Experimental Qualification With a Scale One Mock-Up of the “Measurement and Sorting Unit” for Bituminized Waste Drums

Within the framework of the cleaning operation of the Marcoule reprocessing plant UP1 (France), the CEA (French Atomic Energy Commission) developed a measurement system for 225-liter drums filled with bituminized radioactive sludge originating from the effluent treatment. This work was carried out for the CODEM, which is an economic interest group made up of CEA, EDF (the French public utility) and COGEMA (the operator of UP1). CODEM is in charge of UP1 dismantling operations, especially waste retrieval. The bituminized waste drums mainly contain plutonium, americium, uranium, curium and various beta emitters among which some are responsible for significant gamma irradiation, such as ¹³⁷Cs. The aim of this system is to sort the packages according to their radioactive level, so as to direct them towards the French Aube Center, which is a surface repository. This means they must meet the acceptance criteria related to their activities. Otherwise, they will remain in interim storage in Marcoule, pending the choice of a final mode of management (e.g. underground disposal). The assay system, called UTM (the French acronym for “Measurement and Sorting Unit”), consists of three stations devoted to active gamma imaging, gamma-ray spectroscopy and combined passive / active neutron measurements. After nearly 3 years of optimization and design studies [1], the CEA has built a scale one mock-up of UTM, called SYMETRIC. The purpose was to validate the performances formerly assessed by numerical simulation, mainly with the computer code MCNP [2]. We present here the experimental results obtained with SYMETRIC for five real bituminized waste drums. These confirm the expected performances in the measurement time assigned for each assay, which is limited to 1200 seconds. With the help of gamma imaging, we are able to determine the density of the bituminous mix with an uncertainty of ± 10% for a confidence level of 95%. We can also measure the filling height with an accuracy of ± 2 cm. These data allow us to correct matrix effects in gamma and neutron measurements. For these assays, the main results concern the detection limits and measurement uncertainties on ²⁴¹Am, ²³⁹Pu and ²⁴⁰Pu. These radioisotopes represent the major part of the total alpha activity, which is a very sensitive parameter for surface disposal limited to a maximum level of about 10 GBq per drum. The alpha activity must be calculated after a radioactive decay of 300 years, which is the survey period of the French Aube Center. If we can detect the former isotopes, the uncertainties on their measured activities are roughly 50%. If not, the detection limits are around a few GBq. These performances are sufficient to allow the sorting of the drums to either surface repository or interim storage. However, in order to increase the margin between the detection limits and the acceptance criterion on the total alpha activity, additional studies on the optimization of the measurement performances will be carried out. In this context, the experience gained with the SYMETRIC mock-up will be very helpful.