Smart K_d-values, their uncertainties and sensitivities - Applying a new approach for realistic distribution...

One natural retardation process to be considered in risk assessment for contaminants in the environment is sorption on mineral surfaces. A realistic geochemical modeling is of high relevance in many application areas such as groundwater protection, environmental remediation, or disposal of hazardous waste. Most often concepts with constant distribution coefficients (K_d-values) are applied in geochemical modeling with the advantage to be simple and computationally fast, but not reflecting changes in geochemical conditions. In this paper, we describe an innovative and efficient method, where the smart K_d-concept, a mechanistic approach mainly based on surface complexation modeling, is used (and modified for complex geochemical models) to calculate and apply realistic distribution coefficients. Using the geochemical speciation code PHREEQC, multidimensional smart K_d-matrices are computed as a function of varying (or uncertain) environmental conditions. On the one hand, sensitivity and uncertainty statements for the distribution coefficients can be derived. On the other hand, smart K_d-matrices can be used in reactive transport (or migration) codes (not shown here). This strategy has various benefits: (1) rapid computation of K_d-values for large numbers of environmental parameter combinations; (2) variable geochemistry is taken into account more realistically; (3) efficiency in computing time is ensured, and (4) uncertainty and sensitivity analysis are accessible. Results are presented exemplarily for the sorption of uranium(VI) onto a natural sandy aquifer material and are compared to results based on the conventional K_d-concept. In general, the sorption behavior of U(VI) in dependence of changing geochemical conditions is described quite well.