Summary Bretagne has experimented successfully with CO2 huff ‘n’ puff injection since 1985 in the Shoemaker field in Kentucky. Numerical simulations were performed with a compositional model to review and to quantify the influence of all parameters that could be responsible for the production improvement observed in the wells. Laboratory studies were conducted to get relevant physical data, particularly concerning relative permeabilities with three phases present. These data were integrated into the numerical model, giving good agreement with field results, and predicted production performance more accurately. Finally, the optimization of the huff ‘n’ puff process in Shoemaker is presented.
Introduction Cyclic CO2 injection was initially considered a cheap operation that was useful before performing a CO2 displacement pilot or merely for increasing well productivity. The operation consisted of injecting CO2 into the reservoir, then resuming production from the same well after a shut-in period during which CO2 was dissolved in oil.
Oil production reported in some projects by this CO2 huff ‘n’ puff process, combined with low cost, led operators to consider the technique as an attractive EOR method.
Because lowering oil viscosity is a function of the dissolved CO2 concentration and this mechanism plays the major role in the stimulation potential, field applications were first performed in viscous-oil reservoirs with the more conventional steam huff ‘n’ puff process. Application to recovery from light-oil reservoirs began in 1984. Haskin and Alston presented results of 28 Texas projects in relatively deep (2,600 to 7,800 ft), permeable (400 to 1,700 md), pressured reservoirs (500 to 1,600 psi). Oil viscosity in these reservoirs was between 2 and 33 cp, and water cut was zero. Their study indicated that incremental oil production from CO2 stimulations was the result of oil swelling and viscosity reduction mechanisms. Palmer et al. described the results of two successful CO2 cyclic injections in southern Louisiana in 1983. Again, the reservoirs selected were deep (10,000 ft), with a reservoir pressure close to the minimum miscibility pressure (MMP). No interpretation was developed, but some combination of oil swelling, viscosity reduction, and lowering of interfacial tension (IFT) appeared to be responsible for the success of the operation. In addition, the two candidate wells were perforated immediately above the water/oil contact (WOC); therefore, CO2 injection could have acted as a water-shutoff treatment. Simpson described two CO2 stimulation experiments in a bottomwater-drive reservoir where water coning caused low sweep efficiency. CO2 increased the oil cut during 2 months. Incremental oil production was reported to be the result of oil swelling and oil viscosity reduction effects on fractional flow.
Studies by Monger and Coma, Monger et al., Thomas et al., and Thomas confirmed that, in contrast to heavy oil, the mechanisms of light-oil recovery during CO2 stimulation should include: oil swelling, altered relative permeability from drainage/imbibition hysteresis, altered relative permeability from wettability change, IFT reduction, and hydrocarbon vaporization. Oil swelling was considered to be the most important factor for pressure-depleted reservoirs, while wettability change, IFT reduction, and hydrocarbon vaporization should be beneficial if near-miscible or miscible conditions could be reached.
Field Description The 40,000-acre Shoemaker Ridge field development area in Lee County, KY, operated by Bretagne, represents the southeast extension of the Big Sinking field.
The field was discovered in early 1900 and produces from three 800 to 1,200-ft-deep Silurian age zones. These three zones are known locally as the Corniferous first, second, and third pays and correspond to Upper and Lower Lockport and Keefer, respectively. The first pay is dolomitic and tight in the Shoemaker Ridge area where the reservoir is restricted to the second and third pays.
The second and third pay zones are basically sandstone depositions, which are calcitic or dolomitic in varying degrees and appear to be continuous over the entire field. In the Shoemaker Ridge area, the second pay zone is a microvugular sandy dolomite with occasionally dispersed shale. Second pay gross thickness is 20 to 30 ft.
The third pay zone is a dolomitic sandstone in the same porosity range as the second pay zone but generally is more fine-grained and has lower permeabilities. The gross thickness of the third pay is 30 ft in the Shoemaker Ridge area. The third pay is separated from the second pay in some places by a thin shaly barrier. However, a vertical hydrodynamic communication seems to exist between the second and third pay zones.
Because the area underwent many periods of extreme tectonic activity, the Shoemaker Ridge reservoirs have been faulted. Faulting has caused permeability and fluid flow restriction. Table 1 gives reservoir rock and fluid properties.
Production History Primary Production. More than 800 producing wells have been drilled in the Shoemaker Ridge area, giving an average areal spacing of 400 ft in developed areas. Every well was fractured to increase productivity. A high water cut (90%) was measured as soon as the wells produced because a mobile water saturation existed over the entire structure. This water cut remained steady with time and was not related to the withdrawal rate. Although a WOC was not identified, water saturation increased downdip, and mobile water saturation could be the result of capillary pressures. Mobile water saturation was also interpreted as the result of oil migration still taking place in this geographic area.
Figs. 1 and 2 represent a lease production in the Shoemaker field. A steep decline (caused by low permeability) from 20 to 1 bbl/D-well during the first 3 to 4 months is followed by a more gradual decline because of the influence of liberated gas.
Waterflooding. Initially, the main objective was to maintain pressure to increase oil recovery. Three pilots were designed in 1984 to experiment with waterflooding. Incremental oil recovery was expected to be similar to that obtained in other areas of the Big Sinking field, but performances were poor. Well fracturing that could have interconnected wells and thus caused nonreversible poor sweep efficiency was thought to be the reason for the poor performance.
p. 92–100