摘要

During the solvent vapor extraction (VAPEX) process, a heavy oil reservoir can be divided into three different zones in terms of its fluid saturations, namely, the solvent chamber, transition zone, and untouched heavy oil zone. In the past, the solvent chamber was assumed to be a linear or circular shape in the previous studies. However, it has been observed to be close to a parabolic shape in many laboratory VAPEX tests. In this paper, a new parabolic solvent chamber model in the concave or convex case is formulated to predict the solvent chamber evolution and the heavy oil production in the VAPEX heavy oil recovery process. In the experiment, each recorded digital solvent chamber image at a different time is digitized to determine the solvent chamber shape by analyzing the sudden change of the gray level of each pixel. In theory, the overall discrepancy between the predicted and digitized solvent chambers is minimized by adjusting the transition-zone thickness. It is found that in comparison with the linear and circular solvent chamber models, the parabolic solvent chamber model gives the best prediction of the solvent chamber evolution, especially in the spreading phase. In addition, the maximum transition-zone thickness variation of 13.1% during the entire VAPEX test indicates that the transition-zone thickness can be assumed to be constant. Similar to the other solvent chamber models, the parabolic solvent chamber model can adequately predict the cumulative heavy oil production. The relatively large error of the predicted cumulative heavy oil production is caused by a commonly used assumption. The initial oil saturation in the transition zone is assumed to reduce to the residual oil saturation once the transition zone becomes an incremental part of the solvent chamber. This major theoretical assumption needs to be further investigated.