When realizing thermal enhanced oil recovery projects, it is necessary to have high quality data on thermal and geomechanical properties of reservoir rocks and/or connection between those effective properties at reservoir conditions. Recently a number of important papers were published regarding a connection of the effective properties of rock with unique parameters of rock microstructure using the effective medium theory (EMT). We demonstrate how to connect the dynamic elastic moduli of rock in different saturation states with its thermal conductivity (TC) via these parameters. An EMT-based model of physical properties of Bentheimer sandstone is successfully developed and tested. The model is based on results of microstructure analysis of the sandstone samples and experimental data on elastic wave velocities (Vp and Vs) and TC of the samples in two saturation states (dry and brine-saturated) and under axial stress up to 21 MPa. We use two approaches of EMT: f-model and well known self-consistent approach, to show special aspects of this modeling. We demonstrate how the axial stress affects the rock's microstructure and how it is reflected in the thermal and elastic properties.


Data of rock thermal properties including thermal conductivity (TC) are commonly used in different areas of prospecting geophysics, basin modeling, storage of radioactive materials, etc. While techniques for measuring of TC on core samples in the lab at room temperature and normal atmospheric pressure are well known [1], logging tools for TC are not widely used in the industry and their accuracy is still under investigation [2]. Thus, we utilize theoretical models to connect TC with other well-known properties.

Modeling of effective rock physical properties is a challenging problem. Transport properties (TC and electrical resistivity) as well as elastic properties depend on many factors: mineral and organic composition, cementation, fluid saturation, volume ratios of minerals and voids, size of components and their spatial distribution, and surface properties of grains [3-5].

Five groups of theoretical approaches exist to estimate rock's effective properties: empirical [6, 7], engineering [8-10], effective medium theory (EMT) [11, 12], artificial neural network [13], and numerical simulation approaches [14]. Those approaches have advantages and drawbacks. Empirical approaches use many experimental data to derive dependencies between physical parameters, and they have narrow applicability. Engineering ones use exact solutions for medium model with specific assumptions of its structure, such as two-phase layered medium. Artificial neural network approaches require huge massive of data. Numerical simulation approaches require high-resolution information about the rock structure obtained from X-ray computed tomography. However, the best resolution can be reached on very small samples approximately 8 mm in height. However, for rocks with complex structure representative elementary volume (REV) can be larger.

This content is only available via PDF.
You can access this article if you purchase or spend a download.