Theoretical Study of Water Blocking in Miscible Flooding
- Thomas Muller (BEB Erdgas and Erdol GmbH) | Larry W. Lake (U. of Texas)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- November 1991
- Document Type
- Journal Paper
- 445 - 451
- 1991. Society of Petroleum Engineers
- 4.1.5 Processing Equipment, 5.2.1 Phase Behavior and PVT Measurements, 1.8.5 Phase Trapping, 5.3.1 Flow in Porous Media, 4.1.2 Separation and Treating, 5.4.9 Miscible Methods, 1.8 Formation Damage, 5.3.4 Reduction of Residual Oil Saturation, 5.4.7 Chemical Flooding Methods (e.g., Polymer, Solvent, Nitrogen, Immiscible CO2, Surfactant, Vapex), 5.4.1 Waterflooding, 4.3.4 Scale, 1.6.9 Coring, Fishing, 5.3.2 Multiphase Flow
- 4 in the last 30 days
- 466 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
Miscible displacement processes can leave a substantial amount of residualoil behind the displacement front. This phenomenon has two general causes:instabilities caused by local phenomenon has two general causes: instabilitiescaused by local heterogeneities or viscous fingering and water blocking. Thispaper describes a study of the latter. Numerous laboratory experiments haveshown that significant blocking of oil from the solvent by mobile water canoccur in water-wet media and at large water saturations. Despite this,water-blocking studies have been limited to either simple correction functionsin numerical simulations or microscopic models. To the best of our knowledge,no explicit theoretical model considers the macroscopic bypassing andsubsequent interaction of the solvent stream with a trapped hydrocarbonphase.
In this study, CO2 is the miscible solvent. A numerical model calculates themass flux between flowing and stagnant regions separated by a water film. Themodel considers solvent diffusion and diffusional extraction of oil accompaniedby swelling or shrinking of the stagnant hydrocarbon phase. The various processparameters are represented by a set of dimensionless numbers, which allow theproblem to be scaled to any size. We show that the conditions for the problemto be scaled to any size. We show that the conditions for the mobilization ofwater-trapped oil are determined essentially by two numbers: the ratio of thestagnant water/oil volumes and the ratio of the solvent equilibrium constantsat the phase boundaries. We apply the model to published laboratory miscibleflood experiments. The water-film thickness is used as a matching parameter toreproduce closely the measured residual oil data for different floodvelocities. Results indicate that solvent diffusion and oil swelling mechanismseffectively account for trapped oil on the laboratory scale. Trapping rapidlydecreases at field scale if larger water barriers exist.
A residual oil saturation (ROS) for miscible flood processes has beenobserved under both secondary and tertiary conditions, but it is usually largerunder tertiary conditions. Work on visualization of the oil and waterdistributions for water-wet conditions indicates that trapped oil blobs areblocked from contact with the solvent stream by a continuous water phase.Salter and Mohanty I divided the flowing behavior of each phase into threecategories: flowing, dendritic, and isolated. The flowing fraction representsthe fluid in the main flow channels of a pore network. The dendritic fractionrepresents the part of the fluid volume that interacts with the flowingfraction by mass transfer but is not able to flow (the so-called"dead-end" PV). The isolated fraction is trapped completely by thewetting phase. Both the dendritic and isolated fractions contribute to thestagnant-phase saturation.
Several investigators consider the trapped-phase fraction as the major partof the stagnant phase for water-wet conditions and at large water saturations.Because of its potential significance, several attempts have been made tointroduce this effect into numerical simulation through water-blockingfunctions. These functions describe the experimental observation that theamount of trapped oil increases with increasing water saturation. Attempts toconsider time-dependent effects, such as mass transfer, on the stagnant oilfraction have been performed only for the dendritic fraction. Injecting asolvent with a finite water solubility, such as CO2 , however, might cause apenetration of the water phase by molecular diffusion. Swelling of the trappedoil and subsequent rupture of the water film could then be a possible mechanismto explain experimental data that show a significant decrease in stagnant oilsaturation with increasing solvent contact time
The objectives of this work are (1) to develop a model that combines themicroscopic mass transport into (dissolution) and out of (extraction) twoimmobile phases (a shielding water phase and a stagnant oil phase) with themacroscopic flow of water, oil, and solvent in a main flow channel; (2) tostudy the various mechanisms that influence the interaction between stagnantand flowing phases; and (3) to match measured laboratory residual oil data.phases; and (3) to match measured laboratory residual oil data. The modelformulations are generally applicable to all miscible processes, but this workdeals with only CO2 as the injected processes, but this work deals with onlyCO2 as the injected solvent.
Prior Work Prior Work Water blocking and its effect on displacementefficiency during tertiary miscible flood experiments have been observed innumerous laboratory studies. The blocking effect is significant, especially forwater-wet cores during simultaneous water/solvent injection. The major role ofmolecular diffusion is that of an important mechanism to reach, swell, andreconnect isolated droplets, but this effect is not quantified.
Shearn and Wakeman presented mathematical formulations to calculate the rateof mass transfer through a water film and the rate of the consequent movingboundary. They did not, however, present the solution. Grogan and Pinczewskideveloped a numerical model that simulated the swelling of a single trapped oildrop and compared their calculations with laboratory results from micromodels.They concluded that CO2 diffusion is an important effect in laboratorycorefloods but not in field-scale floods. Their study is restricted to a localsolution of the problem without relating it to the flowing process in the porenetwork system.
|File Size||614 KB||Number of Pages||7|