Groundwater pollution from multiple point sources is often characterised by scattered and unevenly distributed concentrations, which makes source identification more difficult, especially when sampling results are not entirely reliable. The first part of the research focussed on improving scientific knowledge and understanding of characteristics and variability of intraborehole flows (IBFs). The results of borehole logging in the Ljubljansko polje aquifer showed that vertical hydraulic gradients are not limited to complex aquifer systems, but can also occur in thick alluvial aquifers without continuous aquitard layers. Direction and velocity of these flows are strongly influenced by the position and length of the screen sections and the morphology of the bedrock. Numerical modelling has shown that velocities of IBFs can be up to 4 orders of magnitude higher than the vertical component of groundwater flow velocity in the aquifer. Ambient IBFs can also be significantly higher than conventional pumping rates during sampling, which affects the representativeness of sampling. Using a double packer system, the example of hexavalent chromium [Cr(VI)] showed that measured concentrations can deviate by 10-15 µg/l from the actual value. The second part of the research focussed on the spatial interpretation of Cr(VI) contamination in the Ljubljansko polje aquifer. Based on depth-discrete sampling and particle tracking in a numerical model, it was found that there are characteristic differences in the distribution of Cr(VI) along the depth of the aquifer that can be associated with different sources. The aquifer is intertwined with remnants of old contamination from the 1980s and the effects of potentially still active sources. Based on long-term trends, the Cr(VI) retardation factor in the extensive and well-permeable intergranular aquifer was estimated by analytical modelling to be 9.5 with a partition coefficient between 1.06-1.33 l/kg. Rough estimates of the potential locations of still active sources were made in three characteristic longitudinal cross-sections.
|