The Grimsel Test Site, which has been in operation since 1984, is known in the radioactive waste management community for the scope and quality of the experiments and projects performed over the last two decades. This paper will overview Nagra's commitments and the current plans for future activities in association with Grimsel, which are being developed considering time scales of the order of a decade or more.

The extent of reaction of alloy 22 with limited amounts of aqueous calcium chloride (CaCl2) was investigated. Alloy 22 is a highly corrosion-resistant nickel-chromium-molybdenum-tungsten alloy. Specimens were polished to a mirror finish prior to aerosol salt deposition. An aqueous film was formed on the specimen surface by deliquescence of deposited CaCl2 at 150°C and 22.5% relative humidity (RH). The reactant gas was a continuous flow of purified humidified laboratory air. The reaction progress as a function of time was continuously measured in-situ by a microbalance. An initial weight gain due to CaCl2 deliquescence was observed. A steady weight loss was observed over the next 24 hours, after which no further weight change was apparent. During the period of weight loss, white precipitates formed and the specimen surface became visibly dry. Post-test analyses indicated that the precipitates consisted of calcium, chloride, and oxygen.

Uptake of Pb was investigated for different hydrated fresh and leached Portland and high-alumina cements, different CSH phases including tobermorite, portlandite, as well as for several relevant cement Al-minerals. Except for highly degraded cement, CSH minerals are the most important phases for Pb uptake. A systematic analysis of our data and a comparison with recent literature data for CSH phases [1] shows that Pb uptake is characterized by three trends: 1) between pH 10–13, Pb uptake decreases with increasing pH, 2) Pb uptake increases with increasing Fe content in the clinker, and 3) Pb uptake is dependent on the Pb concentration in the solution. Following these findings, an empirical model was developed based on regression analyses of uptake vs. Pb concentration, pH, and Fe content in the clinker. The resulting relation can predict Kd values for a large range of fresh and leached cements, as well as for individual cement minerals under different experimental conditions. Model application to independent data on cement phases, whole cement, and mortar gave very good agreements.

Performance assessment calculations for low-activity waste glass to be disposed at the Hanford site depend on simulations of long-term glass corrosion behavior and contaminant transport that are being performed via reactive chemical transport modeling. Confidence in the underlying physical and chemical processes that are being approximated by the computer model could be significantly enhanced through carefully-controlled field testing, which includes studies of buried ancient glasses. Field tests with simulated low-activity waste glasses have been initiated on the Hanford site and at the Ballidon site in the United Kingdom. In addition, a joint PNNL – SIA RADON research project has been initiated to analyze a unique data set collected over 12+ years during a Russian in-situ testing program with actual low-activity waste glass. The glasses buried at Hanford are scaled down cylinders (45 kg mass) and include a representative glass composition for Hanford low-activity waste and a glass designed to be highly reactive. Each glass was doped with chemical analog tracers (Re, Se, and Mo). Design of the lysimeter test facility and sampling locations were done with aid of predictive calculations performed with the STORM code. These same glasses were also buried at Ballidon but in the form of small glass coupons. The coupons will be retrieved over a period of several years for detailed analysis along with core samples of the surrounding limestone, which will be analyzed for contaminant transport profiles. Analysis of porewater from underneath the Russian burial site ha s been compared with modeling calculations. Good agreement between the model and the field data has been obtained so far using estimated parameters for the glass corrosion. Independent laboratory tests are in progress to parameterize the STORM model for quantitative comparisons.

Transmission electron microscopy results from a sintered ceramics with stoichiometry of Ca(U0.5Ce0.25Hf0.25)Ti2O7 show the material contains both pyrochlore and zirconolite phases and structural intergrowth of zirconolite lamellae within pyrochlore. (001) plane of zirconolite is parallel to (111) plane of pyrochlore because of their structural similarities. The pyrochlore is relatively rich in U, Ce, and Ca with respect to the coexisting zirconolite. Average compositions for the coexisting pyrochlore and zirconolite at 1350 °C are Ca1.01(Ce3+0.13Ce4+0.19U0.52Hf0.18)(Ti1.95Hf0.05)O7 (with U/(U+Hf) = 0.72) and (Ca0.91Ce0.09)(Ce3+0.08U0.26Hf0.66Ti0.01)Ti2.00O7 (with U/(U+Hf) = 0.28) respectively. A single pyrochlore (Ca(U,Hf)Ti2O7) phase may be synthesized at 1350 °C if the ratio of U/(U+Hf) is greater than 0.72, and a single zirconolite (Ca(Hf,U)Ti2O7) phase may be synthesized at 1350 °C if the ratio of U/(U+Hf) is less than 0.28. An amorphous leached layer that is rich in Ti and Hf forms on the surface after the ceramics has been leached in pH 4 buffered solution. The thickness of the layer ranges from 5 nm to 15 nm. The leached layer functions as a protective layer and therefore reduces the leaching rate.

