The modeling of chemical reactions in groundwater transport models can represent complex processes in the subsurface. A common approach is coupling a groundwater flow and transport model with a hydrogeochemical model. This presentation demonstrates the coupling of MODFLOW 6 and PHREEQC 3.
MODFLOW 6 is the latest version of a groundwater modeling framework provided by the U.S. Geological Survey (USGS). It combines most features from previous versions of the MODFLOW series such as MODFLOW-2005, MODFLOW-NWT, and MODFLOW-USG. Furthermore, it adds new capabilities such as groundwater energy transport. It strives to be a unified, flexible, and modular system for simulating multiple hydrologic processes including groundwater flow (GWF), groundwater transport (GWT), groundwater energy transport (GWE), and particle tracking (PRT).
MODFLOW 6 offers access to all model values during runtime via the Basic Model Interface (BMI). This technique is well-suited for coupling with other models. The open-source Python library pymf6 uses this interface to allow interactive access to MODFLOW 6. pymf6 provides a high-level approach to facilitate understanding of model internals. Interactively stepping through a model and introspecting MODFLOW 6 variables helps to understand the internals of a model run. This is essential for model coupling.
PHREEQC is a widely used hydrogeochemical modeling tool that is frequently used for reactive transport modeling. PhreeqPy is a Python interface to IPhreeqc, a PHREEQC variant that runs PHREEQC calculations programmatically without PHREEQC input files. This can considerably facilitate the implementation of reactive transport models. The latest version of PhreeqPy incorporates PhreeqcRM, a PHREEQC version specifically designed for coupling with transport models. PhreeqcRM has a comprehensive programming interface that provides programmatic control over all geochemical modeling features. PHREEQC input files are used to define initial conditions. All computations are performed for multiple cells that can be mapped to the cells of a transport model grid.
This presentation shows how PhreeqPy enables the user to couple PHREEQC computations to transport models. All PHREEQC processes can be controlled in an interactive Python session. This greatly facilitates the development of model coupling. All PHREEQC values can be introspected at all time steps. Jupyter Notebooks simplify the visualization of model values at any stage of the model run. Examples show how PhreeqPy can be used for different use cases. A simple piston flow example of advective 1-D transport demonstrates the general workflow.
A more comprehensive example demonstrates how rtmf6, which combines PhreeqPy and pymf6, can be used to model multi-dimensional reactive transport. The groundwater transport computations run in parallel, using one MODFLOW 6 transport model per chemical constituent. The chemical reaction computations are parallelized with threading in PhreeqcRM. This considerably speeds up computation by taking advantage of modern multi-core systems. The first experiences with the application of rtmf6 show that it provides a performant and flexible system. Its open architecture facilitates accounting for special conditions through custom Python programming.