Abstract
Abstract
Matrix-fracture interaction is a commonly encountered process in fractured subsurface reservoirs during oil, gas, and geothermal fluid recoveries and contaminant transport. It occurs in counter-current manner if the interaction between matrix and fracture takes place from the same side. In the first part of this paper, the effects of the matrix shape factor, wettability, and interfacial tension on the rate of capillary imbibition and development of residual non-wetting phase saturation were studied experimentally. In the second part, secondary oil recovery by diffusion of a solvent from the samples exposed to capillary imbibition is investigated. Primary diffusion experiments were also performed on selected core sizes for comparison.
To achieve all these, experiments were conducted on Berea sandstone samples. Cylindrical samples with different shape factors were obtained by cutting the plugs 1/2, 1 and 2 inches in diameter and 2, 4, and 6 inches in length. All sides were coated with epoxy except one end. Static imbibition experiments were conducted on vertically and laterally situated samples where the imbibition took place upward and lateral directions, respectively. Brine-kerosene, brine-mineral oil and surfactant solution-mineral oil pairs were used as fluids. The rate of imbibition was evaluated for different matrix size, shape factor, wetting/non-wetting phase types and IFT. Conditions at which the shape factor controls the residual oil development were identified.
Diffusion experiments were conducted using n-heptane as a solvent. In addition to the cores exposed to full capillary imbibition, oil saturated (virgin) cores were also tested. To serve as a benchmark, uncoated versions of two different core sizes were used.
It was observed that the residual oil after capillary imbibition is controlled by the matrix shape factor for the vertically positioned samples unlike horizontal samples. Surfactant concentration plays a role on the rate of imbibition and the development of residual oil saturation.
Primary and secondary diffusion experiment showed that it could be more efficient to begin the oil recovery process by capillary imbibition (primary recovery) then continue with diffusion (secondary recovery) depending on the matrix shape factor.
Introduction
Oil recovery from naturally fractured reservoirs (NFR) is a challenging task if the matrix is dominating it and unfavorable conditions such as oil-wet rock, high oil viscosity, or low matrix permeability exist. Matrix-fracture interaction needs to be accelerated using different enhanced oil recovery techniques. If the matrix is water wet and an immiscible wetting phase exists in the fracture, the interaction occurs due to spontaneous imbibition. For weakly water-wet or oil-wet rocks, like many carbonates, injection of heat and solvents to recover matrix oil by energy and mass transfer could be the only way to recovery matrix oil, as the spontaneous imbibition does not take place and the recovery by the gravity drainage is relatively slow process. This is more prominent in case of heavy-oil.
The efficiency of any oil recovery process is evaluated by three parameters:the ultimate recovery,time to recover this amount (the rate of recovery), andthe cost of the process. Obviously, the use of solvent or heat to recover matrix oil by energy or mass transfer will be far costlier than water to recover oil by volume transfer. But, in certain circumstances listed above, expensive techniques are inevitable as cheap water or gas injection does not contribute to matrix recovery significantly.
This study focuses on oil recovery from matrix. The counter-current interaction was considered as it is commonly encountered in NFR. The model used to represent the reservoir matrix is a cylindrical rock sample of which the all sides were coated with epoxy except one end. This, in a sense, can be considered as the worst-case scenario that can be encountered in a NFR as the number of sides of the matrix contacting the fractures has been reduced to only one causing a counter-current interaction. For comparison purpose and to establish a base case, recovery from a matrix with all sides open to flow was also considered.