Alloy 22 (N06022) is the candidate material for the corrosion resistant, outer barrier of the nuclear waste container. Two of the potential corrosion degradation modes of the container are uniform corrosion and localized corrosion. A testing program is under way at the Lawrence Livermore National Laboratory to determine the susceptibility of Alloy 22 to these two forms of corrosion using immersion tests. Metallic coupons are being exposed to several electrolyte solutions simulating concentrated underground water from pH 3 to 10 at 60°C and 90°C. This paper describes the results obtained after more than a five-year exposure of 122 specimens to the testing electrolyte solutions. Results show little general corrosion and the absence of localized corrosion. The maximum general corrosion rate was 23 nm/yr.

A preliminary study of the sorption of reduced plutonium (III-IV) onto TiO2 has been done under anoxic conditions at varying pH and ionic strength. The first approach was to study tetravalent Pu, but the sorption decreases with time in this case. This is probably due to disproportionation of Pu(IV) at the TiO2 surface, which then acts as a catalyst. A further observation that supports this explanation is that the sorption of Pu(IV) onto the walls of the experimental equipment is in some cases higher than the sorption of (disproportioned) Pu onto the oxide.

Pu(III) sorption onto the oxide was also investigated and showed similarities with the previously investigated sorption of Pm(III) in a similar system.

Two disparate reaction mechanisms have been invoked to explain the reactivity of glass in contact with aqueous solution. One model is based on arguments from Transition State Theory (TST), which postulates that glass dissolution rates are surface reaction controlled. Alternatively, the second model argues that release of elements from glass to solution is governed by diffusion through an altered layer that forms on the surface of glass. Vertical Scanning Interferometry (VSI) is a new technique that allows one to observe surface features of specimens exposed to solution and may, potentially, be used to distinguish between competing models. We performed a series of dissolution experiments with a suite of glass compositions from chemically simple (sodium borosilicate) to complex (sixteen component borosilicate). Dissolution rates were determined using single-pass flow-through (SPFT) apparatus at 90°C and pH = 9 and over a solution saturation interval. Upon termination of the experiments, glass coupons were examined by VSI techniques. Effluent chemistry and VSI measurements indicate a nearly constant rate of 2.2 to 3.4 g m-2 d-1 over the solution interval; rates calculated from both methods are identical within experimental uncertainty. We argue that this glass is phase separated, and propose a model for dissolution based on the relative rates of dissolution of the two glass compositions. The data are consistent with a modified version of TST and indicate the potency of VSI methods to elucidate glass reaction mechanisms.

The removal and immobilization of mercury(II) ions from industrial waste streams is a difficult and expensive problem requiring a robust extractant that is resistant to corrosive conditions. We have now developed an acid-resistant thiacrown polymer that has potential utility as a selective and cost-effective Hg2+ extractant. A new crown thioether, 2-hydroxymethyl-1,4,8,11,14-pentathiacycloheptadecane ([17] aneS5-OH), was synthesized through reaction of 2,3-dimercapto-1-propanol with 4,7,10-trithiatridecane-1,13-di-p-toluenesulfonate; then treated with thionyl chloride to form 2-chloromethyl-1,4,8,11,14-pentathiacycloheptadecane ([17] aneS5-Cl); followed by conversion to 2-(N-methyl)aminomethyl-1,4,8,11,14-pentathiacycloheptadecane ([17] aneS5-NHMe), through reaction with methylamine. The synthesis of the 4-vinylbenzyl-substituted thiacrown was readily accomplished by treating with 4-vinylbenzyl chloride. Co-polymerization of the 4-vinylbenzyl-substituted thiacrown with DVB (80% divinylbenzene) using AIBN as the radical initiator generated the highly cross-linked crown thioether polymer. The polymer based material was tested for Hg(II) extraction activity from acidic solutions. The material was found to extract 95+% of the Hg in a pH range of around 1.5 to over 6.0, with contact times of less than 30 min and in the presence of high concentration of competing ions such as Pb, Cd, Al, and Fe. The spent polymer could also be stripped of the Hg2+ and reused without sufficient loss of loading capacity. Characterization of the binding sites by XPS and XAS show only some of the S atoms in the thiacrown participate in the binding of Hg. Our results indicate this polymer has the potential to treat highly acidic wastes efficiently. This thiacrown polymer is far more effective in extracting Hg2+ from aqueous solution than previously reported Hg2+ extrac-tants, not only being very fast and efficient and selective, but also independent of pH.

