An Experimental Study of Steam-Solvent Coinjection for Bitumen Recovery Using a Large-Scale Physical Model

In this paper, we present a solvent-assisted steam-assisted gravity drainage (SA-SAGD) experiment with multicomponent solvent (i.e., condensate) using a large physical model. The sandpack for the experiment had a porosity of 0.33 and a permeability of 5.6 darcies in the cylindrical pressure vessel that was 1.22 m in length and 0.425 m in internal diameter. The sandpack was initially saturated with 93% Athabasca bitumen and 7% deionized water. The main objective of this research was to study the in-situ thermal/compositional flow and produced bitumen properties in SA-SAGD with condensate. After the preheating of the sandpack for 24 hours, SA-SAGD with 2.8-mol% condensate was performed at 50 cm3/min (cold-water equivalent) at 3500 kPa for 3 days. The experimental data of production, injection, and temperature distribution were recorded. Also, 10 samples of produced oil were taken and analyzed for density and asphaltene content. The sandpack was excavated after the experiment to analyze the asphaltene content in the remaining oil at different locations. A numerical simulation model was calibrated based on the data of material balance and temperature distribution, and it was validated with properties of the produced and excavated samples. The simulation model used fluid models based on experimental data of viscosities, densities, and bubblepoints for four condensate/ bitumen mixtures. Results showed that SA-SAGD was efficient in bitumen recovery with a cumulative steam-to-oil ratio (SOR) that was two to three times smaller than that in SAGD using the same physical model. Detailed analysis of the calibrated simulation model indicated that SA-SAGD enabled the steam chamber to expand more efficiently with a smaller amount of water throughput than SAGD. Volatile solvent components tended to remain in the chamber, and the condensed solvent components acted as a miscible carrier for bitumen components. The analysis further showed that the more efficient oil recovery in SA-SAGD occurred with predominantly cocurrent flow of oil and water near the chamber edge. SA-SAGD recovered a larger amount of asphaltene components (i.e., less in-situ upgrading) than SAGD likely because of its lower chamber temperature, shorter production period, and enhanced local displacement efficiency.