Evaluation des flux de biphényles polychlorés (PCBs) dans les affluents du lac Léman à l’aide de capteurs passifs

The aim of this study is to investigate Lake Geneva’s tributaries using passive samplers in order to establish polychlorinated biphenyl (PCB) fluxes. Silicon rubber sheets are therefore deployed during six weeks at the mouth of eight tributaries and in four waste water treatment plant (WWTP) effluents discharging directly into the lake. Concentrations of absorbed PCBs in the passive samplers allow evaluating the dissolved aqueous concentrations. The total (dissolved and particulate) concentrations in the WWTP effluent are also evaluated directly through a liquid-liquid extraction, in order to compare these concentrations to the ones measured by the passive samplers. Results show that Lake Geneva’s tributaries present a low pollution in comparison with international references, with concentrations for the sum of the seven indicator PCBs between 0.09 and 0.49 [ng/L]. However, the Rhône River and the Canal Stockalper, with respectively 0.28 and 0.49 [ng/L] of PCBs, are considered as some of the most polluted Swiss Rivers. The aqueous PCB concentrations measured in WWTPs using passive samplers are low compared to the rivers, between 0.04 and 0.13 [ng/L], excepted the WWTP of Terre Sainte (3.41 [ng/L]). The comparison of the aqueous concentrations measured with the passive samplers with those obtained through liquid-liquid extraction (between 2.11 and 9.98 [ng/L]) show the differences between dissolved and total concentrations. This important difference questions the capacity of passive samplers to evaluate the dissolved concentration in WWTPs. It is likely that this is only possible under certain conditions, fulfilled in the WWTP of Terre Sainte. By evaluating the PCB flux entering the Lake, the four rivers having the highest discharge (i.e. the Rhône, Dranse, Aubonne and Venoge Rivers) represent the major PCB inputs, respectively 4.7, 0.17, 0.36 and 0.24 [g/d]. It would therefore be interesting to investigate further in these rivers in order to determine the precise PCB sources. The extrapolation of the fluxes to other tributaries is based on the discharge, the number of effluents and the number of landfills in the watershed for the rivers. Concerning the WWTPs, the extrapolation is calculated using the population equivalents. These calculations allow a general view of the fluxes entering the lake as well as determining other rivers which represent a large input for further analysis: The Promenthouse, Dullive, Grand Canal, Vuachère and Boiron Rivers. Concerning the WWTPs, those in Montreux, Gland and especially Thonon should be analyzed further, based on the estimated fluxes in these effluents. A projection of the aqueous concentration in the lake gives values of 37.5 [pg/L] in winter and 21.3 [pg/L] in summer. By considering a bioconcentration factor for the different PCBs, the concentration in the fish is determined to be about 4.6 [ng/g] in winter and 2.65 [ng/g] in summer. This is much lower than the limit of 125 [ng/g] issued by the European Union.