Assessing the impact of contaminated sites on karst - Feedback on the exposure of integrative passive samplers in 17 springs of the Jura mountains
Résumé
In Switzerland, many potentially contaminated sites are located in karst, especially in the north-west part of the country, in the Jura mountains. A large part of the contamination is made of chlorinated hydrocarbons used by the watch industry and microtechnology factories. Showing the impact of contaminated sites in these environments is challenging because pollutants often travel very rapidly over long distances (several km), leading to very low and highly fluctuating concentrations in the springs that form the outlets of the karst watersheds. These springs are nevertheless often the only points where groundwater pollution can be monitored, due to the difficulty to reach karstic conduits and the pollution with boreholes, within or at short distances from the contaminated sites. Since 2011, the Swiss federal office for the environment is funding the Chlorokarst project on this topic. The purpose of the project is 1) to test and experiment devices that may be helpful to assess the impact of contaminated sites on karst groundwater, and 2) to elaborate a regulatory approach for estimating the need of monitoring or remediating these sites in these specific environments. Among all the tested devices, integrative passive samplers were found to be useful to measure very low levels of chemicals, despite strong variability with time, over exposure periods of 2 months. Compared to the 2 other types of samplers that were tested in the springs, the Sorbicell® were found to show the best performances for detecting the chlorinated hydrocarbons. A large set of Sorbicell were then installed in 2016 in 17 springs of the Jura mountains, where a hydraulic connection with contaminated sites was suspected or proven. Samplers were exposed in 2 periods of 2 months, in low and high water conditions. Beside the chlorinated hydrocarbons, a broad range of organic and inorganic chemicals were analyzed. Standard snapshot samples were also collected during the period. In the first series of analyzes (low water), chlorinated hydrocarbons were measured in 3 of the 17 springs, at very low levels, close to the limits of quantification (LOQ). Moreover, inorganic chemicals (i.e. lead, mercury, boron, phosphorus) were found in all the springs at low levels (results of the second series are expected in the late 2016). These results show evidence of human disturbance on water, even if it is at very low concentrations. In conclusion, the Sorbicell samplers were found to be very useful to detect the low and highly fluctuating levels of chemicals in karst springs, provided that the limits of quantification were sufficiently low. The LOQ vary for each passive sampler unit, because they mainly depend on the volume of water that percolates in the sampler during the exposure period: the highest the volume, the lowest the LOQ. It is thus important to control exposure conditions in order to get the higher percolation volume without exceeding the maximal capacity of the device.