Abstract
Summary
This paper presents a new method for imaging the 3D microstructure of porous media. The method is based on high-resolution X-ray computer tomography (CT)where a cone-shaped, diverging X-ray beam is used to generate 2D transmission images. Unlike traditional X-ray CT, a 3D reconstruction array is created directly instead of creating a series of 2D slices. This allows direct measurement of 3D geometric and topologic properties of porous media on a microscale. Our results also demonstrate the porous media on a microscale. Our results also demonstrate the effectiveness of the high-resolution system for imaging the distribution of two fluid phases in the void space. The imaging system overcomes the limitations of current methods and has enormous potential for quantifying the microstructure of chaotic porous media. porous media.
Introduction
The microstructure of a porous matrix governs such macroscopic properties asacoustic, electrical, fluid dynamic, mechanical, and properties as acoustic, electrical, fluid dynamic, mechanical, and thermal properties. A detailed description of the microstructure of porous media could be useful to every oil production operation. For example, a priori microstructure knowledge would aid in the design of a drilling fluid that would cause minimal formation damage from particle invasion. The structure of a porous matrix also governs its response to externally imposed stresses. thereby making it useful to researchers working in rock mechanics, For multiphase flow through porous media, a knowledge of phase distribution would enhance understanding of the subject. Of particular interest to researchers working in EOR are the geometric particular interest to researchers working in EOR are the geometric and topologic properties of oil-bearing porous media. It is well-known that shape, size, orientation, and connectedness of the flow paths play a vital role in the transport of various fluid phases through porous media. phases through porous media. Currently, two methods are used to characterize porous media. This paper briefly describes the methods and their advantages and disadvantages. In mercury porosimetry, mercury is forced into a porous sample at constant pressure increments and the volume of porous sample at constant pressure increments and the volume of mercury invaded is recorded as a function of injection pressure. The resulting capillary pressure curve and Laplace equation are used to determine the void size distributions. Although useful to predict certain properties, deconvolution of the capillary predict certain properties, deconvolution of the capillary pressure curve gives nonunique results. pressure curve gives nonunique results. Recently, Yuan and Swanson introduced a volume-controlled mercury injection technique. Mercury is injected at a constant rate, and pressure fluctuations are recorded as functions of time. Toledo et al. discussed the advantages and limitations of this method. Although the rate-controlled method is better than the pressure-controlled system, it can yield only limited information pressure-controlled system, it can yield only limited information about the microstructures For example, it cannot be used to determine the topologic properties of porous media. Two-dimensional polished thin sections and an optical microscope or a scanning electron microscope are used to generate digitized maps of the porous matrix. Stereologic principles are used to estimate 3D properties of the porous sample. DeHoff presents a review of the advantages and limitations of presents a review of the advantages and limitations of applying stereology. Properties such as volume fraction, surface area, intercept length, and integral mean curvature of the surfaces can be estimated accurately with stereologic principles. Size distribution measurements would require assumptions about the void shapes. Dullien and Dhawan used a sphere segment model and a general segment model to estimate the void size distribution of sandstone. In the sphere segment model, the pore space is approximated as a 3D network of intersecting spheres. A comparison of the pore size distribution measurements obtained with the sphere segment model and mercury porosimetry show a wide disparity. The disadvantage of using a shape approximation is that it always will overestimate the size distribution of interest. Ideally, one would like to measure the area distribution such that the areas are orthogonal to the local flow direction. It is virtually impossible to determine such distribution with stereologic principles, although a relationship must exist between the distribution of areas in a 2D thin section and the distribution of areas relevant to fluidflow. Despite the use of stereology, certain properties cannot be measured with stereologic principles. Pathak et al. showed that among such properties, the topology of the matrix was critical. They measured residual oil saturation(ROS) in sintered copper powder packs and showed that the ROS was strongly dependent on the powder packs and showed that the ROS was strongly dependent on the specific genus of porous media. The genus that is a measure of connectivity was varied by controlling the degree of sintering of the copper powder. On the basis of the experimental observations, Pathak et al. stated that "the larger the connectivity, the larger Pathak et al. stated that "the larger the connectivity, the larger the number of interconnections between the pores, the greater the number of alternatives available for oil drainage, the lower the percolation threshold, and the lower the residual saturation." percolation threshold, and the lower the residual saturation." They concluded that fluid distributions and Porous media flow properties depend not only on pore size and shape but also on local connectivity. Sharma and Yortsos studied the effect of connectivity on the permeability decline experienced during particulate injection through porous media-a problem commonly encountered in petroleum recovery operations. They approximated the porous matrix as a network of resistors. They varied connectivity of the network by varying the coordination number. Their results show that the permeability decline rate is strongly dependent on the coordination permeability decline rate is strongly dependent on the coordination number. The above discussion indicates that connectivity is as important as geometry in influencing fluidflow through porous media. However, we have a dearth of information on porous media topology. Currently available techniques are indirect and can provide only limited information. The ultimate goal of microscopic provide only limited information. The ultimate goal of microscopic structure analysis is to correlate such macroscopic properties as permeability, relative permeability, and dispersivity with such permeability, relative permeability, and dispersivity with such microscopic properties as void size distributions. anisotropy, tortuosity, aspect ratio, and connectivity. A high-resolution X-rayimaging system is introduced here as a viable tool to allow direct examination of the 3D microstructure of porous media.
Apparatus Description
Fig. 1 provides a block diagram of the microfocal tomographic system. The imaging system was designed originally for examining small structural defects in ceramic materials. A microfocal X-ray source approximates a point source that emits a diverging X-ray beam, An X-ray acceleration potential of 60 kV was used for all the experiments. The focal spot size is 1 to 5 m. The object to be imaged is placed as close to the source as possible to achieve the desired geometric magnification. The magnified shadow of the object is captured by a 2DX-ray detector. The detection system consists of an X-ray image intensifier, avidicon camera, and a video digitizer. The object can be translated and rotated by Aerotech hardware connected to a general-purpose computer (DEC VAX 11/730)by a serial link. The VAX computer is in command of the entire data acquisition and reconstruction procedure. Computationally intensive portions of there construction procedures are performed by a Floating Point Systems AP-120Barray processor attached to the VAX. processor attached to the VAX.
Publisher
Society of Petroleum Engineers (SPE)
Subject
Process Chemistry and Technology