Experimental Investigation of Foamy Oil Flow Using a High Pressure Etched Glass Micromodel

Abstract

Abstract A series of flow visualization experiments were carried out using a high-pressure etched glass micromodel to gain insight into the pore level processes involved in foamy oil flow. The micromodel incorporated a realistic heterogeneous pore network with well-defined pore size distribution and pore throat size distribution. Solution gas drive experiments were conducted using a crude heavy-oil, a deasphalted fraction of the same crude oil, a synthetic mineral oil and a much lighter crude oil. The experimental results show that the rate of pressure drawdown was the most important parameter that altered the flow behaviour in the pore scale level and induced "foaminess" during the solution gas drive process. The dispersed gas flow occurred only in high rate tests and the dispersion was created by break-up of mobilized gas ganglia. Mathematical expressions for nucleation rate were derived for various oil samples. A metering section at the downstream end of the micromodel was used to measure the volume of fluids expelled from the pore network. These volume measurements were used to estimate the total compressibility of the reservoir fluids before the formation of visible bubbles. The compressibility numbers were used to infer the presence or absence of micro-bubbles that would be too small to be seen but could contribute significantly to oil recovery. The estimated compressibility values suggest thatt some microbubbles were perhaps evolved during the depletion process. However, it appears that most of these microbubbles remained attached to the pore walls; only a handful became detached and grew into larger bubbles. Introduction Foamy oil flow is considered to be an important contributing mechanism in the better than expected performance of solution gas drives in many Canadian and Venezuelan heavy oil reservoirs. The foamy flow occurs when the solution gas released during the depletion is able to flow through the sand while remaining dispersed in the oil. In laboratory depletion tests, it occurs when a high enough rate of pressure decline is used. Several laboratory studies have reported a dramatic effect of depletion rate on the performance of solution gas drive in heavy oil systems (Handy, 1958; Sheng et al.,1999; Pooladi-Darvish and Firoozabadi, 1999; Bayon et al. 2002). The recovery factor are reported to be much higher when higher rates of pressure decline are used. Several theories have been postulated to explain this dependence of recovery factor on rate of pressure decline (Firoozabadi, 2001; Shen and Batycky, 1996; Maini 1999; Smith 1988). The most plausible explanation appears to be based on the formation and flow of a gas-in-liquid dispersion that is often referred to as "foamy oil" (Maini, 2001). The dispersed flow of gas delays the formation of a continuous gas phase that would normally be able to flow at a higher rate and will eventually result in rapid depletion of the reservoir energy. This ability of gas to flow while remaining dispersed in the oil appears to be the mechanism that keeps the gas mobility low during depletion. The factors responsible for creating dispersed flow of gas are not well understood. Several authors have suggested that such dispersions are formed by nucleation of a very large number of bubbles (Arora and Kovscek, 2001; Claridge and Prats, 1995, Smith 1988). The term "explosive nucleation" has been employed by some to dramatize the situation (Gelikman et al., 1995). What happens beyond the nucleation stage has not been fully delineated. It has been suggested that these bubbles remain smaller than the pore-throat size and are produced with the oil (Smith, 1988). An alternate explanation considers the foamy oil flow to be simply a case of two-phase flow at high capillary number where the viscous forces are high enough to mobilize isolated gas ganglia (Maini, 2001).