravity drainage into the pattern.Although wellhead steam quality in Pattern 68BW was higher than in other patterns, it was still within acceptable ranges.Pattern 68BW production baseline was steadily declining, which would facilitate the estimate of incremental production. Geology. The main producing zone in the pilot area is the Miocene Monarch sand. It is an extremely heterogeneous sand-rich reservoir, ranging from silt to very coarse sands with pebbles and cobbles. Individual beds range from a few inches to a few feet and can show excellent fining sequences or massive deposition. Diatomite interbeds are usually < 1 ft thick but appear to be good barriers to steam on a localized basis. The structure in the area is controlled by an anticline and syncline whose axes trend northwest to southeast and plunge to the southeast. p. 297–304