Freiburger Schriften zur Hydrologie

[ << Band/volume 27 ] [ Freiburger Schriften zur Hydrologie ] [ >> Band/volume 1 ]

Band/volume 28: Adolph G. (2009):

Kombination von Isotopenmethoden und Grundwassermodellen in der

This thesis was written as scientific part of the “ERGO” project (“Effective assessment of risks for drinking water supplies from contaminated sites in the Upper Rhine”).
The aim of ERGO is to assess whether and to which degree known or suspected contaminated sites can be hazardous to the drinking water supply in the entire Upper Rhine valley. This thesis was supported by the AWBR (Working Committee of water works) and is intended to provide a clear and reliable basis for the evaluation of contaminated sites. This thesis is the first publication to focus on the special needs of water suppliers transnationally in Switzerland, France and Germany. The spatial domain of this study follows the natural outlines of the system Upper Rhine Valley. The new requirements from the Water Framework Directive and the new Drinking Water Laws (TrinkWW) are also taken into account.
There are several advantages to a regionally integrated view of brownfields for water suppliers. The Analyses show only a small part of known brownfields to be within water protection areas. Brownfields are not distributed randomly but clustered in unprotected areas. Therefore the number of potentially hazardous sites can be reduced and resources can be focused on the remaining ones.
The method developed in this thesis incorporates two approaches.
The “suspected site approach” investigates the transport of contaminants from the site to the well: Known suspected sites get intersected with water protection areas and flow paths get calculated. These data feed into a transport prediction for the sites.
The “supplier oriented approach” focuses on the well and investigates the hydrological dynamics of the aquifer using the Analytical Elements Method for groundwater modelling and a multi-tracer approach.
After an initial prioritisation of suspected sites, the method was applied to four water works with diverse hydrological dynamics. The first test case was the water protection area Zartener Becken. The hydrological dynamics are dominated by two rivers, the Dreisam and the Eschbach. The dual influence is reflected in a bimodal age distribution with peaks at 70 days and three to four years mean residence time (MRT).
At the water works Hausen six wells were investigated which could be assigned to two groups, one of which is heavily affected by river infiltration. Measurements of Krypton-85 and Argon-39/37 indicate the inflow of another groundwater component with higher MRT.
The hydrological dynamics of the two Swiss study sites are heavily influenced by artificial infiltration of Rhine water. In one area, Langen Erlen, the areas of infiltration keep changing whereas in the other, Hardwald, they are fixed. In both areas irrigation gets suspended at high stage in combination with high turbidity and at “Rhine alarm”. Time series analysis of oxygen-18 and deuterium yields good results in these areas. However, the simple AEM approach to groundwater modelling proved insufficient to deal with the large instationarities from irrigation and results come with large uncertainties attached. An age distribution could not or only partially be modelled. This shows the limits of the AEM at the current state of knowledge. Gas tracer analysis showed a local increase in SF6 and FCKW around Basel. As no input data are generally available, a method for data correction was applied. At Langen Erlen, flow times in the meadows could be shown to be below two days at high spate and without irrigation. At the Hard, the hypothesis that only infiltrated Rhine water gets pumped from the well could be rejected. However it was not possible to quantify the contribution from the Rhine, as there was no undisturbed (not influenced by artificial infiltration) end member from the shell limestone.
A risk assessment of the suspected contaminant areas within the study sites was performed in two steps. The flow paths got calculated based on the groundwater modelling. For paths which approach a well or may influence it by dispersion, a dispersion model for the main contaminants of the site was fitted. The maximum permissible value of the TrinkWV or the guide value of the WHO was used as a boundary condition. From this one can calculate the amount of hazardous material which would need to infiltrate at the site to exceed the maximum permissible value at the well.
Groundwater modelling via AEM is still deficient in highly instationary areas and in areas with high slope in aquifers, but it is foreseeable that the AEM-work groups will tackle these problems in the near future. The advantages of the method are the correct flow path delineation and the effective age structure calculation. By assigning an age structure to the pumped water hazard zones and classes can be identified. The fraction of young groundwater, e.g. stemming from indirect groundwater regeneration, can be calculated. A mixture model can be fitted to the explicit age structure and assessed for its risk to drinking water.