Monte Carlo modelling and experimental measurement methodologies to support decommissioning plans of biomedical cyclotrons

When considering the dismantling of a PET cyclotron facility, considerable amount of low level solid radioactive waste has to be characterized and disposed of. Prediction of induced radioactivity is a challenging task since the activity produced varies considerably, depending on the type of accelerator, on its use and on the specific structure of the bunker: for this reason, each facility require its specific decommissioning strategy. This work is aimed at developing a Monte Carlo approach to a preliminary assessment of activation, to define an ad hoc decommissioning strategy and to identify possible countermeasures to be taken during the construction phase of the facility. In this work two main cases studies were analysed: The GE PETtrace facility of the Sant’Orsola-Malpighi Hospital (Bologna) and the IBA CYCLONE 18/9 facility of Inselspital (Bern). The Monte Carlo code FLUKA was used to model accurately the two facilities for activation assessment. Different kinds of experimental measurements were performed to evaluate MC models reliability in terms of neutron field and neutron activation. The main long lived radionuclides founded inside the bunkers walls were 152Eu, 154Eu, 134Cs, 54Mn, 46Sc, 57Co, 65Zn and 60Co, for both facilities. The nuclides with the highest activity concentrations were 60Co and 152Eu. The total activity concentration estimated after 20 years of cyclotron operating life was up to 4.10 Bq/g and up to 3.22 Bq/g respectively in Bologna and in Bern, exceeding in both cases the radiological exemption limit of 1 Bq/g. Monte Carlo simulation proved to be a very powerful and feasible tool in the planning of new biomedical cyclotron installations and in the definition of an optimized decommissioning strategy. The availability of an experimentally validated Monte Carlo model makes it possible to revise the traditional approach to activation assessment.