In a previous MRS paper, the consistency of migration parameters for strontium (Sr) in Boom Clay, obtained by different types of experiments, was examined. No consistent value could be obtained for the product ηR of the diffusion accessible porosity η and the retardation factor R. Furthermore the nearly flat concentration profile measured in one of the through diffusion experiments could not be explained by the traditional through diffusion model. A reason is that the filter plates confining the clay sample have not been taken into account.

Therefore, for Sr and tritiated water (HTO), the apparent diffusion coefficient and the product ηR in the filters are measured in through diffusion experiments on filter plates.

Taking into account the filter plates, the outlet fluxes and the Sr profiles in the clay of both Sr through diffusion experiments, are described well with (i) the previously estimated Sr apparent diffusion coefficient in the clay of 7 × 10-12 m2/s, (ii) an apparent filter diffusion coefficient in the range 2 × 10-12 m2/s to 5 × 10-11 m2/s (vs. 1 × 10-11 m2/s measured in the filter through diffusion experiments), (iii) a clay capacity factor ηR in the range between 5000 and 22000, and (iv) a filter capacity factor between 0.3 and 0.6 (in agreement with the filter through diffusion measurements). However, using the above parameters, the evolution at the inlet could not be described. So although inconsistency diminished, some inconsistency remains.

Yttrium aluminosilicate (YAS) glasses have been proposed as host matrices for the immobilization of radioactive elements. In addition, yttrium has been used to simulate actinides [1]. It is well known that these glasses are resistant to water corrosion and exhibit high Tg and good mechanical properties [2]. As shown in [3], on heating, yttrium disilicate and mullite / sillimanite crystals grow from the pre-existing nucleation sites on the surface, until each glass particle volume is fully crystallized (volume-homogeneous nucleation was not observed), decreasing the glassy surface available for sintering by viscous flow. Sintering takes place simultaneously, by viscous flow but competes with surface crystallization; thus, if thermal treatment is not carefully designed a vitroceramic is obtained. In this paper we study the isothermal sintering kinetics of a YAS glass-powder-size distribution and non-isothermal sintering kinetics at 1, 3, 5, 10 and 15 K/min of two YAS glass-powder-size distributions. From the experimental evidence obtained, and crystallization data from [3], we design a sintering procedure in order to achieve a high-density glass monolith with submicrometric crystalline phases.

Thorium phosphate diphosphate Th4(PO4)4P2O7 (β-TPD) was already proposed for plutonium and minor actinides immobilization. Synthesis of Th4-xPux(PO4)4P2O7 with x∼1.5 corresponding to roughly 26 wt % of Pu was achieved demonstrating the thermodynamic stability of this phosphate-based material even for high plutonium mole loadings. We established reliable inter-atomic potentials in the shell-model approach for the β-TPD in order to calculate the Frenkel defects formation energies as well as the threshold displacement energies. Furthermore we carried out a detailed analysis of the energetic pathways for the corresponding Frenkel defects annealing. We deduced that the diphosphate sub-lattices may be easily displaced while the annealing mechanisms are revealed to be complex requiring highly energetic pathways, thus constituting the main defects configurations that may lead to the amorphous state.

Nuclear graphite components are produced from polycrystalline artificial graphite manufactured from binder and filler coke material with approximately 20% porosity. During the operational lifetime of a nuclear reactor the graphite moderator is subjected to fast neutron irradiation which contributes to changes in material and physical properties such as thermal expansion co-efficient, young’s modulus and dimensional change. These changes are directly driven by irradiation induced changes to the crystal structure as reflected through the bulk microstructure. Therefore it is important that irradiation changes and there implications on component property changes are understood. Work carried out under the FP7 CARBOWASTE consortium under work package three is underway to characterize both structural and radiological damage in graphite. This study examines a range of irradiated graphite samples removed from the British Experimental Pile Zero (BEPO) reactor. Raman spectroscopy and Transmission Electron Microscopy (TEM) have been used to compare the effect of increased irradiation Fluence on graphite microstructure. Irradiation induced crystal defects and changes in crystallite size are observed using TEM and related to Raman Spectroscopy, comparisons are also made to virgin nuclear grade graphite.

The assessment of the main changes expected for spent nuclear fuel from its discharge to its deposition in a deep geological repository is of the outmost relevance to establish the initial conditions of the disposal. In this work, a literature review and a critical discussion of the main processes that will affect the structure and the inventory of the spent nuclear fuel during its interim dry storage is presented. Once the irradiation period is finished, the following changes are observed: i) the fuel pellet is fragmented due to the temperature gradient established during the irradiation stage. On average between 10-15 fragments are observed per pellet. ii) the initial gap existing between the pellet and the cladding decreases or disappears depending on the burnup. iii) a radial zonation is observed in the microstructure of the pellet. For burnup over 40MWd/KgU, the rim develops a porosity increase due to the high local burnup and the low temperature in the periphery. The rim also presents small bubbles of fission gases. This high burnup structure implies a degradation of the thermic conductivity in the pellet, that leads to a temperature increase in the center of the pellet with a subsequent migration of the fission gases and other impurities to the grain boundaries. The implications that all these changes may have on the spent fuel behaviour is presented and discussed.

