ABSTRACT

The processes of migration and entrapment of multiphase hydrocarbons are poorly known. Key issues include (a) estimating reasonable rates of charge under geologic conditions; (b) the effect of varying GOR on hydrocarbon retention in the reservoir, and the distribution of trapped oil vs. gas; and (c) the effect of fault permeability.

To gain insight into the above issues, we have used off-the-shelf reservoir simulation software (Eclipse) to examine simultaneous migration of oil and gas through a series of stacked permeable reservoirs separated by impermeable shale and bounded by faults that may act as conduits, baffles, or barriers to flow (Figure 1). A similar reservoir geometry is observed throughout the Ship Shoal 274 field, Louisiana OCS, Gulf of Mexico. The migration history in Ship Shoal 274 consists of an early pulse of oil and gas which charges water-saturated reservoirs, followed by a late gas pulse which migrates into oil-charged reservoirs. Appropriate reservoir and seal properties are taken from the Ship Shoal 274 data. Composition of migrating fluid is varied in our model from black oil (charging a water-saturated reservoir) to dry gas (charging an oil-saturated reservoir) to mixed oil and gas of varying GOR (charging a water-saturated reservoir). These scenarios reflect, in a simplified way, phases in the migration history of Ship Shoal 274. Fault properties are also varied in the model; fault permeability ranges from 0.1 to 1000 md, and appropriate porosities and capillary pressure profiles are assigned for each permeability.

Figure 1. Model geometry and properties. Fault width is exaggerated for illustrative purposes. Boundary conditions are described in text. Reservoir permeability (k) is on the order of 100 md for all models, reservoir porosity is 30%, and reservoir capillary pressure is shown at right as capillary pressure in a mercury-air system (pcma) with respect to water saturation (S_w). For the shale seal, permeability is 10^-6 md and porosity is 15%. Fault permeability varies from 0.1 ("low k") to 1000 ("high k") md, porosity varies from 15 to 30%, and variability in capillary pressure is shown at right.

Our results suggest that the total time to fully charge all reservoirs with gas, oil, or a mixture of both, is less than 20000 years under most conditions, given a sufficient source. We have assumed that flow is driven by at least a small degree of overpressuring; migration due to buoyancy alone would take considerably longer. For our model, gas migrates into the reservoir at a rate of on the order of tens to hundreds of mcf/day. Oil charges the reservoir more slowly than gas, perhaps at a rate of tens of STB/day or less.

To test the distribution of oil and gas resulting from two-phase migration, we varied the GOR of the migrating hydrocarbons while keeping the migration rate and fault and reservoir properties constant. At low GOR's, hydrocarbons are somewhat segregated so that gas is chiefly found in the upper reservoir (Fig. 2). At higher GOR's, gas also charges the lower reservoir (Figure 3), but even at very high GOR's, the theoretically maximum obtainable gas saturation is not achieved in either reservoir, even after several hundred thousand years.

Preliminary results indicate that fault properties affect hydrocarbon migration to varying degrees, depending on the hydrocarbon type. In general, there is a threshold permeability above which fault behavior changes dramatically from conduit to baffle or barrier. Figure 4 shows an abrupt decrease in gas saturation throughout both reservoirs when permeability of the charging fault is lowered from 1 md to 0.1 md. In a system that is being charged with oil, the change in fault behavior from conduit to barrier or baffle occurs at a higher permeability, probably between 10 and 100 md, as suggested in Figure 5. Actual threshold values in a particular field are expected to depend on a number of other conditions as well, notably fault-zone thickness and continuity, fluid properties, and the driving mechanism for migration.

End_Page 12------------------------

Figure 2. Gas saturation (Fig. 2a) and oil saturation (Fig. 2b) distribution in the same reservoirs after migration paths have been established. In this model, both faults have a permeability of 1000 md. Refer to Figure 1 for position of faults. The gas/oil ration (GOR) is 200.

Figure 3. Gas saturation (Fig. 3a) and oil saturation (Fig. 3b) distribution in the same reservoirs after migration paths have been established. In this model, both faults have a permeability of 1000 md. Refer to Figure 1 for position of faults. The gas/oil ration (GOR) is 10000.

Figure 4. Effect of varying charging (left-hand) fault permeability (k), porosity (f), and capillary pressure on gas saturation, when gas is allowed to charge an initially oil-saturated reservoir. Permeability and porosity of the right-hand fault are constant. Figure 4a shows gas saturation (Sgas) as a function of fault permeability for each of 4 cells in the model. Position of the cells is shown in Figure 4b.

Figure 5. Time to fully charge a water-saturated reservoir with oil as a function of the charging fault permeability (k).

End_of_Record - Last_Page 13-------