A new sampling system has been developed for the measurement of time-averaged concentrations of organic micropollutants in aquatic environments. The system is based on the diffusion of targeted organic compounds through a rate-limiting membrane and the subsequent accumulation of these species in a bound, hydrophobic, solid-phase material. It provides a novel and robust solution to the problem of monitoring in situations where large temporal fluctuations in pollutant levels may occur. Accumulation rates are regulated by choice of diffusion-limiting membrane and bound solid-phase material and have been found to be dependent on the physico-chemical properties of individual target analytes. Two separate prototype systems are described: one suitable for the sampling of non-polar organic species with log octanol/water partition coefficient (log P) values greater than 4, the other for more polar species with log P values between 2 and 4. Both systems use the same solid-phase material (47 mm C18 Empore disk) as a receiving phase but are fitted with different rate-limiting membrane materials (polysulfone for the polar and polyethylene for the non-polar analytes). The two systems complement each other and together can be used for sampling a wider range of organic analytes than generally possible using current passive sampling techniques. Calibration data are presented for both devices. In each case, linear uptake kinetics were sustained, under constant conditions, for deployment periods of between 1 and 9 days. The effects of water temperature and turbulence on sampling rates have been quantitatively assessed. The performance of the system was further investigated by means of field exposures for one and two weeks in marine environments where calibrated samplers were used to determine the time-averaged concentrations of the polar biocides diuron and irgarol 1051. The quantitative results obtained using the passive sampler were compared with those obtained using spot sampling.

The treatment performance of the Keiskammahoek Sewage Treatment Plant (KSTP), was assessed in terms of pH, conductivity, and COD and nutrients removal from the influent. The contributions from this and other smaller point sources in the town to these parameters in the receiving Keiskamma River were determined by simultaneously monitoring the parameters in the river over a period of about 1 month. The COD and orthophosphate in effluents exceed the SA Effluent Quality Standards for these parameters in effluents to be discharged into a river. Also, significant pollution of the receiving Keiskamma River was indicated for orthophosphate, COD and NH 4-N.

Passive sampling systems are an emerging technology for detecting pollutants in the aquatic environment. A passive sampling system has been developed based on diffusion through a porous surface to a receiving phase, where the analyte is removed by chelation at a solid phase. The diffusion process can be described by Fickian diffusion through the sampler. The sampler has a well-defined surface area, which permits calibration in terms of concentration. Passive sampling systems can be used to determine pollutant concentrations if the diffusion process can be described and understood within environmental limits. In natural water systems, diffusion coefficients for metal transport across the porous membrane will be affected by external conditions, including biofouling and variation in turbulence and temperature. Uptake rates for the analytes Cu, Cd and Pb have been determined for the complete passive sampling system. Two different cases have been investigated, a batch case, where the bulk concentration decreases with time, and a flow-through case, where the bulk concentration remains constant. Diffusion coefficients were determined for the two conditions and compared with the calculated value obtained for the Stokes-Einstein equation in pure water. Diffusion coefficients for metals were found to be lower than for diffusion in pure water, a difference attributed to the effect of the porous membrane. The effect of the hydraulic conditions on the metal diffusion was studied for both a conventional magnetic stirrer creating turbulence in the system and for a rotated sampler, the latter providing a well-defined boundary layer system. The boundary layer was found to be negligible compared with the diffusion limiting membrane in the presence of sufficient turbulence or if the rotation of the sampler was high.