After the closure of a high-level waste repository, corrosion of the carbon steel overpack will occur. The corrosion products can then migrate into bentonite and affect the migration behavior of radionuclides in bentonite. Therefore, electrochemical experiments, with Fe2+ supplied by anodic corrosion of carbon steel, were carried out to study trivalent lanthanides in compacted bentonite. The interface between a carbon steel coupon and bentonite (dry density, 1.5 Mg/m3) was spiked with a tracer solution containing Nd(NO3)3, Eu(NO3)3, Dy(NO3)3, and Er(NO3)3. The carbon steel coupon was connected as the working electrode to a potentiostat and held at a constant potential between -550 and 0 mV (vs. Ag/AgCl reference electrode) for 7 days. A model using dispersion and electromigration could explain the measured profiles in the bentonite specimens. The best-fit electromigration velocity was related to the applied electric potential and was 1.0–3.8 nm/s for Nd, Eu, Dy, and Er ions. For these lanthanides, the best-fit dispersion coefficient was also related to the applied potential and was 0.8–1.6 μm2/s, and the dispersion length was calculated as 0.2 mm from the linear relationship between the dispersion coefficient and electromigration velocity. Finally, the apparent diffusion coefficient for these lanthanides was estimated as 0.6–0.9 μm2/s.

Determining the redox conditions in the near field of deep underground radioactive waste disposal cells is a key question regarding the performance of metallic components (e.g. waste overpack), which may undergo drastic corrosion processes in oxic conditions. This oxic transient is supposed to be short due notably to the oxygen consumption by corrosion and pyrite oxidation. However, the observed precipitation of Fe(III)-minerals as well as localized corrosion patterns on steel coupons placed during 6 years in a borehole drilled in the Toarcian argillite of Tournemire (France) may suggest that in-situ oxic conditions lasted several years, which is not consistent with reactive transport simulations performed with usual hypotheses (perfect contact between materials, high pyrite accessibility, water saturated conditions). Multicomponent reactive transport simulations considering gas diffusion were performed with the code HYTEC and reproduce correctly the observations made in Tournemire while considering imperfect interfaces and resaturation processes. The model was then applied to a disposal cell for high-level waste (HLW) representative of the design developed in France, putting into evidence the possibility of a redox contrast between the front and back of a disposal cell in an argillaceous medium, as well as a duration of the oxic stage within the cell as long as the ventilation of handling drifts is maintained.

Self-disposal option for heat-generating radioactive waste (HLW, spent fuel, sealed radioactive sources) known also as rock melting concept was considered in the 70s as a viable but alternative disposal option by both DOE in the USA and Atomic Industry Ministry in the USSR. Self-disposal is currently reconsidered with a novel purpose – to penetrate into the very deep Earth’s layers beneath the Moho’s discontinuity and to explore Earth interior. Self-descending heat generating capsules can be used for disposal of dangerous radioactive wastes in extremely deep layers of the Earth preventing any release of radionuclides into the biosphere. Descending of capsules continues until enough heat is generated by radionuclides to provide partial melting of surrounding rock. Estimates show that extreme depths of several tens and up to hundred km can be reached by capsules which could never be achieved by other techniques.

In order to observe the structural change in the interior of irradiated fuel assembly, the non-destructive post irradiation examination technique using X-ray computer tomography (X-ray CT) was developed.

In this X-ray CT system, the 12 MeV X-ray pulses were used in synchronization with the switch-in of the detector in order to minimize the effects of the gamma ray emissions from the irradiated fuel assembly then clear cross section CT image of irradiated fuel assembly could be successfully obtained. Also, this non-destructive technique can be applied to observe the inner condition of the high radioactive materials such as a radioactive waste.

Corrosion tests of Zircaloy-4 were performed in a dilute NaOH solution (pH =12.5) at 303 K for 90 days using the gas flow system (oxygen; < 1 ppb) and a batch method (oxygen; < 0.1 ppm). The corrosion rate was determined by measuring gaseous hydrogen and the hydrogen absorbed into Zircaloy-4 assuming the following reaction:

where x represents the Zircaloy-4 hydrogen absorption ratio. The initial hydrogen content in the Zircaloy-4 specimen was controlled to be below 10 ppm. The corrosion rate decreased with time (90-day values: 2.46×10-3 and 2.37×10-3 μm/y for the gas flow method and 6.72×10-2 μm/y for the batch test). The Zircaloy-4 hydrogen absorption ratio during corrosion was over 90%. The large amount of hydrogen absorbed in Zircaloy-4 will play an important role in the long-term safety for the disposal of irradiated Zircaloy materials.

