Numerical Simulation Study of SAGD Experiment and Investigating Possibility of Solvent Co-Injection

Ashrafi, Mohammad (Norwegian University of Science and Technology) | Souraki, Yaser (Norwegian University of Science and Technology) | Karimaie, Hassan (Norwegian University of Science and Technology) | Torsaeter, Ole (Norwegian University of Science and Technology) | Kleppe, Jon (Norwegian University of Science and Technology)

OnePetro 


Bitumen resources constitute a high portion of the total world oil resources. The main recovery mechanism for these high viscous fluids is to reduce their viscosity by the application of heat, mostly by introducing steam.

Among different steam injection schemes, steam assisted gravity drainage (SAGD) has become the method of choice applicable to bitumen and oil sand reservoirs. In these extra heavy oil resources, the reservoir has almost no injectivity due to high oil viscosity, and therefore conventional steam flooding is hard to conduct. SAGD, however, reduces the viscosity of bitumen in place and the heated bitumen drains due to gravity forces towards the production well and is then being produced. Recently hybrid processes are attracting more attentions in the industry. These processes benefit from co-injection of a solvent together with steam. The solvent can diffuse into the bitumen and make it even lighter by reducing the viscosity.

Our simulation study is based on the experimental work done by Chung (1988) and the simulation model of this experiment by Chow (1993). Chung's physical experiment was a 2-D model to simulate SAGD experiment in laboratory. The Chung's experiment was done with Cold Lake crude oil. A reservoir simulation model was built using a numerical thermal reservoir simulator. The model was then tested and validated with Chung's physical model. Having a valid model, sensitivity analysis was run to examine the effect of different simulation parameters on recovery and steam oil ratio.

The sensitivity parameters tested are steam temperature and quality, the porosity of the model, different well placement schemes, and the effect of shale barrier. Different steam temperatures and qualities were examined. The best injection condition was found to be 130 °C and 90% quality, beyond which no increased recovery was achieved. Different injector and producer placements were tested. Placing injector and producer diagonally in the model showed the best horizontal sweep efficiency in the laboratory model. Horizontal shale barrier had a dramatic negative effect on the oil recovery. Vertical shale, however, had a smaller effect. This is because in horizontal case the steam chamber cannot reach to the top layers. Porosity was found to be inversely proportional to the oil recovery and steam oil ratio. Results showed that solvent can help to improve oil recovery and steam-oil ratio. In addition most of the injected solvent could be recovered from production stream. Sensitivity analyses on solvent type and concentration indicated significant effects on performance of process. Among the solvents used in this study, hexane showed the best recovery performance.