Previous: what_we_do

PEREN - Project for an Experiment

The demonstration that the molten salt reactor system offers a great deal of potential has led our group to submit a project to create, at our Institute, an experimental platform dedicated to the gathering of fundamental nuclear and chemical data that are a prerequisite to the development of such a system. This project will be carried out within a collaboration that includes the CENBG¹, IPNO² laboratories, the ENSEEG³ laboratories LTPCM⁴ and LEPMI⁵, and EdF⁶.

Materials and Means

The experimental platform that we intend to build in Grenoble does not contain fissile matter so that it will not be a Nuclear Installation (Installation Nucléaire de Base), leaving wide possibilities open for manipulations on the platform. It will include a pulsed neutron generator and a configurable massive moderator block representing the core of a molten salt reactor with a graphite structure and several channels that can be used in various configurations and filled with various elements. In order to allow work with massive salts, a melting and molding system will have to be designed, permitting the introduction of salts in graphite cylinders. The purpose of the platform is to allow us to address and rapidly solve a number of neutronics and chemistry issues that arise in the development of such molten salt reactor systems.

Pulsed Neutron Generator

In order to study neutron moderation in space and time, the instant and location where a neutron is created must be known precisely. This implies that the charged particle beam is pulsed and has narrow pulse widths.

The specifications for the new neutron source are similar to the characteristics of GENEPI, except for the pulse shape:

: Peak deuteron current 50 mA.
: Pulse frequency adjustable to 10 KHz.
: Pulse descent time a few 10^-8 seconds.
: Pulse width on the order of 10^-7 seconds.

The last two requirements are imposed by the need to measure high lethargy light materials placed in the diffusion block.

The implementation of the new neutron generator will benefit from the experience accumulated in the design and development of GENEPI.

One of the solutions being considered in order to meet the pulse shape requirements specified for the new generator is to install an electrostatic deflection stage beyond the magnetic deflection system: a high voltage of 5 KV would be applied to the deflection plates within a 10 nanosecond time span. Indeed, it is unlikely that the pulse duration could be narrowed further as compared to the GENEPI one just by playing on the ion source parameters.

The electrostatic ion guide can be simplified in this project. In the MUSE experiment, the ions have to be guided over a 2 meter distance, this length being set by the dimensions of the reactor, while in the present project, the distance from the deflection outlet to the center of the diffusion block is only about 60 cm.

Ideally, Deuterium targets would be used most of the time, since the energy carried by the neutrons they produce is close to that of fission neutrons. However, their neutron yield may prove insufficient so that the possibility of Tritium targets will have to be kept. These would be used also for the measurement of inelastic scattering cross sections, if need be. Note that the tritium gas recovery installation is still available at ISN.

Massive Moderator Block

The scattering block will be composed essentially of high density graphite. It is ortho-cylindrical, its diameter is 1 meter. In it, seven 15 cm. diameter measurement channels will be drilled, and one central channel for beam entrance. Each measurement channel will be filled with the scattering material to be measured (C, CF2, Be, etc.) except for a 3 cm diameter hollow cylinder in the center in which a detector or a sample can be introduced in order to measure (n, $\gamma$ ) or (n,fission) events.

$\begin{figure}\begin{center}{\centering\includegraphics{bloc_graphite_web.eps} \par } \end{center}\end{figure}$

Handling of the Fluorides and Alloys

Two conditions must be met if dense filling of the channels with LiF-BeF2 salts is to be achieved:

: Operate in a closed environment, because of the toxicity of Be salt dust and because of the hydroscopicity of BeF2.
: Achieve uniform salt solidification in graphite channels.

This requires two devices:

: A glove box in which to handle and weigh the fluoride powders.
: An oriented solidification oven with a neutral atmosphere. (four à solidification orientée sous atmosphere neutre)

The glove box is a device that will have long term, multiple uses, for anything related to fluorides and metals sensitive to humidity and oxidation. It is likely that commercially available glove boxes will have to be modified to ease the transfer of materials from the glove box to the oven.

The oven will be useful not only for the filling (in several steps) of the salt channels but also for the preparation, in a vacuum, and the pre-melting of the salts and alloys. These preparations are always necessary for the chemical exchange or the physico-chemical separation measurements that we operate on radioactive material. The installation for the study of the metal-salt separation will also benefit form this basic equipment.

