Imaging of Pore Scale Distribution of fluids and Wettability

Abstract

Wettability has a profound effect on reservoir recovery and productivity. It determines the microscopic distribution of fluids in the pore-space which, in turn, determine important global multiphase properties such as capillary pressure, relative permeability, residual saturation and resistivity index. Complexities in porespace geometry, rock-fluid and fluid-fluid interactions have limited descriptions of wettability to highly simplified model systems and wettability in real porous systems remains a poorly understood phenomenon. This paper utilizes two new techniques which have the potential to greatly improve our understanding of wettability in real porous systems. The first is a technique to reproducibly clean and modify the surface energy of clastic and carbonate cores to produce well defined wettability states. The second is a technique for directly imaging the pore-scale distribution of fluids in reservoir cores using high resolution tomography and a newly developed 3D registration technique which allows voxel perfect alignment of a set of images of the same core. We present results for a preliminary study of drainage and imbibition in Fontainebleau sandstone, sucrosic dolomite and oomoldic grainstone cores at well defined wettability states using air and water to demonstrate the applicability of the techniques. The imaged fluid distributions show that gas is preferentially located in larger pores with water occupying smaller pores. The gas saturations, measured compare well with those calculated from the imaged fluid distributions. The imaged pore-scale fluid distributions are also compared with predictions based on computations made directly on dry images of the pore-space and in equivalent network models. The computations use simple percolation concepts to model the pore-scale distributions. Drainage saturations and fluid distributions compare well with invasion percolation. Imbibition fluid distributions compare well with ordinary percolation. The comparisons show, for the first time, the feasibility of testing the validity of network models for multi-phase flow by directly comparing model fluid saturations with imaged saturations in real systems on a pore-to-pore basis.