NON DESTRUCTIVE MEASUREMENT OF URANIUM INSIDE
LARGE CONTAINERS FOR THE DISMANTLING OF
AN URANIUM ENRICHMENT PLANT
J.Romeyer Dherbey, M.Butez, R.Pasquali and M.Vilaine
Commissariat à l’Energie Atomique -
CE CADARACHE-DER/SSAE/bat 225 -
13108 St Paul lez durance -France-
E-mail :ldmncad@macadam.cea.fr
ABSTRACT
In the frame of the dismantling of an uranium enrichment plant using the gaseous diffusion technology, a non destructive assay was developed, in order to quantify the residual mass of uranium inside diffusion exchangers. The measurement campaigns were performed before and after ClF3 cleanings so as to get the cleaning efficiency. Two measurements were performed : A local measurement by active neutron interrogation using a californium neutron source, and axial/angular gamma scanning in order to check the homogeneity of the uranium distribution. Two diffusion exchangers were measured. The external dimensions were about 3.7 m long, 0.5 m diameter for the first one and 2.3 m long, 0.2 m diameter for the second one. The validation of the 2 techniques were made, at scale one, on spare diffusion exchangers (with no uranium inside), using reference sources( 235U and 152Eu). A comparison between calculated and experimental results permitted us to validate the computerised model. The final calibration was carried out using Monte Carlo calculations. The main results show a good homogeneity of the uranium distribution, and a good coherency between the gamma and neutron measurements (about 10%).
INTRODUCTION
In the frame of the dismantling of an uranium enrichment plant using the gaseous diffusion technology, a ClF3 cleaning is a way to recover the residual uranium inside diffusion exchangers. In order to evaluate the ClF3 cleaning efficiency, a non destructive assay (NDA) was developed. The NDA consist of:
A measuring campaign was performed on two diffusion exchangers, before and after ClF3 cleaning. The paper presents the description of the devices, the calibration process, and the comparison between the techniques.
DESCRIPTION OF THE PHYSICAL METHODS APPLIED AND OF THE
FACILITY MEASUREMENT
The external dimensions of the diffusion exchangers were about 3.7 m long, 0.5 m diameter for the first one (D1) and 2.3 m long, 0.2 m diameter for the second one (D2). The neutron system (Fig 2) consisted of fifteen 3He detectors (65 cm long, 2.5 cm diameter), surrounded by a cylindrical polyethylene envelop.
Figure 1. Delayed Neutron Signal as a Function of Time
Figure 2. Active Neutron Measurements
A 1.7 108 n/s Californium source was moved inside a guide tube by a spiral cable near the measurement area in order to induce 235U fission [1],[2],[3].We measured the delayed neutrons after the source was displaced inside a polyethylene cask. Using 3 guide tubes enable us to perform 3 angular measurements (120 °), near the half length of the two diffusion exchangers. Each measurement consisted of 25 source cycling (1800 s). The delayed neutron signal may be considered as the sum of six decreasing exponential, with a half life between 0.2 and 60 S, and mean energies between 25 and 500 Kev. The signal processing consists of searching for an amplitude term and a background term, by fitting a sum of exponential to the signal, the amplitude term being proportional to the fissile mass content (Fig 1). The gamma spectrometry measurements (fig 3) have been made with NaI detector, using the 185 Kev ray.
Figure 3. Gamma Measurements
8 angular measurements have been performed, at the same axial position than for the neutron assay, in order to verify the homogeneity of the uranium distribution. 7 axial measurements have also been made in order to verify the absence of accumulation. Each measuring time was 600 s.
CALIBRATION
The calibration has been made by Monte-Carlo calculation, considering an homogeneous distribution of Uranium inside the diffusion exchangers. As the gamma calculation was sensitive to the knowledge of the mean density, experiments have been performed on spare diffusion exchangers (having no uranium inside) to measure the real mean density, using an europium 152 source (122 Kev and 344 Kev rays). A 40 % variation in density, will give a 21% change in the gamma signal, and less than 1% change in the neutron signal. The neutron calculations have been validated by comparing an experiment with an uranium sample to a specific calculation. In this case the sample consisted of a 10 g uranium (93% enriched) plate (4.5 cm length, 1.6 cm wide, 0.075 cm depth), positioned at 120 ° of the californium source. The calculation /experiment ratio was -1.5 %for the D1 spare diffusion exchanger, and 12% for the D2 spare exchanger (the smaller one).
EXPERIMENTAL RESULTS
Gamma Spectrometry Measurements
The table 1 hereunder gives for each diffusion exchanger (D1 & D2), and for three cleaning steps (before cleaning, after one cleaning, after two cleanings) the following results :
The good agreement between the axial and radial measurements shows the absence of singular accumulation.
Table I. Gamma Spectrometry Results
Active Neutrons Measurements
The table 2 hereunder gives for each diffusion exchanger (D1 & D2), and for three cleaning steps (before cleaning, after one cleaning, after two cleanings) the following results :
Table II. Active Neutron Results
Comparison Between the Gamma and the Neutron Results
The table 3 hereafter gives for each diffusion exchanger (D1 & D2), and for three cleaning steps (before cleaning, after one cleaning, after two cleanings) the following results :
The good agreement between the gamma and neutron measurements confirm the homogeneity of the uranium distribution.
Table III. Gamma and Neutron Results
CONCLUSION
The present work enables us to evaluate the residual mass of uranium inside diffusion exchangers before and after cleanup. The experimental results showed a good homogeneity of the uranium distribution, and a good consistency between the gamma and neutron measurements. The neutron interrogation system turned out to feature low detection limits and suitable accuracy (in the present case the device could have detected a 99% cleaning efficiency). Furthermore, it is not difficult to operate in current industrial use.
REFERENCES