Comparison of RIMPUFF, HYSPLIT, ADMS atmospheric dispersion model outputs, using emergency response procedures, with 85Kr measurements made in the vicinity of nuclear reprocessing plant

• Published in: Journal of environmental radioactivity Vol. 124; pp. 266 - 277
• Main Authors: Connan, Olivier, Smith, Kilian, Organo, Catherine, Solier, Luc, Maro, Denis, Hébert, Didier
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01-10-2013; Elsevier
• Subjects: 85Kr; ADMS; air; Air sampling; Analysis methods; Applied sciences; Atmospheric pollution; Atmospherics; Dispersions; Emergency response; Exact sciences and technology; Experimental measurement; HYSPLIT; Life Sciences; Numerical control; Pollution; Radioactivity; Reprocessing; RIMPUFF; wind direction; wind speed; France; Europe; Experimental measurement; RIMPUFF; 85Kr; HYSPLIT; ADMS; Performance evaluation; Spent fuels; Radioactive waste; Fuel reprocessing plants; Pollutant behavior; Kr; Measurement campaign; Nuclear fuel; Hazardous waste; Radioisotope; Forecast model; Krypton 85; Emergency procedure; Fission reactor fuel reprocessing; Air pollution; Atmospheric dispersion; Comparative study
• ISSN: 0265-931X; 1879-1700
• DOI: 10.1016/j.jenvrad.2013.06.004

The Institut de Radioprotection et de Sureté Nucléaire (IRSN) performed a series of 85Kr air sampling campaigns at mesoscale distances (18–50 km) from the AREVA NC La Hague nuclear reprocessing plant (North West France) between 2007 and 2009. The samples were collected in order to test and optimise a technique to measure low krypton-85 (85Kr) air concentrations and to investigate the performance of three atmospheric dispersion models (RIMPUFF, HYSPLIT, and ADMS),
This paper presents the 85Kr air concentrations measured at three sampling locations which varied from 2 to 8000 Bq m−3, along with the 85Kr air concentrations output by the dispersion models. The dispersion models made reasonable estimates of the mean concentrations of 85Kr field measurements during steady wind conditions. In contrast, the models failed to accurately predict peaks in 85Kr air concentration during periods of rapid and large changes in wind speed and/or wind direction. At distances where we made the comparisons (18–50 km), in all cases, the models underestimated the air concentration activities.
•Measurements of 85Kr in air between 18 and 50 km from the release source.•Comparison of experimental results with RIMPUFF, HYSPLIT and ADMS model outputs.•Utilisation of atmospheric dispersion models under emergency response conditions.•In steady wind conditions, the models produce realistic results.•Overall, the models underestimate the measured activity concentrations.