Untersuchung der Ausgasung organischer Chemikalien in einer Fließgewässersimulationsanlage (English shop)

Preview

Table of Contents, Datei (45 KB)
Extract, Datei (60 KB)

Background: In running waters, volatilisation from the water phase is the most important loss process for some organic chemicals. The kinetics of volatilisation depend on substance properties (diffusivity, air/water partition coefficient Kaw) and environmental conditions (flow velocity, temperature, wind speed, geometry of the river bed, roughness of the stream bottom). To date, experimental investigations of volatilisation have only been available at laboratory scale.
Aim and research question: A systematic investigation of the volatilisation of reference chemicals from flowing water bodies.
What influence do water temperature, flow velocity and further hydraulic conditions have on volatilisation?
Materials and methods: Volatilisation was investigated at the running and standing water simulation facility (Fließ- und Stillgewässersimulationsanlage, FSA) of the Umweltbundesamt (UBA) in Berlin-Marienfelde. Individual environmental conditions were varied separately and in as controlled a manner as possible. The experiments were carried out in circulating flumes (length 104 and 154 m respectively, width 1 m). Constant flow velocities in the range of 0.15–0.45 m/s were set by means of a screw pump. The water temperature could not be regulated, but was recorded continuously. In some flumes the hydraulic conditions were varied, e.g. by filling with sediment. Via storage vessels, four highly volatile substances (MTBE, ethylbenzene, 1,2- and 1,3-dichloropropane) and four less volatile substances (2-methyl-1-propanol, 2-methyl-1-butanol, cyclopentanol, cyclohexanol) were homogeneously mixed into the water body (30 and 45 m³ respectively), after which water samples were taken at regular intervals. Analysis was performed by headspace GC/MS after addition of internal standards. The rate constant of volatilisation was determined by exponential regression; multiplied by the water depth, it yields the specific volatilisation velocity (vaw in m/s).
Results: For the highly volatile substances, relatively short half-lives of between 4 and 32 hours were observed. Volatilisation of the methyl alcohols and of the cyclic alcohols was considerably slower, with half-lives of between 2 and 40 days.
Modelling was carried out with an extended boundary layer model after Deacon (1977):
1/vaw = 1/vw + 1/(va * Kaw)
where vw and va are the phase-internal velocities in water and air respectively.
The air/water partition coefficient Kaw was estimated as a function of temperature by means of pp-LFER equations (poly-parameter Linear Free Energy Relationships) (Goss, 2006). The water-side resistance 1/vw depends on the diffusion coefficient in water, the viscosity of water, flow velocity, hydraulic radius and river bed roughness. The air-side resistance 1/va depends on the wind speed (at 10 m height), the flow velocity and the diffusion coefficient in air.
The model was fitted by means of non-linear optimisation, with only two regression constants and five different river bed roughnesses being adjusted. The best fit of the model was obtained at 0.157 for the first regression constant – this is only slightly lower than the value of 0.161 given by Moog & Jirka (1999).
Discussion: For the highly volatile substances a clear effect of the flow velocity was observed, which is presumably attributable to the influence of turbulence at the boundary layer between air and water on the water-side resistance. Especially for MTBE, volatilisation slowed down at the minimum water temperature of 5 °C, but the effect was considerably less pronounced. Because of the temperature dependence of Kaw, the air-side resistance additionally comes into play here at low temperatures.
For the volatilisation velocity of the alcohols, Kaw is of primary importance. The air-side resistance is indirectly reduced not only by the actual wind, but also by the “relative wind” of the water current.
Conclusions: Overall it can be stated that, with the exception of Kaw, the variability of environmental conditions plays a greater role for the volatilisation of organic chemicals than the variability of substance parameters.
In running waters, considerably more substances are subject to air-side control than is the case in standing waters. Thus, in running waters the critical Kaw is around 0.03 instead of 0.001.
The FSA is well suited for investigating the volatilisation of highly volatile substances. For less volatile substances such as cyclic alcohols the volatilisation velocities are too slow, so that distortion by long-term temperature changes is possible.