Residual oil saturation is an important parameter in evaluating tertiary oil potential for a depleted reservoir. Among various methods, single-well tracer testing has been widely accepted as a standard method for measuring residual oil saturation to waterflood. The test involves the injection of a reacting tracer into a reservoir and matching of the tracer production profiles using a simulator. Of the two existing simulators for single-well tracer test interpretation, Exxon's single-porosity model was developed for homogeneous sandstone reservoirs. It can handle fluid drift but fails to model the dead-end porosity that is typical of carbonate reservoirs. Conversely, Deans' double-porosity model allows transfer of tracers between the flowing and non-flowing pores but cannot account for fluid drift. It also suffers from the drawback of simulating physical dispersion by numerical dispersion, i.e., by varying the grid size. Considering the limitations of these two models, there is an urgent need to develop a new general simulator by combining the specific features of these two models for direct application to carbonate reservoirs with strong fluid drift and severe dispersion. This paper described the formulation of a new double-porosity model, which has encompassed the mechanisms of physical dispersion and fluid drift, as well as a discussion of the characteristics of the model.
The amount of oil left in a reservoir after secondary operations is needed to evaluate the potential of enhanced oil recovery processes. Of all the methods available to date1,2, SWTT is unique in its large and variable depth of investigation, and relatively free of the near well-bore effect. SWTT is still the most widely accepted method in the industry for measuring residual oil saturation, though an increasing number of successful interwell tracer testing has been reported recently3,4. The single-well tracer testing method for measuring Sorw, which has been described fully by Deans5–7and Tomich8, is based on chromatographic separation of partitioning and nonpartitioning tracers in the test zone, a phenomenon well understood in chemistry for many years. Although single-well tracer test has been well demonstrated for homogeneous sandstone reservoirs, it has some shortcomings when applied to carbonate reservoirs owing to the complicated pore structure and microscopic conformance, which could render long production tails and extreme dilution of the tracer curves.
Two models, namely a double-porosity model9 where tracer could distribute between the flowing and nonflowing pores through mass transfer and a single-porosity model10,11 where a fictitious water drift rate was assumed in the test zone were used to cope with the abnormal dispersion exhibited in the carbonate SWTT. The single porosity model can handle fluid drift but fails to model the dead-end porosity that is typical of carbonate reservoirs. Conversely, Deans' double-porosity model allows transfer of tracers between the flowing and nonflowing pores but it cannot account for fluid drift. It also suffers from the drawback of simulating physical dispersion by numerical dispersion, i.e., by varying the grid size.