|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Core photographs become more valuable when scanned and digitized into a data base. The ability to zoom, enhance the color contrasts and compare directly to other log and well data allows details to be identified and interpreted that are barely visible on the original photographs. Examples of core images are shown that compare various photograph scales and various scanning densities to illustrate effects on image quality. Core images are compared to micro-resistivity and acoustic images demonstrating the interpretations available by combining various data sets. Thin section and SEM images, which have been stored in the same data base, are also displayed.
The various steps in making digital images of core photographs are discussed including core preparation, photography and transfer of data sets. High quality digital images produce large blocks of data. Data set sizes are compared for variations in scanning density, photographic scale and color resolution.
Some applications are discussed that use quantified image data including thin bed corrections, definition of pore geometry and improved relative depth control.
Image data represents the only remaining well data to be routinely stored in digital data bases. There are two main reasons for this delay. The quality of good photographs (the most common media for storing images) is difficult to capture and when using high density scanners in an attempt to capture this quality, the sizes of the resulting data sets become very large. However the benefits of digital images accrue very quickly, even with existing technology. These benefits included those attributable to all digital data bases, such quick and easy distribution and access. In addition images derived from photographs can be manipulated to yield information not available from the original images.
This is realized by color enhancement, zooming and by combining images to make a synthetic image. It also opens the option of quantifying image characteristics to reflect reservoir properties.
A quantitative image analysis system has been configured for characterizing pore space of reservoir rocks. The technique can be applied to a magnified image of a rock thin section or polished section. With this system, numerous accurate measurements of porosity, pore size, shape, and internal rock-surface area can be obtained and easily summarized as distributions of pores. This quantity and quality of information is not attainable with conventional thin-section point-count techniques. Porosity values obtained by image analysis are in agreement with those obtained from core-plug analysis.
Reservoir properties of rocks depend strongly on their pore space geometries, yet very little is known about the pore space geometries, yet very little is known about the geometry and connectivity of pores because of their complexity and the consequent difficulty in obtaining measurements of sufficient quantity and accuracy to be statistically meaningful. Qualitative, descriptive terms currently applied to pore space in reservoir rocks are usually vague and inadequate. Manual point counting, the commonly used method for describing pores in thin sections, is tedious, time consuming, and too imprecise to solve many problems that involve reservoir studies. A number of laboratory macroanalysis methods are available for determining effective PV fraction and pore-size distribution, but these methods do not give information about the shape and spatial distribution of pores. For example, capillary-pressure measurements give limited information on pore size and connectivity. Recent advances in computerized quantitative image analysis have made possible the rapid and highly accurate measurement of rock microstructure. An image-analysis system that uses magnified images of rock thin sections or polished sections has been configured for the measurement of porosity percentage, specific surface, pore size, and pore-shape distributions. These measurements are made with a commercial image analyzer. The optical image is scanned by a Vidicon TM camera, converted to digital form, stored, and processed on a microcomputer. Each pixel, or picture element, is allocated to 1 of 256 intervals on the gray scale. Individual pixels of the digital image are classified as rock or pore on the basis of gray level. Pore sizes are obtained by measurement of Feret's diameter in several directions for individual pores. Feret's diameter is the maximum spacing between parallel tangents to a feature in a given direction. The maximum Feret diameter is considered as length, the minimum Feret diameter as breadth. Pore shape is computed by any of a number of indices--e.g., aspect ratio, or as a normalized form factor where a sphere = 1. Rocks with a wide range of pore sizes can be analyzed at more than one level of magnification. Used in conjunction with routine scanning electron microscopy (SEM) or light microscopy, the analysis provides a direct means of measuring pore size, shape, and provides a direct means of measuring pore size, shape, and distribution for a wide range of rock types. The method can serve as a descriptive aid in the evaluation of rock with complex pore systems or other rocks that are inadequately described by other methods. Measurement of pore structure can also be valuable in the identification of geologic controls on porosity and permeability distributions.
The operating principle of the image-analysis equipment is a sequence of steps shown schematically in Fig. 1. The input image, which may be either a light-microscope image or SEM photo, is projected onto the face of a scanner tube, such as a Vidicon camera, where it is converted into electrical pulses on the basis of gray level. Scanners designed for image analysis give optimum balance among speed of operation, resolution, and noise. The Vidicon scanner scans the 704-line raster of a full frame 10.5 times per second, giving up to 630,784 pixels per standard per second, giving up to 630,784 pixels per standard image. Both image frame size and pixel resolution are variable. In practice, SEM photographs are placed on an illuminated copy stand above which the Vidicon camera is rigidly attached. Image focus and magnification are controlled by a conventional macrozoom lens. Illumination is controlled automatically by instrument software. The instrument has the capability for automatic software-controlled focussing and x-y stage movements. Direct interface to an SEM and/or light microscope is also available. Variations of illumination across the view field and scanner variations are compensated by a shading corrector routine. The amplitude of the video signal is a function of light intensity derived from the various phases or components within the input optical image (Fig. 1). An analog of the electronic image is transmitted to a detector that allocates each pixel to 1 of 256 levels on the gray scale. A value of zero corresponds to black; a value of 255 corresponds to white. To detect the phase or features of interest, a gray-level threshold and band width is set by the operator.
This paper was prepared for presentation at the 1999 SPE/EPA Exploration and Production Environmental Conference held in Austin, Texas, 28 February-3 March 1999.