Modelling supported non-destructive in situ depth profiling of Cs-137 contaminated concrete

Abstract

This thesis evaluated the performance of in situ depth profiling of contaminated concrete components. In chapter 1, the introductory framework of this thesis was covered. Due to the ageing of the nuclear power plants (NPPs) in Europe, it is expected that during next decade more and more reactors will have to be decommissioned. [247] Especially for Belgium, there is an expected increase in decommissioning activities due to the nuclear phase out. Dismantling of a nuclear facility is different from the dismantling of a conventional facility due to the presence of radioactive material inside of the installation. The presence of this radioactive material has its implications on safety of works, and strict regulatory requirements apply to avoid exposure of the environment and public to ionizing radiation. This includes monitoring of dose to workers and follow-up of contamination levels. Also, the generation of RAW generation plays an important role during decommissioning projects. A RAW management system is required that can characterise the properties of waste in every step of the project, from initial characterisation, to final characterisations after the waste has been conditioned. The decommissioning strategy is often based on a waste lead approach, meaning that the minimisation of RAW is the central theme of performing the decommissioning works. This strategy is applied due to the large costs that are accompanied by the conditioning of RAW. For nuclear power plants, most of the volumes of generated RAW are low-level wastes. Within this waste fraction, concrete is identified as a major contributor to the overall volume of waste. This concrete waste includes activated concrete coming from areas that have been exposed to a neutron flux (such as the biological shield) and concrete which has, due to contact with radioactive materials become contaminated. In order to limit the amount of generated waste, contaminated concrete structures are frequently decontaminated by removing a top layer from the surface. As a result, only the removed fraction will be disposed of as RAW whilst the remaining structure can be disposed of as conventional waste. In order to avoid having successive steps of material removal followed by characterisation measurements to assess whether free release levels have been reached, it is important to perform contamination depth profiling before starting decontaminations. The scope of the thesis was further specified based on two cases which cover the characterisation and decontamination strategy of concrete materials. The first covered the decommissioning of Auxiliary buildings 6A and 6B of the reprocessing plant Eurochemic. Here, measurements were performed by using handheld contamination monitors and code drilling samples were taken from regions with the highest contamination levels. The second case was the decommissioning of the concrete floor of a waste gas surge tank and liquid tank collection room. Here, contamination depth estimations were performed with non-destructive in situ measurements using the multiple photon peak method applied to 137Cs. The main reason for the evolution of supplementing core drilling sampling with in situ measurements is the limited information on spatial distribution that can be obtained from core drilling. although in situ measurements are a supplementary method which cannot fully replace the core drilling, the volume reduction as a result of the in situ measurements however clearly shows the value of this technique. Although in situ measurements have already proven their use in decommissioning projects, improving the performance of in situ measurements in decommissioning projects has been identified as a key R&D field that would improve the waste management in decommissioning projects. In chapter 2, the materials and methods were described. The multiple photon peak method was selected to be the in situ technique that will be evaluated. This method has its origin in in situ measurements of contaminated soils and was first developed for environmental monitoring of fields that were contaminated after the Chernobyl accident. There are fundamental differences when applying this method to concrete, mainly with respect to the inhomogeneous nature of concrete due to its internal structure consisting of aggregates, mortar and voids. It was opted to study the performance of the multiple photon peak method by using a Monte Carlo model. Mainly the diversity of the modelling approach with respect to concrete composition, radionuclides inventory and contamination depths favoured the modelling approach combined with practical limitations that were linked to the possible contamination of lab equipment. The main objective of the thesis was: the development of a Monte Carlo (MC) model that is capable of quantifying the impact of the concrete internal composition on the performance of non-destructive in situ depth profiling methods. An extended-range (XtRa) coaxial p-type HPGe detector was selected to be modelled in the MC model due to its high resolution and capability of measuring X-rays. As no standard type of concrete mix exists, the concrete mixing that was used during this work was based on the recipe of Chooz B NPP, consisting of CEMI type cement and limestone aggregates. CT was chosen as a technique that could visualise the internal composition of concrete. The CT scans were performed using a Philips Brilliance Big Bore 16-Slice CT Scanner with a “SRS-HEAD” protocol. The CT-scan images of the concrete samples were integrated in the MC model. TOPAS MC was selected as the MC code that was used in this thesis because of its origin in medical applications which allows easy incorporation of CT-scan images into the model. TOPAS also provides the flexibility of creating custom physics processes and particle sources which were modelled in order to create contamination profiles in the concrete CT-scan images. A series of sub objectives were identified which were the scope of the next chapters. Chapter 3 covered the