Assessment of the Chemical Salt Metal Separation

Thanks to the platform, it will be possible to measure the separation between metal and molten salts of the species Pa (from irradiation of Th), Th and U and to measure their respective concentrations in the cooled metal and salt, thanks to radioactivity measurements done at ISN. In this view, the tools that are commonly used at ENSEEG will have to be duplicated at the ISN site.

Later on, this assay facility can be used to study electrochemical mechanisms implying these three elements, in particular in a collaboration with the Transuranian Institute in Karlsruhe.

Immediate results

Elastic Scattering cross sections

The history and the slowing down of a neutron are determined by the elastic scattering cross sections of the various nuclei. This process necessarily induces a correlation between a neutron's instantaneous energy and the time elapsed since its creation.

Preceding studies, dealing with the moderation of neutrons in a lead block, have shown that the shape of the energy-time correlation curve allows the measurement of the variation of the elastic scattering cross section of the neutrons as a function of their energy with a sensitivity close to 5 %. Simulations based on the specific geometrical characteristics of the graphite moderator block provide the expected sensitivity for a 10 % variation of the elastic scattering cross section of neutrons in graphite. This is shown on the following distributions of neutron capture rate as a function of time, in a 0.1 mm thick gold target placed in the central measurement channel:

$\begin{figure}\begin{center}{\centering\includegraphics{comp_xsec.eps} \par } \end{center}\end{figure}$

We thus have a unique method to confront the results from a simulation of neutron transport in a given medium with experimental reality.

Capture Cross Sections

The validation of data bases relating either to the structure materials or to the fuels will have to be achieved by irradiating thin targets within the lead block, which will temporarily replace the 'molten salt' model, in order to take advantage of the better time and energy resolutions associated to slowing down the neutrons in lead.

Protactinium Extraction and Chemical Selectivity of the Separation Procedures

The ability of Thorium based molten salt reactors to operate as breeders depends heavily on the efficient extraction of ²³³Pa from the neutron flux zone, in order to avoid (n, $\gamma$ ) captures and let it decay to ²³³U away from the neutron flux.

The efficiency and selectivity of the extraction procedure can be measured, in collaboration with ENSEEG laboratories in the following way: a small amount of ThF₄ (a few grams) is mixed with the base components (LiF, BeF₂) with the molar proportions characteristic of the salt. The target thus made is irradiated in the graphite block, within the solid fluorides that fill the channels. This target is then assayed using gamma spectrometry before and after ²³³Pa extraction. The ratio between Thorium and Protactinium activities will give a direct measurement of the efficiency of the separation procedure.

Likewise, ²³³U retrieval can be measured by adding to the irradiated target a natural Uranium based tracer.

Long Term Program

The experiments described above can be extended on the platform in several ways, for a longer term program:

: The beam can be injected in a lead block (already available) in view of obtaining more accurate measurements of (n, $\gamma$ ) or (n,fission) events.
: Fuel processing and fission product transmutation tests can be performed in ways similar to those used for Pa extraction. According to the US research program published in the years 60-70, it is likely that Pa extraction will not be a major problem, the issue being only to find efficient technological solutions. On the contrary, this is not true of the separation of Th/rare earths (lanthanides). A large research effort will have to be dedicated to the evaluation of the problem: thermodynamic measurements, electrochemical measurements, find alloys that favor lanthanide extraction, phase diagrams, etc. Parallel to this, how well these reactors tolerate high lanthanide ratios will have to be evaluated, as well as the possibility of a partial rare earth extraction, e.g. valence 3 rare earths.
: A study of the neutronic and chemical behavior of the salt with temperature can be considered in the longer term, with a molten salt loop set up in one of the measurement channels.
: Various reactor configurations can also be studied, such as pseudo-fuel pellets (again with no fissile matter).

All these developments should allow us to validate and refine our exploration of possible scenarios concerning future nuclear electricity production. In the long term, computer simulations will have to handle fuel processing and its chemical aspects with as good a precision as the neutronics in the reactor core.

... CENBG ¹: Centre d'Etudes Nucléaires de Bordeaux Gradignan
... IPNO ²: Institut de Physique Nucléaire d'Orsay
... ENSEEG ³: Ecole Nationale Supérieure d'Electrochimie et Electrométallurgie de Grenoble
... LTPCM ⁴: Laboratoire de Thermodynamique et de Physico-Chimie Métallurgiques
... LEPMI ⁵: Laboratoire d'Electro-chimie et de Physico-chimie des Matériaux et des Interfaces
... EdF ⁶: Electricité de France

Last update: 12 June 2002

Previous: what_we_do