Abstract and Introduction

Steam Assisted Gravity Drainage (SAGD) is a steam-flood process that relies on thermal diffusion and gravity to mobilise bitumen in oil sands reservoirs and allow for recovery. A pair of parallel horizontal wells are drilled near the bottom of the pay zone with approximately 5m vertical separation between them. After an initial circulation phase to establish communication, high quality steam is continuously injected down the injection well while the heated bitumen and condensate flow, due to gravity, to the production well. The volume of low oil saturation left behind is filled with steam and referred to as the steam chamber. The steam chamber grows vertically until it hits the cap-rock at which point it grows laterally away from the horizontal wells(1).

A major cost of producing bitumen by SAGD is the energy required to produce steam; thus there is a great drive to improve the energy efficiency of a SAGD project. Solvent addition is expected to increase oil production and improve the energy efficiency of SAGD by combining the thermal process with the diluent mechanism of a solvent (2). A solvent is added to the steam chamber by injecting the solvent-steam mixture continuously from the injector well (3).

Reservoir simulations of solvent-steam co-injection were done with a pseudo-compositional thermal reservoir simulator. The effects of co-injecting a multi-component solvent, naphtha, and a single component solvent, propane or pentane, were compared to SAGD. Three pseudo-components were used to model the multi-component behaviour of naphtha.

Co-injection of any of the solvents studied (propane, pentane, naphtha) resulted in an improved SOR. The greater the mole ratio of solvent to steam the greater the improvement in SOR. Only naphtha co-injection resulted in an improved oil production rate. All components of naphtha travelled freely in the vapour chamber and accumulated along the vapour chamber front in both the vapour and oil phases. At higher solvent to steam ratios oil production rate decreased due to an accumulation of solvent gas at the vapour chamber front.

Reservoir Model Description

Simulations of a SAGD well pair were performed using a homogeneous, two-dimensional model of a vertical cross section perpendicular to the wells. Figure 1 shows the layout of the 100m by 1m by 20m model made of 1 metre cubes.

Reservoir values typical to MacKay River were used and are listed below in Table 1. The horizontal production well is in grid cell 50, 1, 18 and the horizontal injector is 5 meters above the producer in grid cell 50, 1, 13. The model is bound on the top and bottom by impermeable cap and base rock. No mass or heat transfer can occur through the lateral boundaries of the model.

The numerical models considered in this paper provide a preliminary analysis of the complex, three-dimensional system that is expected to exist in reality(4). The following assumptions were made in the course of the numerical study:

  • Bitumen is modelled as "dead" oil, existing only in the oil phase.

  • Water and oil are immiscible.

This content is only available via PDF.
You can access this article if you purchase or spend a download.