Inverse estimation of rock thermal conductivity based on numerical microscale modeling from sandstone thin sections

The effective thermal conductivity (ETC) of porous rocks is primarily controlled by the matrix thermal conductivity (MTC) and the thermal conductivity of the pore fluid as a function of porosity. Rock ETC has been investigated widely, however, information on MTC is rather sparse. This study proposes a new approach using a numerical model to determine the MTC of porous rocks by parameter fitting to the measured ETC. The model represents realistic rock textures as detailed information on grain and pore geometries was obtained from microscopic images of thin sections. MTC fitted from water saturated condition (water as pore fluid) showed plausible values between 4.9 and 7.6 W/(m ∙ K), while MTC values from dry condition (air as pore fluid) were disproportionate, except for one model with consistent MTC for both conditions. This supports the idea that at constant temperature and pressure, MTC is an intrinsic property independent of fluid saturation. For this reason, MTC from water saturated condition was used to calculate dry ETC. The model showed lower dry ETC than the lab measurements, but within a reasonable range for dry sandstones. Investigation of ETC regarding saturation degree revealed that the samples may contain 12%-45% residual moisture, which explained the higher measured dry ETC. Additionally, the models allowed observation of textural influences, showing that grain size, distribution, and connectivity affected the heat conduction of porous rocks.