A critical parameter of transport modeling is the selection of the appropriate thermodynamic data base or the determination of acid / base constants and complex formation constants (e.g. FITEQL, PHREEQC), especially in unknown systems. Frequently a definite model solution is only possible in combination with spectroscopic methods. The description of the kinetics of solid phase / metal interaction is still incomplete, but model approaches (e.g., KICAM, PHREEQC) exist for simplified description for selected natural groundwater systems. In addition to the kinetics of the sorption / desorption reaction, the 3D geometry of the pore space / fracture structure is of crucial importance for the mobility of environmentally relevant pollutants. Especially the integration of the fracture geometry and the change by formation of secondary phases in transport models is often not possible due to the absence of detailed information. New 3D examination methods, e.g. Computed tomography (CT) can provide quantitative information on the fracture network / pore space structure. The incorporation of this information into computational fluid dynamic (CFD) transport models and the calculation of the reactive transport along individual streamline paths from particle tracking realizations allows the differentiation of physical and chemical retention mechanisms and the temporal change of the hydraulic system in a further step. Another focus is lead on the predictive modeling of "clogging" phenomena by mineral precipitation reactions in porous media. Here, the empirical Archie's law, implemented in reactive transport codes, was tested for validity by comparison to experimental data (under defined boundary conditions) predicting transport calculations using PHREEQC or TOUGHREACT. Such scientific work on the direct interaction between experimenters and modelers is extremely important to reduce uncertainties concerning the implemented reaction formalisms in existing reactive transport codes, thereby strengthening the trustworthiness of numerical codes as prognostic tools.
Project leader: Thorsten Schäfer