Coffinite (USiO4•nH2O) is a common mineral in uranium ores. It is often observed to replace uraninite during alteration under reducing conditions. However, it has proven difficult to synthesize coffinite in the laboratory and quantitative thermodynamic data for coffinite are lacking. Despite these experimental difficulties, there is ample evidence in nature that many uranium deposits have encountered conditions where formation of coffinite has been favoured over uraninite during postgenetic alteration events. Coffinite is also found as a primary mineral in sandstone uranium deposits. This review elucidates the spatial relation between coffinite and uraninite as seen on different scales with different analytical methods. Some further insight into the mechanism of uraninite alteration in natural, reducing, Si-rich environments is gained and some new arguments put forward, including the question of the effect of impurities and dopants, defects, and grain size. The replacement of uraninite by coffinite is discussed in terms of solid-fluid interaction.

Concepts for the disposal of high-level radioactive waste (HLW) and spent fuel (SF) in several countries include a massive steel overpack within a bentonite buffer. In past conservative safety assessments to demonstrate feasibility of geological disposal, overpacks are assumed to provide complete containment for a given lifetime, after which all fail simultaneously. After failure, they are ignored as physical barriers to radionuclide transport. In order to compare different repository designs for specific sites, however, a more realistic treatment of overpack failure and its subsequent behaviour is needed. In addition to arguing for much longer lifetimes before mechanical failure and a distribution of overpack failure times, such assessment indicates that the presence of the failed overpack greatly constrains radionuclide release from the waste matrix and subsequent migration through the engineered barrier system. It also emphasises the key role of the bentonite buffer and the need to be able to assure its performance over relevant timescales.

A bentonite buffer is a part of the engineering barrier system of the geological disposal concept for spent nuclear fuel in Finland. The chemical conditions in the bentonite porewater determine the solubility, speciation and diffusion/sorption behaviour of the radionuclides in the bentonite. The OH-groups on the edge sites of the montmorillonite, the main component of bentonite, can buffer the pH conditions especially in the case that the buffering capacity of the accessory minerals is small. In this work, the pH conditions created by the interaction of sodium montmorillonite and an acetate buffer solution were studied experimentally in batch experiments and in compacted sodium montmorillonite. In the experiment with the compacted sample, the ends of an 18.4 mm long cylinder were exposed to an external solution of 0.3 M NaCl and 0.1 M acetate adjusted to pH 5 with NaOH. The effects were monitored by measuring the pH in the montmorillonite sample at 5 mm and 9.2 mm from the solution-montmorillonite interface as a function of time. The batch experiment was modelled considering the surface reactions of montmorillonite and the dissolution reactions of calcite and the added constituents to explain the observed phenomena.

Layered hydrazinium titanate LHT-9, (N2H5)1/2Ti1.87O4 is a new nanohybrid material related to lepidocrocite-type titanates. Unique combination of ion exchange, reductive properties, surface activity due to Brønsted acid sites and occurrence of surface titanyl groups allows exploring LHT-9 for simultaneous uptake of almost all components of liquid nuclear wastes. LHT-9 irreversibly removes technetium, molybdenum, palladium and selenium from their aqueous solutions by specific mechanism of reductive adsorption. For removal of cesium, strontium, transition elements, actinides and lanthanides LHT-9 provides mechanisms of ion exchange and surface complexation. Products of adsorption are nanocrystalline and homogeneous powders loaded with 5 to 15 wt. % of radionuclides and non-radioactive elements. LHT-9 can be applied as ready-to-use precursor for one-step synthesis of durable titanate ceramic waste forms similar to SYNROC. An essential advantage of LHT-9 in comparison with other titanate sorbents (monosodium titanate and peroxo-titanate materials) is the absence of Na in its composition that permits arbitrary tailoring of sorbent properties by simple pre-treatment with the desired elements. Results on sorption of americium, cesium, strontium and lanthanides by LHT-9 are discussed.

Dedicated integrated experiments have been used to determine in situ the corrosion rate of low alloyed steels in geological repository mockups (90°C, anaerobic, 40 bars, slow synthetic water flux…). Two possible situations have been investigated: the steel can either be in direct contact with the host rock or in a solution rich in clay suspension. Electrochemical Impedance Spectroscopy (EIS), sampling gravimetry and microanalyses have been performed to determine the corrosion rate evolution with the exposure duration of the specimens. Kinetic regime modifications have been linked to structural modifications of the corrosion products interface.

