The displacement of oil from a consolidated sandstone core has been investigated using microemulsion slugs located in a single phase region. Phase diagram was prepared at the salinity lower than the optimum salinity of the system. Formation of an oil bank and resulting fractional flow of oil was dependent upon the oil content and size of the microemulsion slug. For a low salinity polymer buffer, the sulfonate adsorption was low but no increase in oil recovery was obtained. Breakthrough of the oil bank occured at 0.135 to 0.25 injected pore volumes which was followed by a white macroemulsion. The displacement process has been discribed qualitatively on the basis of analysis of the produced fluids. Screening criteria based on solubilisation parameters and fluid viscosity measurements were found to be adequate for the selection of surfactant and alcohol. Polyacrylamide was found to be inadequate. A polysachharide polymer provided adequate mobility control acheiving a higher effective viscosity in consolidated core than that calculated from simplified models. The capillary number concept was investigated over a wide range of velocity, viscosity and interfacial tension. Release of oil at the onset of mobilisation was governed by the interfacial forces and by viscous forces at ultralow interfacial tension. Dispersion in single phase flow through sandpacks and consolidated core were carried out for both saturated and partially saturated condition. Surfactant, sucrose and tritiated water were employed as dispersion tracers. Longitudinal dispersion coefficient, D1, increased with increasing sandpack permeability but was independent of sulfonate and alcohol concentration, salinity gradient, interfacial tension, sulfonate solubility and for a limited range of favourable viscosity ratio. increased with increasing immobile oil saturation in sandpacks but there was no effect in the consolidated core. The breakthrough curves for sandpacks showed good agreement with the prediction of a convective-diffusion model while the consolidated core exhibited a slight capacitance effect. Convective-diffusion and capacitance models were solved numerically using a Crank-Nicholson finite difference technique with different boundary conditions. The former exhibited a small measure of sensitivity with respect to boundary conditions used but the latter was fairly insensitive to change in boundary conditions.
|Date of Award