This project simulates long-term, 2D reactive transport driving cementitious grout degradation within the bedrock foundations of hydropower dams. Preventing under-dam seepage is critical for structural safety and operational longevity; however, constant hydraulic gradients induce severe calcium leaching that degrades the grout over decades. We use a split-operator approach combining the finite element library FEniCS dolfinx as transport solver and PhreeqcRM library for chemical equilibrium and kinetics calculation. Our model also incorporates micro-scale boundary roughness at the rock-grout interface - requiring a highly fine local mesh refinement and a greater computational power. The workflow is executed via a Python script, calling high-performance C++ backends natively optimized for distributed-memory architectures via the Message Passing Interface (MPI). FEniCS dolfinx is invoked via mpirun, and PhreeqcRM has built-in MPI support. We estimate a typical job utilizes standard multi-core configurations, allowing us to find the ideal balance between kinetic computation and MPI communication overhead. The aggregated memory requirements will comfortably fit within the standard default memory allocation. The main supervisor is Liangchao Zou, associate professor at The Department of Sustainable Development, Environmental Science and Engineering, KTH.