The heat of hydration of a number of composite cement systems has been studied using isothermal conduction calorimetry (ICC) from ambient temperature to 90°C. The resulting hardened cement paste, from the high temperature regime, was then examined by scanning electron microscopy. Results showed that increasing the hydration temperature increased the rate of heat output for all systems and, at early ages, decreased the porosity of the sample.

For safety evaluation of deep repositories, the evolution of chemical containment properties of clayey barriers in spent fuel disposal tunnels are assessed using reactive transport modelling. The disturbances related to cement components are more particularly studied for relevant time scales (100,000 years) and dimensions. Theoretical distribution coefficients (Kd) and maximum concentrations are derived for Cs, Tc and U and their sensitivity to the system evolution estimated. Mineralogical transformations and ion sorption are shown to be interdependent mechanisms controlling the intensity and spatial expansion of the alkaline plume. Simulations for a normal diffusive scenario and an altered one involving an advective flow lead to limited perturbations of the mineralogy and containment properties of the multi-barriers system, but emphasize the possibility of a migration pathway through the excavation damaged zone.

Magnesium oxide (MgO) is the only engineered barrier being emplaced in the Waste Isolation Pilot Plant (WIPP), a U.S. Department of Energy repository for transuranic waste. MgO will lower dissolved concentrations of actinides by consuming CO2 from possible microbial activity, by buffering the pH between 8.5 and 9.5, and by reducing the amount of free water in the repository. This paper discusses results from experiments measuring the hydration of MgO. Results suggest that periclase (MgO) hydrates rapidly to brucite (Mg(OH)2) in de-ionized water and 4 M NaCl solution at 90°C; the hydration rate decreases as temperature decreases. In ERDA-6, a NaCl-rich WIPP brine, MgO hydrates directly to brucite; in GWB, a high-Mg brine, periclase hydrates to magnesium chloride hydroxide hydrate(s) until the dissolved Mg2+ concentration decreases, and brucite becomes the stable phase. Under humid conditions MgO fully hydrates at higher humidities(> 50%). All data are consistent with diffusion-limited hydration reactions.

A new class of getter materials has been synthesized for immobilization of long-lived radionuclides such as 129I. These novel materials consist of nanoporous ceramic substrates with tailored pore sizes ranging from 2 – 20 nm. These high surface area (∼1000 m2/g) ceramic substrates have been functionalized with self-assembled monolayers consisting of soft cation-capped thiol-functionality. The resulting getter materials exhibit highly dense binding sites, and excellent selectivity for iodide. The effectiveness of these novel getter materials was evaluated using radioiodide-spiked samples of surface water and concrete leachate and adsorption performance was compared with natural sulfide mineral getter materials. The data indicated that the novel getter materials have very high affinity for radioiodide (Kd: 4 × 104 – 3 × 105ml/g and 6 × 105 ml/g in surface and concrete leachate respectively). Comparatively, the radioiodide Kd values for natural mineral getters were typically two to three orders magnitude less than the novel getters. The results indicated that the synthetic getter materials have the potential to immobilize and therefore retard the migration of 129I in the subsurface environment. Additional studies are being conducted to evaluate thelong-term stability of these materials in waste disposal environments.

Attention has recently been given to the immobilization of special categories of radioactive wastes, some of which contain high concentrations of actinide chlorides. Although vitrification in phosphate glass has been proposed, this was rejected because of the high losses of chloride through the mobilization of volatile species.

On the basis of non-radioactive and, more recently, radioactive studies, we have shown that calcium phosphate is an effective host for immobilizing the chloride constituents [1]. In this instance, the chlorine is retained as chloride, rather than evolved as a chlorine-bearing gas. The immobilized product is in the form of a free-flowing, non-hygroscopic powder, in which the chlorides are chemically combined within the synthetic mineral phases chlorapatite [Ca5(PO4)3Cl] and spodiosite [Ca2(PO4)Cl]. Data from studies on non-radioactive simulated waste consisting of a mixture of CaCl2 and SmCl3, and radioactive simulated waste composed of CaCl2 with PuCl3 or PuCl3 and AmCl3, are presented and compared.

The XRD data confirm the presence of chlorapatite and spodiosite in the non-radioactive and radioactive materials. The durability of all specimens was measured with a modified MCC-1 test. Normalized releases of Cl after 28 days were 1.6 × 10-3 g m-2 for the non-radioactive specimens and 7 × 10-3 g m-2 for the Pu bearing specimens. Releases of Ca after 28 days were 0.3 × 10-3 and 2.0 × 10-3 g m-2 for the non-radioactive composition and the Pu composition, respectively, whilst release of Pu from the radioactive specimens was higher for the mixed Pu/Am specimen at 1.2 × 10-5g m-2 than for the Pu only specimens. The release of Am from the mixed Pu/Am composition was exceptionally low at 2.4 × 10-7 g m-2. Overall, the release rate data suggest that the ceramics dissolve congruently, followed by precipitation of Sm, Pu and Am as less soluble phases, possibly oxides or phosphates. The differences in behaviour noted between non-radioactive and radioactive specimens are interpreted in terms of the crystal chemistry of the individual systems.

