Dimethyl Ether as an Additive to Steam for Improved Steam-Assisted Gravity Drainage

Coinjection of solvent with steam results in lower chamber-edge temperatures than those in steam-assisted gravity drainage (SAGD), which enables the decrease of heat losses to the overlying formation rocks. However, use of highly volatile solvents, such as propane, can yield significantly slow bitumen production because of low chamber-edge temperatures. The suitability of alkane solvents for SAGD in terms of phase behavior has been reported to increase with increasing carbon number and tends to level off at a certain carbon number, which is approximately C-6 for Athabasca bitumen reservoirs. The main objective of this research is to investigate the potential of dimethyl ether (DME), a water-soluble solvent, as an additive to steam for reducing steam/oil ratio (SOR) while keeping SAGD-like rates of bitumen production. The chamber-edge temperature for a given overall composition and operating pressure is defined as the temperature at which the vapor phase completely condenses with decreasing temperature. Thermodynamic predictions show that the chamber-edge temperature so defined will increase substantially if the solvent can partition into the aqueous phase at chamber-edge conditions. This is confirmed in numerical-reservoir simulation for coinjection of steam with DME, as a water-soluble solvent, for Athabasca bitumen. In simulation case studies, coinjection of steam with DME (DME-SAGD) is compared with SAGD and coinjection of steam with C-4 (C-4-SAGD), in terms of SOR, bitumen production, local displacement efficiency, and solvent recovery. The steam-injection pressure is 35 bar for all cases, and 2 mol% of solvent is coinjected in solvent-SAGD simulations until the steam chamber reaches the side boundary of a 2D homogeneous reservoir model. Because the DME volatility is between C-3 and C-4, C-4 is selected as the alkane counterpart in this simulation study to see the effect of the solvent solubility in water on oil recovery in solvent-SAGD. DME is more volatile and less soluble in bitumen than C-4 at their corresponding chamber-edge conditions. However, results show that DME-SAGD results in 35% lower SOR than SAGD while being able to increase bitumen-production rates of SAGD. Analysis of simulation results indicates that the solubility of DME in water not only makes the chamber-edge temperature higher than that of C-4-SAGD, but also yields 15% higher solvent-recovery factor than C-4-SAGD. The main reason for the latter observation is that a much-smaller fraction of the injected solvent is present in the vapor phase in DME-SAGD than in C-4-SAGD. Also, DME dissolves in both water and bitumen, which results in the aqueous and oleic phases of nearly equal density within the gravity-drainage zone near the edge of a steam chamber. This is the neutral regime of oil/water two-phase flow along the chamber edge between the two extreme cases: SAGD and C-4-SAGD. Unlike in C-4-SAGD, the reduced gravity segregation in DME-SAGD is expected to facilitate the mixing of condensed solvent with bitumen near the edge of a steam chamber.