At the end of process equipment dismantling, the complete decontamination of nuclear facilities requires the radiological assessment of residual activity levels of building structures. As stated by the IAEA: “Segregation and characterization of contaminated materials are the key elements of waste minimization”.
From this point of view, the set up of an appropriate evaluation methodology is of prime importance. The radiological characterization of contaminated premises can be divided into three steps. First, the most exhaustive facility analysis provides historical and qualitative information. Then, a systematic (exhaustive or not) control of the radiation signal is performed by means of in situ measurement methods such as surface control device combined with in situ gamma spectrometry. Besides, in order to assess the contamination depth, samples can be collected at several locations within the premises and analysed. Combined with historical information and radiation maps, such data improve and reinforce the preliminary waste zoning.
The relevance of the geostatistical methodology relies on the presence of a spatial continuity for radiological contamination. In this case, geostatistics provides reliable methods for activity estimation, uncertainty quantification and risk analysis, which are essential decision-making tools for decommissioning and dismantling projects of nuclear installations.
Besides, the geostatistical framework provides answers to several key issues that generally occur during the clean-up preparation phase: How to optimise the investigation costs? How to deal with data quality issues? How to consistently take into account auxiliary information such as historical inventory? How to integrate the remediation support into the modelling? How to quantify uncertainties in the remediation costs while computing contaminated volumes?
This geostatistical approach is currently applied to several former nuclear facilities of the CEA in France. The ATUE (enriched uranium workshops) premise, located in Cadarache, is a case in point. Focusing on this premise, the presentation deals with the geostatistical methodology and its added value to get a reliable mapping of the contaminated areas and estimate the corresponding waste surfaces and volumes.