Long term bentonite erosion experiments: Radionuclide diffusion in FEBEX bentonite

INTRODUCTION

Natural bentonite is envisaged to be used as engineered barrier for repositories of high-level radioactive waste in, e.g., crystalline rocks [1]. Steel canisters with or without copper coating containing the spent fuels rods are planned to be emplaced in a host rock with bentonite surrounding the canisters as a backfill material. In some disposal concepts (e.g. the Scandinavian concept), glacial melt water intrusion scenarios are discussed. The intrusion of low mineralized water would lead to the swelling and the erosion of the bentonite [2], with consecutive formation of bentonite colloids. These would in term act as carrier for radionuclides, increasing their mobility, in such cases, where the container corrodes and loses integrity. The present study investigates radionuclide diffusion through the bentonite, bentonite erosion and the colloid-associated radionuclide migration.


DESCRIPTION OF THE WORK

FEBEX bentonite erosion experiments have been conducted in an artificial horizontal fracture set-up (mock-up experiments) to simulate the intrusion of groundwater in a repository. A ring of compacted bentonite (40 mm inner diameter, 80 mm outer diameter, 25 mm height) was emplaced between two Plexiglas plates spaced by a 1 mm height aperture to simulate a parallel fracture around the bentonite. ... mehrGroundwater from the Grimsel Test Site, pumped through the fracture at a flow rate of 50 µL/min, led to bentonite swelling and formation of a gel in the aperture [3]. This inactive experiment was carried out under atmospheric conditions.

A similar experiment was performed with Zn-labelled montmorillonite, a cocktail of radionuclide tracers (containing 241Am(III), 137Cs(I), 242Pu(III), 45Ca(II), 75Se(IV), 99Tc(VII), 233U(VI), 237Np(V)) and a conservative tracer, Amino-G, that was mixed with a bentonite slurry and inserted in open glass vials within the bentonite. This active experiment was carried out in an argon glove box. In the collected effluent, analysis of colloids, elemental composition and radionuclide breakthrough were performed [3].
The Plexiglas boxes containing the radioactive and the inactive mock-up test were opened after ca. 5 years from the end of the experiments for post mortem analysis.


RESULTS AND DISCUSSION

In the inactive experiment, coarse grains of accessory minerals of various colors were visible in the bentonite gel and sampled for SEM-EDX and XRD analysis. Some minerals that were originally present in the bentonite (i.e. gypsum) were not identified, suggesting the dissolution of these minerals ̶ in agreement with enhanced SO4 concentrations in the collected water samples, while some other secondary phases were formed in the ring.
In the active experiment, 49 ± 2% of the Amino G tracer was released from the bentonite source. Radionuclide release was detected in case of 99Tc, and 237Np. In the post mortem analysis of the active experiment aiming at the investigation of other radiotracers, the bentonite ring and gel were preserved as two samples of the same size that adhere onto the two Plexiglas plates (top and bottom plates of the Plexiglas box as seen in Fig. 1). Two of the vials were intact and the two others broken due to the pressure rise during the experiment. Autoradiography was performed on the ring sample adhering to the top plate to detect gamma-emitter radionuclide tracers (namely 241Am and 137Cs). Two different types of diffusion distances were identified as shown in Fig. 1. For the intact vial, the radionuclides were still mostly contained in the vial, with diffusion at a millimeter scale around the vial. In the case of the broken vial, the radionuclides were diffusing further away from the source on a centimeter scale. Autoradiographs of gamma emitting radionuclides could not be taken to represent diffusion profiles due to the relatively long range of gamma radiation. Such picture of the gamma-emitters distribution in the proximity of the vials will, however, allow planning the forthcoming steps of sample preparation and analysis by other analytical methods. In particular, in order to investigate the diffusion profiles of the actinide tracers at different distances from the vials, the bentonite sample will be cut into ca. 100 μm slices by abrasive peeling. After desorption of the actinide tracers with concentrated nitric acid, their concentration will be determined with (SF)ICP-MS.

Figure

Fig. 1: Post mortem analysis of the FEBEX erosion experiment containing the radioactive tracer cocktail. Autoradiography images of the bentonite ring on the right.



REFERENCES

[1] Shelton et al., SKB Technical report TR-17-17, (2018).
[2] Bouby et al., Applied Clay Science, 198, 105797, (2020).
[3] Rinderknecht, Doctoral thesis, KIT, (2017).