The fine grained nature of shale makes Scanning Electron Microscopy (SEM) an essential part of the core analysis suite of techniques. Typically, the SEM is used for both imaging and for elemental analysis by Energy Dispersive X-Ray spectroscopy (EDS). Both the imaging and the EDS analysis require a specific approach when working on shale samples.
In shale reservoirs, the pores in the organic matter can be as small as 2nm, what would be the lower limit that can be visualized in the SEM. Imaging these small features comes at the cost of a very limited Field of View—the higher the magnification, the smaller the area that will be imaged. This raises serious questions about the representative nature of the SEM images. In the ideal case, one makes a mosaic by stitching as many SEM images as needed to have both the fine details and the fabric. However, for a 1 inch core plug, it would take 900 million (2kx2k) pixel images to cover the entire surface with a resolution that sees the smallest pores. While technically possible, it would not be practical.
More practical would be a methodology whereby mosaics with increasing resolution and decreasing Field of View combined with EDS maps are analyzed to uncover the representative features or building blocks that make up the rock fabric. Integrating rock fabric analysis with the constituent mineralogy from EDS mapping provides a pathway towards estimating the mechanical and chemical properties of shale reservoirs. Furthermore, the ability to identify representative textures by utilizing a multi-scale imaging approach aids in identifying prime candidates for further analysis in 3D with the Focused Ion Beam SEM (FIBSEM); a feat that was not possible with SEM imaging until recently. However, even an SEM mosaic of a few hundred images would be gigapixels in size, and that may pose problems for the efficiency of most image analysis software packages.
EDS-based mineralogical analysis has some serious limitations as well. The interaction volume of the electron beam with a sample is a pear shaped figure with a diameter of between 1 and 2 micrometer and a depth approaching 4 µm at high accelerating voltage. For a shale sample where grain sizes of clays, micas and organic particles are submicron, this means that the EDS spectrum will often contain multiple phases. The EDS analysis is therefore complicated, not only by the clay minerals themselves, but also because the spectra may contain different clay species.
This paper presents the first results in which the EDS spectrum acquired at each pixel was analyzed and deconvolved to reveal the relative modal proportions in each measured pixel. This approach provides insight into the mineralogy and mineral textures not attainable by any other technique.