Multiscale Pore Structure Characterization By Combining Image Analysis And Mercury Porosimetry

Abstract

Abstract This article introduces a multiscale pore structure characterization method using a combination of mercury porosimetry and image analysis. The method was used to determine the distribution of pore volume by pore size and to estimate the pore-to-throat size aspect ratio. The key idea of the method is that pore size distribution obeys a fractal scaling law over a range of pore size. On this basis, scattering intensity data computed from the measured two-point correlation function and those measured from mercury porosimetry are extrapolated in the size range 0.01 µm < r < 1000 µm, using the known fractal scaling law. A set of siltstone samples taken from Daqing Oilfield was analyzed through this method. Distribution of pore volume by pore size over the entire range of pore length scales was determined. The results demonstrated significant similarities in the pore structure of all samples. The image analysis results were in qualitative agreement with the results of mercury intrusion/extrusion tests. The results were also compared with some other samples (including siltstone, sandstone, and dolomite) that had been analyzed using similar methods. It is shown that the surface fractal dimension obtained by analysis of MIP data is consistent with the value obtained by image analysis for different samples with different porosity and permeability. Novel information on the pore-to-throat aspect ratio is obtained by comparing the complete pore volume distribution (PVD) to the MIP data. Introduction Many methods of pore size distributions measurement are available. They are mainly mercury porosimetry methods, photomicrographic analysis methods, and sorption-desorption isotherms methods. However, no single experimental technique can yet provide a quantitative description of rock micro-architecture over length scales spanning four to five orders of magnitude. Direct imaging methods (Ruzyla, 1984) cannot provide statistically significant microstructure data at length scales smaller than 1 micrometer. Indirect imaging methods, such as small-angle neutron scattering (SANS/USANS) or small-angle X-ray scattering (SAXS), reveal pores ranging from one nanometer to about 10 micrometers (Radlinski et al., 2002). Nuclear magnetic resonance (NMR) relaxation methods (Shen, 1992) can only semiquantitatively provide characteristics of pore sizes and results of such analysis can be distorted by diffusional averaging of magnetization between pores (Chang, 2002). The mercury intrusion porosimetry (MIP) method may be used to probe the distribution of pore volume in the range 20 nm to 100 µm; unfortunately, the method provides only the distribution of pore volume accessible to mercury through pore throats of different size. A combination of mercury porosimetry and photo-micrographic analysis can lead to a better understanding of pore accessibility and pore structure characterization in general (Dullien, 1979). Radlinski et al. (2002) presented a method of determining the pore size distribution based on the statistical fusion of small-angle neutron scattering (SANS) and backscatter SEM (BSEM) data by utilizing surface fractal property of rocks (Radlinski et al. 1999). The results have provided the pore size distribution in the range 1nm to 1mm. Amirtharaj et al. (2003) presented a MIP and BSEM combined method using information obtained from MIP instead of SANS. This paper introduces a similar method that combines mercury intrusion porosimetry (MIP) and backscatter SEM (BSEM) data by using surface fractal law. The difference is that the MIP procedure used in this experiment included high-pressure (up to 50000 psi) parts and low-pressure (1–50 psi) parts, instead of one continuous MIP procedure. The experimental results were analyzed and compared with other previous experimental results; furthermore, this paper discusses the micro-heterogeneity of the samples from a fractal dimension point of view.