A new finite element code was developed for modeling thermo–hydro–mechanical coupled processes in continuum-based geomechanical applications. The development was conducted using open-source software and libraries for input of model geometry and mesh, numerical calculation, and data management to take advantage of the accessibility and flexibility of the source code for updates and revisions. Implemented primarily with the PETSc library for parallel mesh discretization and numerical solvers in C++, the code incorporates physics models including heat transfer, Darcy's flow and storage, and linear elasticity and their coupled processes. The physics models were verified by comparing them with analytical solutions and a commercial FEM code. The code was applied to the simulation of deep borehole disposal for high-level radioactive waste, modeling the generated thermal stress due to decay heat emitted from spent nuclear fuels. The simulated results of changes in coulomb failure stress and ground uplifts in the deep borehole disposal model were compared with those of the reference disposal concept. This study demonstrates the applicability and future extensibility of the new finite element code to field-scale geomechanical applications.
Thermo-hydro-mechanical (THM) coupled processes are a key behavior in geomechanical applications such as nuclear waste disposal in geological media, underground CO2 sequestration, and enhanced geothermal systems. Understanding the complex interaction between heat transfer, fluid flow, and mechanical responses in porous media is essential for accurately analyzing and predicting the long-term performance and safety in subsurface environments. Numerical modeling can improve the understanding of complicated coupled processes by capitalizing on the benefits of advanced computer technology.
The active development of open-source software for numerical computation as well as pre- and post-processing has facilitated the emergence of large-scale projects aimed at developing numerical simulators for coupled processes in subsurface environments. Notable examples include OpenGeoSys (Bilke et al., 2022), MOOSE (Lindsay et al., 2022), PFLOTRAN (Lichtner et al., 2020), and PyLith (Aagaard et al., 2023). The development of open-source frameworks offers several advantages, including easy access to the source code and, thus, flexibility in writing and revising the code suitable for specific needs. Furthermore, another significant advantage is the ability to obtain independent simulators from commercial software.