design and validation of the MC model of the HPGe detector. The sub objective of this chapter was: Designing, validating and optimizing a High Purity Germanium (HPGe) detector model in a Monte Carlo simulation tool . Although the rest of the thesis focussed on 137Cs, the goal of the HPGe model was to create a model that can also be used for other radionuclides. The HPGe geometry was modelled based on the technical drawings provided by the manufacturer. The model was validated by comparing experimental FEP efficiencies with efficiencies calculated from the TOPAS MC simulations. Three different geometries of radionuclide sources, placed at different heights from the detector endcap, were used to validate the model. The model was tested for radionuclides: 134Cs, 152Eu, 139Ce, 60Co, 137Cs, 113Sn, 85Sr, 57Co, 51Cr, 88Y, 133Ba, 109Cd, 241Am and 210Pb. These sources were chosen as they had well known activity levels and also provided gamma-ray emissions over a broad energy range. As TOPAS MC had not been used for gamma-ray spectrometry applications, a custom physics list was designed which would enable the simulation of radioactive decays. The simulations of decays were used in combination with simulations of single photons in order to calculate and correct for coincidences in the HPGe detector. The results show that the MC model performs well and adheres to 5% relative difference for the efficiency of gamma rays with an energy of between 100 keV and 2000 keV and 10% relative difference for gamma rays with energies lower than 100 keV or higher than 2000 keV. The designed HPGe detector model was next used in chapter 4. The sub objective of this chapter was: Incorporation of the concrete sample internal structure into the MC model Similar to chapter 3, the goal was to design a model that can also be used for applications beyond the scope of this thesis. For chapter 4 this meant that the method for incorporating the concrete internal structure in the MC model should also be transferable to other types of concrete materials. The model was also created so that it can incorporate a broad number of contamination and/or activation profiles. A CT density reconstruction was made specifically to be used for the CT scans of the concrete samples. The density reconstruction determines the relation between CT number and material density that will be used in the MC model. The performance of the incorporation of the concrete in the model was evaluated by comparing experimentally determined linear attenuation coefficient to linear attenuation coefficient calculated from the MC model. The attenuation coefficient of different concrete samples was experimentally determined with a narrow-beam setup in an energy range of 609 keV to 2447 keV. Fluctuations of up to 16.8% in the attenuation coefficients were observed between the different samples. The same setup was modelled in TOPAS MC. The concrete structure was included in the model by incorporating the CT scan images and the CT density reconstruction. Results of the attenuation coefficients determined by the MC model closely match experimental values and the same fluctuating trend between the attenuation coefficients of the different samples was observed. The fluctuations in the attenuation coefficients were also linked back to the concrete internal structure. For samples with low linear attenuation coefficients, mortar was visible in the CT-scan images at the beam location. Correspondingly, aggregates were found to be present at the beam location for samples with high attenuation coefficients. The results of chapter four show that the chosen methodology of incorporating the CT-scan images in the MC model is useful for studying the internal composition of concrete and gives realistic results for the attenuation properties of different phases in the concrete. The designed model of the HPGe detector and the methodology for incorporating the CT-scan images of concrete were combined to construct the MC model that is described in chapter 5. The sub objective of this chapter was: Quantifying and evaluating the impact of the concrete’s internal structure on depth estimation parameters. In chapter 5 the performance of the multiple photon peak method was evaluated, for determining the contamination depths of 137Cs in concrete materials. The MC model was used to examine the influence of the collimator design, the depth of the contamination and the influence of the concrete internal structure on the multiple photon peak method. The performance of the multiple photon peak method was shown to be influenced by contaminations that were present at the edge of the field of view of the collimator. Here, an increase in the X/ ratio was observed which is linked to the field of view (FOV) of the 32 keV and 662 keV photons which changes differently for both energies. The high self-absorption of the 32 keV X-ray spectral peak proved to be the limiting factor of the multiple photon peak method. Especially for deeper contamination depths, the detection efficiency of the 32 keV X-ray peak was low. This self-absorption of the X-rays of 137Cs was also the main cause of fluctuations of the estimated contamination depths which were a result of differences in the concrete internal structure. Contamination depths were overestimated for aggregate-rich regions and underestimated for mortar-rich regions. When using the multiple photon peak method, care has to be taken to avoid selecting collimators with apertures that have aperture openings that are of comparable size to the aggregates that are used in the concrete as the concrete internal composition will influence the depth estimations. On the other hand, choosing too large collimators apertures will result in activity levels and depth estimations that averaged over a too large surface area. This has its implications on the amount of waste being generated during the decontamination. Also, for deeper contaminations, the self-absorption in the concrete makes it difficult to achieve an adequate statistical uncertainty on the 32 keV spectral peak that is required to calculate a correct contamination depth.