Integrative Passive Samplers to Detect Chlorinated Hydrocarbon Contamination in Karst

A large number of former industrial sites and landfills are responsible for aquifer contamination with chlorinated solvents in Switzerland. 16.7 % of them are located in karstic areas. But, the detection of such contaminations in springs or boreholes can be difficult because concentrations are often varying highly and rapidly. Measuring low concentrations during a few measurement campaigns is not sufficient to prove the absence of contamination. To overcome this difficulty, the performances of integrative passive samplers were investigated in the framework of the Chlorokarst project funded by the Swiss Federal Office for the Environment. Three types of devices were tested. They allow measuring time-weighted average concentrations of pollutants over long periods. Experiments in karst springs show that the Sorbicell samplers developed by the Sorbisense company were capable of measuring average concentrations over a 2-month period, matching the averages of the concentrations measured in daily snapshot samples collected with an automatic sampler taken as reference. Equipment of the springs with the Sorbicell devices was found to be easy. More difficulties were encountered when exposing the Sorbicell samplers in piezometers, because of the impact of the high fluctuations of the water column above the sampler on the filling rate. Adaptation of the device was thus carried out with the manufacturer, to reduce the effect of the water column, using air-vents and cartridges of high resistance. Satisfactory results were then obtained in terms of sampling rate. Uncertainty still remains on the estimated concentrations provided by the system since they can vary by a factor of two depending on how the volume of water that passed through the cartridge is estimated. Two other types of passive samplers were tested, the ceramic dosimeters developed at the University of Tubingen and commercialized by IMW, and the Tips developed by the AGI company. They were both found to be inappropriate to estimate reasonably well the mean concentrations in chlorinated solvents in these types of environment, even if they can still be used to detect the presence of contamination.