Three Dimensional Visualisation of Solvent Chamber Growth in Solvent Injection Processes: An Experimental Approach

Abstract

Solvent based methods have been recommended as an alternative to thermal techniques for heavy oil and bitumen recovery. However, the cost of solvent injection processes is high and therefore requires a careful design of field scale applications. Yet, the efforts have been limited to numerical or analytical models in clarifying the complex physics of solvent processes due to difficulties in experimentation. Vapor extraction (VAPEX) is one of the solvent applications suggested and tested for heavy-oil/bitumen recovery from unconsolidated oilsands. In this method, solvents are injected into the reservoir through a horizontal well and gravity drainage takes place resulting in the production from the lower production well. Theoretical analyses suggest that a solvent chamber develops and the displacement is controlled by its growth during this process. Previous studies, however, showed the possibility of unusual behavior at the reservoir scale reservoirs such as stopping, shrinking, and even disappearing of chambers associated with severe fingering phenomena. These kinds of anomalies can only be identified through experimental designs and, so far, have been limited to 2-D models with minimal visual support. In this paper, a three dimensional imaging technique using laser is introduced. Refractive index matching of solvent, oil, and glass beads pack enabled imaging the progress of solvent displacement. 2-D images of slices with little spacing in the scaled VAPEX model were recorded within a significantly short time using laser sheets. Then, 3-D images were generated by integrating the 2-D images. In simulating the VAPEX model, different injection rates were applied and different sizes of glass beads were used to create different permeability media to also study the physics of the process parametrically. 3-D images were analyzed to understand the chamber growth process during VAPEX. The observations are expected to be useful in clarifying many uncertainties of the VAPEX process and also provide data for further computational studies.