An atmospheric dispersion codes comparison applied to incineration of urban solid waste containing high activity radioactive sources

Incineration of urban solid waste, accidentally contaminated by orphan sources or radioactive material, represents a potential risk for environment and public health. Moreover, production and emission of radioactive smoke plumes can determine a heavy contamination of the plant, leading to important economic damage. In order to prevent these hazards, in February 2004 a radiometric portal to detect radioactive material in incoming garbage trucks was installed at AMSA “Silla 2” urban solid waste incineration plant of Milan. In this paper the dose committed to population and the environmental effects, due to incineration of a high activity source in absence of the radiometric portal, have been evaluated. Furthermore, results provided by the easy-to-use software Hotspot, specifically designed to supply a fast and simple tool for radioactive atmospheric releases evaluation, have been compared with the results provided by the more complex code Calpuff, the EPA reference code for atmospheric dispersion. A 3,7 1010 Bq (1 Ci) source of 137Cs has been considered as reference. The maximum total dose committed to a single individual, due to inhalation and cloud submersion, appears to be negligible [less than 10-8 Sv]; nevertheless a light environmental contamination would involve a large area, leading to a significant exposure due to ingestion of contaminated foods. Hotspot and Calpuff have been compared varying the most important meteorological conditions, mainly stability class and wind speed. Considering the high uncertainties affecting atmospheric dispersion calculation, the agreement between the two codes appears to be fine, although in low wind conditions [i.e. lower than 1 m s-1] relevant differences arise up to one order of magnitude. Moreover Hotspot results are strongly influenced by some input variables such ad deposition velocity and scavenging coefficient. These parameters show a large variability depending on the physical-chemical form of the considered radioisotope. Exact determination of these values is neither fast nor easy and default values are frequently inadequate. Uncertainties due to improper values of these parameters can be much greater than uncertainties due to the atmospheric dispersion modelling. Therefore this aspect of the easy-to-use software Hotspot arises to be critical.