Internal Dosimetry in Nuclear Fuel Fabrication : Occupational Exposure to Uranium Aerosols

Sammanfattning: The production of nuclear fuel can be associated with occupational exposure to ionizing radiation from radioactive decay of uranium. Such exposure must be sufficiently low and radiation doses adequately determined. Radiation doses from internal exposure, i.e., following intake (usually inhalation), cannot be estimated using dosimeters, but must be calculated based on indirect measurements in combination with biokinetic models.Such biokinetic models have been developed and refined for decades. Good knowledge of the material characteristics is crucial. However, the physicochemical properties of chemical compounds can vary between different production facilities. Aerosol size distributions and dissolution characteristics in lung fluid are of particular importance. The latter is important since dissolved material is absorbed to blood, whereupon a large fraction reaches the urine after filtering by the kidneys. This enables urine sampling as a method to monitor occupational exposure.The aim of this thesis was to investigate the physicochemical properties of uranium aerosols and their implication on internal dose assessments at a nuclear fuel fabrication plant in Sweden. Uranium aerosols were sampled and size fractionated using personal cascade impactors carried by workers at the factory’s different main workshops. Aerosols were studied using scanning electron microscopy in Paper I. In Paper II the activity size distributions were determined and in Paper III dissolution rates in simulated lung fluid were investigated. Paper IV is an internal dose assessment based on records of urine sample analyses from about 10 years of routine occupational exposure monitoring of uranium pelletizing workers at the site.For a median worker, the urinary daily excretion rate of uranium increased due to chronic exposure for about 1000 days, after which the excretion rate stabilized. This suggests that inhaled material dissolves in the respiratory tract rapidly enough to prevent a net buildup in the lung after several years of exposure. This could be modelled using the default recommendations for uranium oxide materials provided by the International Commission on Radiological Protection. However, the best model fit to measurement data was obtained using a different set of parameters, that showed some discrepancies with results from Papers II-III. For individual cases, excretion rates could vary between sampling occasions to a greater extent than predicted using the default recommendations, which could indicate a more rapid body clearance than expected. Whether this is an effect of experimental methods or simplifications in the biokinetic models should be further investigated in future work.