Utilization of Percolation Theory and Multiphysics' Concept to Develop Dynamic Modeling and Accurate Upscaling Approaches for Two-Phase Flow during Immiscible Displacement of Heavy Oil in Porous Media

稠油采收率通常很低，并且大量未采出的稠油留在储层中。稠油开采需要提高采收率（EOR）工艺，这从根本上导致了多孔介质中连续的两相或三相流动（非混相或混相驱油）。此类EOR技术的成功应用需要对多孔介质中的流动机制有更详细的了解。利用动态渗流网络模型研究稠油饱和多孔介质中两相渗流（非混相驱替）的基本机理，并探讨其在稠油油藏中的应用。由于各种提高采收率技术中涉及许多流动和输送机制，因此本研究的重点将放在稠油系统中热水和溶剂注入的动态非混相驱替上，以便实现所有提出的短期和长期目标。该研究项目将有助于填补多孔介质多相流领域的知识空白。尽管人们对动态孔隙网络建模进行了多年的研究，但对其在稠油系统中的应用却鲜有研究。我们将开发稠油系统中热水和溶剂注入综合热导率的多物理体动态渗流网络建模方法。借助多物性动态渗流网络模型，研究孔隙尺度非均质性（尤其是孔隙连通性和孔径分布）、黏度比M、毛管数Ca、重力、热导率以及稠油和溶剂（CO2）性质对（热水和溶剂）驱替的影响。渗流理论中的一个基本概念是渗透率k遵循普遍的标度规律（k~(z-zc)，其中z为多孔介质配位数，zc为渗流阈值，与孔径分布有关）。通用标度律与网络晶格无关，在规则网络和无序网络（岩石孔隙空间为无序网络）中均有效，并可推广到两相流。基于网络模拟，可以推导出稠油体系中孔隙尺度非均质性、M、Ca、热导率与一些宏观性质（相对渗透率、剩余油饱和度、粘性指动）之间的“通用”性质和高级关系（或数学模型）。将通过微模型和沙包实验对模型和数学模型进行验证。所有结果，包括多物理体动态渗流网络模型和升级模型，将有助于研究提高采收率过程中更复杂的相关现象，例如向含盐水的孔隙空间注入溶剂。这些结果使研究人员和工业工程师能够优化石油生产的性能。

Heavy oil recovery is typically low, and a significant amount of unrecovered heavy oil is left in the reservoir. Heavy oil production requires enhanced oil recovery (EOR) processes, which fundamentally cause continuous two- or three- phase flow (immiscible or miscible displacement) in porous media. The successful applications of such EOR techniques require more detailed knowledge of flow mechanisms in porous media.The objective of this study is to investigate the fundamental mechanisms of two phase flow (immiscible displacement) in porous media saturated with heavy oil using dynamic percolation network modeling, and its application in heavy oil reservoirs. Since there are many flow and transport mechanisms involved in various EOR techniques, the focus of this study will be on the dynamic immiscible displacement with hot-water, and solvent injection in heavy oil systems so that all proposed short and long term objectives can be achieved. This research program will help fill the knowledge gap in the area of multiphase flow in porous media.Although dynamic pore-network modeling has been studied for years, its application in heavy oil systems has rarely been investigated. We will develop a multiphysical dynamic percolation network modeling method with integrated thermal conductivity of hot-water and solvent injection in heavy oil systems. With the help of this multiphysical dynamic percolation network modeling, the effects of pore scale heterogeneity, particularly pore connectivity and pore-size distribution, viscosity ratio M, capillary number Ca, gravity, thermal conductance and effects of properties of heavy oil and solvent (CO2) on the (hot-water and solvent) displacement will be studied.A basic concept in percolation theory is that permeability k obeys the universal' scaling law (k~(z-zc), where z is coordination number of porous media, zc is percolation threshold and is related to pore size distribution). Universal' scaling law is independent of network lattice and is valid in both regular and disorder networks (pore space of rocks is disorder network), which can be extended to two-phase flow. 'Universal' properties and upscale relations (or mathematical models) between pore-scale heterogeneity, M, Ca, thermal conductance and some macroscopic properties (relative permeability, residual oil saturation, viscous fingering) in heavy oil systems can be derived based on network simulations. The upscaled relations and mathematical models will be validated with experiments on micromodel and sandpacks.All results, including the multiphysical dynamic percolation network modeling and upscaling models, will enable study of more complex related phenomena in EOR processes, such as injection of solvent into pore space containing saline water. Such results allow researchers and industry engineers to optimize the performance of oil production.