In accordance with the Belgian “supercontainer design”, spent nuclear fuel (SNF) will be encapsulated in carbon steel canisters, surrounded by a concrete overpack for disposal in poorly-indurated clay. After re-saturation of the barriers by porewater, interactions with the concrete will result in solutions rich in NaOH, KOH and Ca(OH)2. Corrosion studies of SNF in ECW-type solution (Evolved Cement Water) and YCWCa-type solution (Young Cement Water with Ca) were performed under externally applied H2 overpressures over 426 days. Directly after H2 application, Tc concentrations decreased from >10-8 M to concentrations below detection limit. Based on the fractional release of selected fission products, low matrix dissolution rates of ~10-8/day were found in both experiments. U concentrations decreased finally to 1.5•10-9 M (YCWCa) and to 2.1•10-10 M (ECW), respectively. Am, Np and Pu concentrations were found throughout the experiments below their detection limits indicating an effective retention process.

This study aims at gaining a better understanding of the behaviour of montmorillonite in contact with different ground waters; alteration of montmorillonite and possible formation of secondary minerals. Batch experiments were conducted with purified Swy-2 montmorillonite in simulated fresh (I=0.05 M, pH 8) and saline (I=0.1 M, pH 11) waters at 25 and 60ºC in anaerobic (Ar(g)) conditions. The concentrations of Al, Fe; Mg and Si were analysed from ultra-filtered solution samples with HR-ICP-MS (High Resolution Inductively Coupled Plasma Mass Spectrometry). The amount of released Si depended strongly on the experimental conditions. The Si concentrations at 60oC in the saline and fresh waters showed a difference greater than an order of magnitude. The initial purified montmorillonite and the solid materials from experiments were analysed with XRD. The analysis indicated that the nature of smectite did not change, but the experimental conditions, more or less, modified the structure of montmorillonite, e.g., in fresh waters the XRD spectra showed peaks typical of mixed layer minerals, which can refer to the presence of either randomly ordered illite/smectite or randomly ordered collapsed smectite/ hydrated smectite layers. The dissolution of montmorillonite was studied also by modelling with TOUGHREACT. The experimental and modelled results were compared revealing a need to develop the model e.g. in respect of the evolution of pH.

A range of advanced imaging techniques have been brought together to provide a comprehensive picture of cement microstructure for nuclear waste immobilization. Image analysis of Nirex Reference Vault Backfill (NRVB) has been used to characterize the Calcium-Silicate-Hydrate (C-S-H) matrix fraction. Through weight loss measurements and digital image correlation of OPC-based cement blends we have quantified the development of microstructure surface strains during the initial 48 hrs hardening period. The build-up of displacements on the microstructure scale indicated grain-like compressive areas, surrounded by a network of tensile regions. Serial sectioning of NRVB using ultra-microtome cutting has been explored for advanced high-resolution 3D microstructure characterization, while X-ray Computed Tomography (XCT) has been used to obtain information of the 3-D pore space and size distribution of air pores in NRVB non-destructively.

In the framework of the successive 1991 and 2006 Waste Management Act, French government supported a very significant R&D program on partitioning and transmutation of minor actinides (MA). This program aims to study potential solutions for still minimizing the quantity and the hazardousness of final waste, by MA recycling. Indeed, MA recycling can reduce the heat load and the half-life of most of the waste to be buried to a couple of hundred years, overcoming the concerns of the public related to the long-life of the waste.

Within this framework, this paper aims to present the most recent progress obtained in CEA on the development of innovative actinide partitioning hydrometallurgical processes in support of their recycling, either in an homogeneous mode (MA are recycled at low concentration in all the standard reactor fuel) or in an heterogeneous mode (MA are recycled at higher concentration in specific targets, at the periphery of the reactor core). Recovery performances obtained on recent tests in high active conditions of the so-called GANEX and DIAMEX-SANEX process will be presented and discussed in light of the potential P&T scenarios. Finally, recent developments regarding the recycling of the sole Am will be presented as well as the results obtained on highly active solutions for this so-called EXAM process.

This set of results gives to the French government a portfolio of potential recycling processes which could be separately and progressively implemented if decided.

Higher burn-up (> 50 GWd/t) spent nuclear fuels (SNF) present problems for long-term management and disposal in mined repositories, principally because of their higher heat output. Here we present results from heat flow modeling of an alternative scheme for disposing of SNF - deep borehole disposal (DBD). We focus on how temperatures on the outer surface of the containers evolve, affect the melting and re-solidification of the high density support matrix (HDSM) and their consequences for the feasibility of this disposal concept. We conclude that not only is DBD a viable option for higher burn-up SNF, but it could be a practical disposal route for a range of combinations of SNF ages and number of fuel pins per container.