Steam-Foam Pilot Project in Dome-Tumbador, Midway-Sunset Field

Abstract

Summary. This paper describes a steam-foam pilot project in the Potter sand, Midway-Sunset field. The pilot consists of four inverted five-spot patterns with a confined producer covering 5.2 acres [2.1 ha]. Steam foam was generated by continuous injection of steam with NaCl, alpha olefin sodium sulfonate, and nitrogen. Production and subsurface data, obtained from two observation wells, were used as monitoring tools in the pilot. Overall, during the first 2 years of foam injection, 207,000 bbl [32 900 m3] of incremental oil was produced. Introduction Thermal recovery is the most commercially successful EOR method in use in heavy-oil reservoirs. Both steam soak and steamdrive have been used for many years. The efficiency of steam recovery is adversely affected by two factors: the presence of high-permeability channels and gravity segregation of the injected and displaced fluids. High-permeability zones tend to act as steam thieves. Once swept of oil, the pressure drops between the injector and the producer, and these zones receive most of the injected steam. The less-permeable, oil-saturated beds remain virtually untouched by continued steaming. In reservoirs with little or no dip, the relatively low-density steam rises to the top of the reservoir, forming a channel beneath the caprock to the producer. As in high-permeability zones, little pressure differential exists between injector and producer once steam breakthrough occurs. Steam then follows this established path. During maturity, a combination of heat losses to the overburden and low pressure drop impair the process efficiency. A solution to the channeling is to decrease steam mobility by combining steam, a noncondensable gas, and a foaming agent simultaneously to generate a foam in the reservoir. Use of foam for gas diversion in porous media was reported as early as 1958.1 The first field application of foam was reported in 1970. Dilgren et al., Brigham et al., Ploeg and Duerksen, and Mohammadi and McCollum reported on the use of foam in reservoirs under steam-drive. The main differences between these studies are reservoir char-acteristics, mode of injection (i.e., slug vs. continuous), duration of the injection phase, and chemical structure of the foaming agent. A close inspection of these projects shows a need for identification of a mechanism describing fluid displacement by steam foam. This mechanism can assist in process optimization and differentiation of the so-called "accelerated recovery" from true "incremental recovery." Another equally important need is accurate determination of fluid production from the pilot producers. Brigham et al. 4 emphasized this point relative to their pilot project in the Kem River field. One of the recommendations of their study was the use of a production test tank and a sampling tube for measurement of oil and water production. Another recommendation was the use of neutron logs in observation wells for determination of liquid displacement. Our laboratory studies have shown that foam appears to propagate as a front through the porous medium. The pressure drop behind the front is considerably higher than the pressure drop ahead of the front. Additionally, the quantity of dilute (i.e., 0.5 to 1.0 wt% active) foaming agent required to sustain foam propagation is greater than 1 swept PV. Because of these results, we chose to inject the chemical continuously in the pilot test. The injection procedure used was similar to the one described by Dilgren et al. We also adopted the production sampling techniques described by Brigham et al. The primary objective of the pilot project was to correct gross steam override in a thick interval and improve the drive. A secondary objective was to find an acceptable mechanism to describe liquid displacement by steam-foam propagation. Laboratory Studies A number of foaming agents were tested for their ability to improve preferential oil recovery from a low-permeability sandpack in a parallel-flow apparatus. The experiments were conducted with Dome crude under reservoir conditions. Details of the experimental apparatus and procedures are discussed elsewhere. 6 As a result of these experiments, Enordet AOS-1618 TM, a C 16–18 alpha olefin sodium sulfonate, was selected as the most suitable foaming agent for the conditions of the pilot. Additional tests were conducted for identification of foam propagation characteristics in an oil-free porous medium. These tests involved the continuous injection of water and nitrogen followed by foam into a 2-ft [0. 6-m] -long Dundee core. Dundee cores were selected because of their high degree of consolidation, high permeability, and uniform structure. These cores are particularly suited for high-velocity comparative core-flow studies of foaming agents. In a typical run, water and gas were injected at 40 and 360 mi/h, respectively, into a core saturated with water. The permeability and porosity were 2,150 md and 38% PV, respectively. At a steady-state water saturation of about 0. 72 % PV, AOS was added to the flowing water and allowed to generate a foam in situ. Four intermediate pressure taps were used for continuous monitoring of the pressure drop across the various lengths of the core. An example of the results obtained from this experiment is presented in Fig. 1. While water and nitrogen were being injected, the pressure drop stabilized at a relatively low level. After the start of foam injection, the pressure drop across the first 6-in. [15-cm] -long section of core increased gradually. This increase was followed by an increase in the second 6-in. [1 5-cm] -long section and so forth through the third and fourth sections of the core. The results indicated that in the absence of oil, foam propagation, as determined by the pressure gradient, appeared to proceed as a front. Field Description and Steamflood History The Potter interval is an upper Miocene sand that has produced nearly 11x10(6) bbl [1.75 × 10(6) M3] oil at Dome-Tumbador. It constitutes a series of stacked, submarine fan channel complexes that have been divided into four productive zones, These zones, Zones A through D, are characterized by poorly sorted sands ranging from silt size to gravels and are separated by intraformational silts or relatively impermeable sand layers that exhibit low resistivity on dual induction logs. These relatively impermeable layers are laterally continuous across Dome-Tumbador and partially restrict vertical fluid migration within the formation. The Potter sand at Dome-Tumbador dips 14 to 18 degrees to an apparent oil/water contact on the eastern portion of the properties. Analyses of individual beds within the Potter sand and overlying formations indicate a uniform decrease in dip angle with decreasing depth. Fig. 2 shows the structural contours on top of the Potter sand. The steam-foam pilot area consists of four inverted five-spot patterns including nine producers and four injectors. The surface locations of the pilot wells are included in Fig. 2. Fig. 3 shows a three-dimensional perspective of the pilot. In addition, Table 1 lists key reservoir characteristics of the Potter sands in the pilot area. The pilot area has undergone steamdrive since 1977. Cumulative production before foam addition was 1.96 × 106 bbl [311 600 M3] of oil from the nine producers, including 878 × 103 bbl [139 600 M3] from within the pilot area. SPERE P. 7^