Microfluidic investigation of cyclic solvent injection: from reservoir-on-the-chip to large scale

Abstract

Cyclic Solvent Injection (CSI) is an effective Enhanced Oil Recovery (EOR) technique with several economic and environmental benefits. Understanding the key mechanisms and developing reliable scaling criteria are required to successfully implement CSI commercially. In this study, CSI was assessed from both microscopic and macroscopic perspectives using experimental and numerical methods. Regarding the experimental studies, the characteristics of the heavy oil/solvent systems used were evaluated through a detailed PVT analysis that included Constant Composition Expansion (CCE) and Differential Liberation (DL) tests. Moreover, this study examined the microscopic behaviour of foamy oil flow on microfluidic platforms, examining the mechanisms involved in bubble evolution. Based on the visualization studies conducted on the microfluidic systems, it was found that solvent type, pressure depletion rate, and reservoir characteristics had a significant influence on the extension of foamy oil flow. Accordingly, solvents containing a higher proportion of CO2 exhibited superior performance, primarily due to their ability to lower viscosity, enhance swelling, and deliver more gas molecules. Additionally, a higher pressure-depletion rate increases the driving force for bubble nucleation while limiting the time available for bubble coalescence. Moreover, lower reservoir porosity interferes with bubble movement and slows down coalescence, prolonging the foamy oil flow. In addition, Sandpack experiments showed that the Cumulative Gas Oil Ratio (CGOR) is a key indicator of foamy oil flow. Even below bubble point pressure, CGOR remains nearly constant as exsolved gas disperses rather than forming free gas immediately. As part of the simulation study, a numerical model was developed using the CMG software package that captures the non-equilibrium behaviour of the foamy oil flow by utilizing two pseudo-chemical reactions including bubble generation and bubble coalescence. Additionally, to minimize the discrepancy between simulation predictions and experimental results, CMG CMOST was used to tune the oil and gas relative permeabilities as well as reaction rate frequency factors. CSI has been also formulated comprehensively by integrating material balance, mass transfer, and pseudo-chemical reaction equations to derive key dimensionless scaling terms based on the Buckingham π Theorem. 11 dimensionless terms have been identified, which encompass a wide range of phenomena, including foamy oil mobility and its intricate dynamics, as well as solvent exsolution processes. As a result, a comprehensive procedure has been developed for scaling up laboratory results to larger systems. In addition, to account for pressure propagation delay in larger reservoirs, an effective workflow was established to systematically modify the permeability in lab settings in such a manner that its results can be translatable into larger models. Based on analysis of two synthetic reservoirs, a reasonable match in terms of recovery factor, cumulative gas production per unit pore volume, and CGOR versus dimensionless time between the synthetic reservoirs and the Sandpack models was demonstrated, which highlights the robustness and effectiveness of the proposed scaling method. The proposed scaling workflow offers a foundation for future research on scaling methodologies in solvent-based heavy oil recovery processes. Moreover, it can be used to optimize recovery strategies and reservoir management by enabling more accurate predictions of CSI performance at larger scales.