Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Evaluation of two sampling methods for the determination of gas oil mist and vapour in workplace air

Abstract Number 4898 at Conference "Environment and Health". Bridging South, North, East and West. Basel, Switzerland, August 19-23, 2013
: Urbanus, Jan; Berger-Preiß, Edith; Elend, Manfred; Koch, Wolfgang

Conference "Environment and Health" <2013, Basel>
Fraunhofer ITEM ()

Background: Inhalation exposure of workers to gas oil can occur in the refinery as well as during transport and use of the product. Exposure is characterized by the presence of aerosol determined by the high boiling compounds and vapour from the light end of the product and is typical for high-boiling liquids often referred to as semi-volatiles. The assessment of workplace exposure requires the determination of the total concentration, as well as partitioning of the vapour and the droplet fractions because of differences in the toxicity. Aim: Investigate performance of an existing and a novel sampling method under laboratory conditions: a conventional system of filter and XAD-adsorbent cartridge operated in series and a virtual-impactor (VI) combined with filter-XAD cartridges.
Methods: Using a test stand for reproducible and stable generation of gas oil mists from liquid samples the two sampling method were tested for three gas oil products of different boiling ranges at different concentrations and thermodynamic conditions. The material collected was analyzed using an infrared method. Results: The conventional method allows for measuring the total concentration of the gas oil atmospheres for concentrations larger than 0.5 mg/m³ for 8-hour sampling. Losses between 15 and 30 % were encountered when measuring the total concentration with the VI-method. In the conventional sampling system evaporation effects cause a shift of the aerosol-vapour partitioning towards the vapour phase of up to 50% and depends on the boiling range of the gas oil, the thermodynamic conditions during sampling and the sampling duration. In the VI-method the aerosol-vapour partitioning is nearly independent of these parameters. However, the partitioning coefficient may be influenced by the aerosol mass fraction in the submicron size range, falsely counted as vapour phase.
Conclusions: Although the novel method is better able to quantify the vapour-mist partitioning, losses in the sampler may lead to underestimation of ambient concentrations if unaccounted for.