Relative Permeability Estimation in Steam Assisted Gravity Drainage (SAGD) Using a Fractional Flow Model

Morte, M. (Texas A&M University) | Hascakir, B. (Texas A&M University)

OnePetro 

Abstract

Enhanced oil displacement in a reservoir is highly affected by wettability alterations in conjunction with the lowering of viscosities during steam assisted gravity drainage (SAGD) for bitumen extraction. The impartation of energy in the form of heat to the fluid by injecting steam triggers an alteration to a more water-wet state during SAGD. However, the presence of three distinct phases in the reservoir has implications for the effective modeling of the complex fluid dynamics. Dependency of the relative permeability endpoints on the temperature realized as a function of the introduction of steam is difficult to model. Optimization of any steam process requires simulation in order to adequately characterize years of flow and so a model that is capable of representing three phase flow is necessary. To obtain this a pseudo-two phase relative permeability is proposed that assumes fractional flow theory is valid and treats the experiments as a waterflood.

In this study, experimental recovery data for two SAGD experiments and one hot water flood are empirically matched by manipulating relative permeabilities. The analytical approach implemented allows for the representation of fluid flow in the reservoir by achieving a pseudo-two phase relative permeability that results in comparable performance to the experiments. Waterflooding techniques were utilized which allowed for the negation of the steam phase in the model and so two-phase flow was established.

The sensitivity of the relative permeability curves to temperature change results in the inability to formulate a generic three-phase curve and so the pseudo-two phase curve is valuable for the purpose of simulation. The methodology presented enables the formulation of a simplified relative permeability that is unique to each process used and in that specific location. The model that was established was validated and proven credible by the good match with the experimentally obtained values.