Effects of Waterflooding, Solvent Injection, and Solvent Convective Dispersion on Vapour Extraction (VAPEX) Heavy Oil Recovery
MetadataShow full item record
In this study, the detailed effects of waterflooding and solvent injection on vapour extraction (VAPEX) heavy oil recovery are studied by using a visual rectangular sandpacked high-pressure VAPEX physical model with a low permeability. The physical model is sandpacked and then saturated with a heavy oil sample at the connate water saturation. Pure propane and a mixture of n-butane and methane are used as respective solvents to extract two different heavy oils. The waterflooding effect is examined by performing a series of VAPEX tests with the initial waterflooding prior to the subsequent solvent injection. In addition to the visual observation of the solvent chamber evolution, the heavy oil production rate, solvent–oil ratio, and asphaltene content of the produced heavy oil are measured during the waterflooding and solvent injection. It is found that the initial waterflooding causes an oil production reduction during the subsequent solvent injection. Also, solvent breakthrough occurs earlier and a small amount of water is produced afterwards. This is because the waterflooding creates some low-resistance channels for the subsequently injected solvent to bypass the untouched heavy oil. As a result, the heavy oil is not diluted enough to be produced during the solvent injection. In the absence of waterflooding, however, solvent injection can increase the heavy oil production, in comparison with the so-called solvent-soaking process. Moreover, it is visually observed that solvent injection leads to less asphaltene deposition onto the porous media. This is also quantitatively verified by the measured higher asphaltene content of the produced heavy oil after the solvent injection. Theoretically, a new analytical model is developed to determine the solvent convective dispersion coefficient in a solvent vapour extraction (VAPEX) heavy oil recovery process. It is assumed that solvent mass transfer by convective dispersion takes place along the transition zone between the solvent chamber and untouched heavy oil, whereas solvent mass transfer by molecular diffusion occurs in the direction normal to the transition zone. It is also assumed that the solvent-diluted heavy oil gravity drainage through the transition zone has a linear or quadratic velocity profile in order to obtain analytical solutions for the solvent convective dispersion coefficients in the solvent chamber spreading and falling phases. As a result, this analytical model correlates the solvent convective dispersion coefficient to the maximum apparent oil gravity drainage velocity at the interface between the solvent chamber and transition zone, solvent molecular diffusion coefficient, transition-zone thickness, and porosity of the porous medium. To determine the solvent convective dispersion coefficient, the maximum apparent oil gravity drainage velocity is calculated by using Darcy’s law and the transition-zone thickness is obtained either from a previous study or by using a time similarity between the solvent molecular diffusion and oil gravity drainage. It is found that such determined solvent convective dispersion coefficients are two to five orders larger than the solvent molecular diffusion coefficient, depending on the detailed experimental conditions of a specific VAPEX test.