·Home ·Table of Contents ·Nuclear Industry

Application of Transmission Tomography to Nuclear Waste Management Ph. Rizo, Ch. Robert-Coutant, V. Moulin, R. Sauze, M. Antonakios. LETI (CEA-Technologies avancées) 17 rue des Martyrs 38054 Grenoble, FRANCE. Contact

Abstract

This communication presents different possible applications of X-ray transmission computed tomography (CT) for the inspection of nuclear waste. Each tomographic system is specific to the size and attenuation of the container to be examined. For light containers containing low activity waste, dual energy transmission (active) CT allows to correct for the attenuation in spectroscopic measurements (passive CT) and to characterize the structure of the waste. For large cylindrical concrete containers up to 120 cm diameter, tomography can detects voids and cracks, it also allows to compute the attenuation correction for radio elements emitting at energies higher than 200 keV. On large parallelepiped containers of 5 or 10m³, X-ray attenuation is too high with respect to transmission CT with a translate-rotate system. This last type of container will require an alternative to classical transmission CT, such as a tomosynthesis approach.

Introduction

Nuclear waste management requires a perfect knowledge of the contents of each container and of its internal structure. This information is required in order to ensure that the waste container can be stored safely in the dedicated storage area. Three types of information must be provided on the container and its contents
• Activity of the different radio-elements and position of these radio elements.
• Structure of the container; detection and characterisation of voids and crack in the concrete matrix, localisation of waste, sizing of the void between the concrete inside the container and the cover of the container...
• Characteristics of the waste materials. For lightweight waste, metallic parts must be clearly identified. For waste to be compacted the presence of aerosol bomb must be detected, liquids and putrescible organic materials must also be detected.
Today only some characteristics are checked on all the containers and some others are obtained by a complete destructive examination of a few containers selected in the overall production. This kind of destructive examination is time consuming and very expensive. Moreover, it generates a large amount of potentially radioactive waste. In order to reduce the number of these destructive controls and to increase the quality of the systematic control, tomographic systems are progressively adapted to the examination of nuclear waste containers.
In this paper, we first briefly recall the principle and potentiality of transmission tomography. We explain the kind of information which can be extracted from tomographic examinations according to the size and density of the container. Then dividing the nuclear waste container in three groups: light weight containers, heavy cylindrical concrete containers and large containers (5 m³ and 10 m³) with a parallelepiped shape, we present tomographic approaches dedicated to each family. We will conclude by stressing the limitations of tomography in the on line examination of waste containers, giving orientation towards other non-destructive X-ray methods which could possibly be coupled to tomography.

Potentialities of tomography

Principle of tomography


Fig 1: principle of translate rotate tomographic systems

In transmission X-ray tomography, X-rays are attenuated through the object and detected by a dedicated system. From the set of all the attenuation measurements across a plane of the object, the attenuation coefficient of each point of the object in this plane can be reconstructed. In order to provide reliable reconstructed values, the attenuation must be measured through the object and out of the object. With translation rotation scanners (Figure 1) such as those used for large objects, the detector must perform measurements in and out of the container. This therefore leads to a very high dynamic range for the detectors. Today the maximum dynamic range which can be reached is about 5 decades. The attenuation of X-rays being a decreasing function of the energy, for energies lower than 10 MeV, the larger the object, the higher the required energy.(Figure 2)

Fig 2: mass attenuation of X-ray through aluminium.

Information provided by tomography according to the attenuation of the object.
Tomography allows to reconstruct the attenuation coefficient of each pixel in the object. Therefore it can give access to the detection of voids, cracks and any structure leading to a sensible variation of the attenuation coefficient. It can allow to locate radiation protections and actual position of waste in concrete containers. The main limitation of this technique is the detection of liquids and the chemical characterisation of materials.
Another information to be taken into account in tomography is that the attenuation coefficient depends on the density r and atomic effective number Z of the material at low energy up to 100 keV (photoelectric domain) and mainly on the density for energy from 200 keV up to 6 Mev (Compton domain). Models of this attenuation using energy, density and atomic effectives numbers have been presented by Mc. Cullough [1].
Therefore using dual energy tomographic acquisition with objects of low attenuation, photoelectric and Compton information can be combined in order to obtain the effective atomic number and the density on each pixel of the slice. In that case it will be possible to distinguish organic materials, mineral and metals. For heavy objects, low energy X-rays are completely attenuated through the object therefore only the density of each pixel of the slice can be estimated.
The geometric resolution of tomographic images is limited by the system geometry, the number of projections acquired and the resolution of these projections [2]. The density resolution of the reconstruction is limited by the number of photons reaching the detectors. Taking into account the preceding constraints, it appears that the type of information that will be extracted from tomographic measurements will depend on the container size and density. We distinguish hereafter three groups of containers requiring three different examination systems.

System dedicated to lightweight nuclear wastes

In the case of light containers containing low activity nuclear waste, the problem is to estimate precisely the actual activity contained in the container in order to define its processing. Another problem is to ensure that there is no massive metallic part or forbidden items such as aerosol bombs inside the container. The system (Figure 3) developed by LETI is based on two types of detectors [3].
• A linear detector using photodiode arrays coupled to Gd₂O₂S scintillator converting X-ray into visible light to perform radiographic and/or tomographic examination of the container in order to identify forbidden items. (Figure 4, 5)
• A set of 22 BGO scintillators coupled to photo-multipliers performing helical dual energy tomography of the container.

Fig 3: Acquisition system for radiography and dual energy tomography of light weight containers.	Fig 4: Detail of the linear detector based on elementary modules of photodiode arrays. 20 modules are coupled for the radiographic acquisition
	Fig 5: Photograph of a phantom of the light weight container. Centre: transmission tomography of this container. Right : radiographic image obtained on the light weight container with the linear array.

