Numerical Modeling of Effective Thermal Conductivity for Polymineralic Rocks using Lattice Element Method

Abstract

Accurate prediction of rock thermal conductivity under in-situ conditions is essential for characterizing subsurface heat flow. This study presents a numerical framework based on the Lattice Element Method (LEM) for simulating the effective thermal conductivity of polymineralic rocks under coupled pressure-temperature conditions. The model resolves interactions among heat transfer, grain contacts, and mechanical deformation within a microstructure-representative lattice. The methodology enables consistent treatment of heat conduction, nonlinear contact evolution, and thermally induced intergranular fracturing. Heterogeneity is introduced through a stochastic, volume-fraction-constrained discretization that preserves the measured mineral composition and porosity, while mineral anisotropy and fracture behavior are captured through element-level constitutive laws. The framework is evaluated using experimental data for two dry sandstones under ambient and elevated pressures and temperatures. Effective thermal conductivity is computed over the same pressure-temperature ranges and compared directly with the measurements. The results indicate that the predictions are capable of reproducing the characteristic trends and absolute levels. The close agreement between experimental observations and model predictions confirms that the thermo-mechanical coupled LEM provides a physically grounded and transferable approach for modeling heat transport in heterogeneous, polymineralic media.