The dual-energy measurement are obtained from two successive scans with different high-voltage, current abd filtration of the X-ray source.

These two scans high and low energy are performed using two helical acquisitions. These two tomographies provide an equivalent low resolution image of the container describing each voxel as a mixture of reference materials material, here Plexiglas and iron (Figure 6 ). This material decomposition allows to compute, for each pixel, its attenuation value from 60 keV to 1 MeV. Therefore this dual energy reconstruction provides the exact attenuation correction factor to be used to compute the actual activity from spectroscopic measurements. It is estimated that a precision of 20% can be reached for the activity estimation with such an attenuation correction. To be compared to values up to 300% on non homogeneous containers [3]. This system provides a complete computation of the attenuation coefficient of each voxel of the container (45cm diameter, 80cm high) with a geometric resolution of 5cm in about 5mn. The goal of this system is to perform a systematic examination of all the container (20000 to 30000) containers per year.

Fig 6: Decomposition in material basis of the content of the container. The material basis used in this example is Iron and Plexiglas. Fig 7 (right): High energy tomograph which is being mechanically adapted to the examination of cylindrical containers with diameter up to 1.4 m.

System dedicated to waste in heavy cylindric concrete containers

For cylindric concrete containers with diameter lower than 120cm, the X-ray energy to be used is of few MeV. In that case the attenuation coefficient depends mainly of the density. Tomography will then be used mainly to detect voids, cracks. It can also be used for attenuation correction but only for radioelements emitting in the Compton domain.

The system developed by LETI is based on the following elements. (Figure 8).

• An 8 MeV linear accelerator providing 2000 Rad at 1m from the source.
• A series of 25 CdTe monocristaline semiconductors as detector (Figure 9). The flux detected by the crystals is linked to the variation of the resistance of the crystal.

Fig 8: Detail of the high energy CdTe detectors. The 25 detectors are focused on the X-ray source Fig 9: Reconstruction obtained on Container phantoms. Left: this phantom presents an attenuation similar to the most attenuating containers. Right: Reconstruction with the same system of a concrete phantom.

This tomograph is a translation rotation tomograph [4]. The maximum dynamics of the detectors is close to 10⁵. The presented mechanics is limited to objects 50cm in diameter. This mechanics is under modification to handle objects 1.4m diameter, 1.5 m high with a weight of 7.5 t. The estimated performances (extrapolated from representative scaled sample) are the following: geometric resolution 1.1cm³ , density resolution better than 3%. These values have been evaluated by simulation and validated using experimental acquisition on scaled phantons (Figure 9)

The acquisition time required for a complete examination of a concrete container using such a tomograph is about 1h30. This is not compatible with a systematic examination of all the containers. This kind of tomographic examination will be dedicated to complete examination of containers previously selected and will help to define the location of the destructive examination. Tomography will indicate where to saw the container. The on line examination of the complete set of containers if required should be done using a faster method such as high energy radiography and possibly an algebraic reconstruction using a limited number of projections.

System dedicated to heavy waste in 5 m³ and 10 m³ parallelipedic containers.

The last type of nuclear waste containers are parallelepipeds containing a large amount of heavy waste with a volume of 5m³ or 10m³. The dynamic range required on the detectors with such objects is much higher than 10⁵. This dynamics is therefore not compatible with the up to date high energy detectors technology. The required acquisition time for a tomographic reconstruction (about 50 h) is also not realistic. Computed tomography is therefore not realistic with such objects. The only possibility to obtain 3D localization and characterisation of structures is to use techniques which do not require very high dynamic detectors. This mean that local radiography of the container must be used. With local radiography the contrast on the image is generally low. The main problem is to ensure a high enough sensitivity of the sensor. The technique under study is tomosynthesis. (described by M. Antonakios in another paper of this conference). The main advantage of this technique [5] is that it can reconstruct relative density variations using local radiographs with a reduced geometric resolution and a loss of the actual density value of the elements in the object. This technique has been fully tested on solid rocket motors and is in the process of adaptation to nuclear waste containers.

Conclusion

In this presentation we presented the potentialities of computed tomography in the process of nuclear waste management. We showed that this technique can be used in an on-line complete examination for light weight containers, providing information on the density and the atomic effective number of the contents. We showed that high resolution tomography able to detect critical defects in concrete containers is not compatible with on line examination. Only the expertise of selected containers is realistic. Tomography is no more possible in the state of the art of the accelerator sources and detectors on 5m³ and 10m³ containers. The proposed alternative is to use high energy radiography based on large Gd2O2S converter screens optically coupled to CCD cameras. These systems can give access to radiographic images in a few seconds and are compatible with an algebraic reconstruction from a limited number of projections or with tomosynthesis. Such systems cannot reach the resolution of classical tomographs, but should be able to give access to the main structural information on the containers.

Acknowledgments:

Authors would like to thank ANDRA, COGEMA, CEA/DCC and DRET who have motivated and partly funded the development of the presented systems.

References:

1. Mc Cullough, Med, Phys, 2 (6) (1975) 307
2. G.T. Hermann, "image reconstruction from projections: the fundamentals of computed tomography", Academic press, New York, (1980)
3. Robert-Coutant & al., " Estimation of the matrix attenuation in heterogeneous radioactive waste drum using dual energy computed tomography", Nuclear Inst. And Methods in Phys. Res., A422 (1999) 949-956.
4. F. Glasser & al., "Application of Cadnium Telluride detectors to high energy computed tomography", Nuclear Inst. And Methods, A322 (1992) 619-622
5. M. Antonakios & al., "Real time tomosynthesis system applied to solid rocket motors examination", proceedings of the 7^th ECNDT, Copenhagen 26-29 May (1998) 186-192

© AIPnD , created by NDT.net

|Home|    |Top|