Closure of tanks containing high-level radioactive waste (HLW) is a challenging problem involving potentially competing influences from economic, societal, and technological considerations. The U.S. Department of Energy (DOE) is faced with protecting public health and the environment while making economically responsible decisions. Risk (i.e., annual dose) is becoming more prominent as DOE's metric to evaluate the economic consequences of its decisions. Risks are assessed through modeling and calculations commonly known as performance assessment (PA). In the process of tank closure, the U.S. Nuclear Regulatory Commission (NRC) is typically consulted to perform an independent review of DOE's PAs.

The NRC staff developed a generic PA model, applicable to HLW tank closure, which NRC utilizes to complete its independent review. The model was developed using the generic simulation software, GoldSim, because of its probabilistic capabilities and its adaptability to different problems [1]. The NRC staff uses the resultant risk from the generic models to evaluate the reasonableness of performance assessment models submitted by DOE. Large differences in the estimates of risk between the generic PA model and the DOE PA would likely indicate a need for stronger technical basis for processes significantly contributing to annual dose (risk) reduction.

A metallic waste form (MWF) will be used to immobilize cladding hulls from spent sodium-bonded nuclear fuel rods. Laboratory tests were conducted to determine if the HLW glass degradation rate model used in Total System Performance Assessment (TSPA) calculations can be used as an upper bound for the degradation rate of the MWF. Static tests were conducted with monolithic MWF samples in pH buffer solutions to measure the degradation rates at 50, 70, and 90°C. The measured degradation rates of the MWF were compared to the rates calculated with the HLW glass degradation model. The degradation rates were calculated as the product of the specific dissolution rates and the anticipated exposed surface areas of glass and MWF in a breeched canister. This comparison showed that the rate calculated using the HLW glass degradation model bounds the rates measured for the MWF at all pH values and temperatures. These results support the potential use of HLW glass as a surrogate for MWF degradation for the purpose of performance assessment.

The system C – Al – TiO2 has been demonstrated to be a strong candidate for the processing of irradiated reactor graphite waste with the retention of biologic hazardous carbon-14 in chemically and thermal stable corundum-carbide ceramics. The corundum-carbide ceramics is obtained from the powdered precursors blend through self-sustaining thermochemical reactions. Investigations of the system C – Al – TiO2 were carried out both theoretically and experimentally. The refining thermodynamic calculations of the phase composition of resulting end product were performed for a wide variety of components content in the system being investigated. Aluminium oxycarbides production was taken into account in the calculations. Thermodynamic functions of aluminium oxycarbides Al4O4C and Al2OC have been calculated for this purpose using currently available literature evidences and own assessments of missing data. On the basis of thermodynamic simulation the proportions of the source substances were determined, which result in the aluminium oxycarbides production. These simulation results have been supported by XRD-analysis of produced specimens. The experimental processing of reactor graphite was conducted by the use of self-sustaining reactions in C – Al – TiO2 powder blends. Test specimens were produced by mass ranging from 0.1 to 3 kg in the argon atmosphere. Various techniques were used to characterize the produced specimens. The compressive strength of specimens of corundum-carbide matrices produced ranges from 7 to 13 MPa. The leaching rates of Cs-137 and Sr-90 from specimens ranged between 10-4 and 10-5 g/(cm2.day) respectively. The carry-over of the carbon combined in carbon monoxide from the reacting mixtures during exothermic process may run up to 1% wt. that appropriates roughly to less than 0.01% wt. of the carbon-14 in the irradiated reactor graphite.

This paper compares the results of three different fuel corrosion experiments by taking the existing corrosion data reported for various temperatures and recalculating corrosion rates at a single temperature of 25°C using a temperature dependent model developed elsewhere. Three types of light water reactor fuel corrosion tests (sometimes called dissolution or alteration tests) were performed in support of Yucca Mountain Project. The tests used three water contact modes and various fuel burnups. All measurements were adjusted for temperature differences and then compared. Five different isotopes (cesium, technetium, iodine, strontium, and, in the flow-through tests, uranium) were considered as a measure of corrosion. The data used represent over ten years of experiments with about nine different fuel types. Most experiments were with repository type fluids, containing dissolved constituents such as carbonate, calcium and silicon. The results show that all of the experiments predict similar fuel corrosion rates. Small differences in the isotope release rates are observed and incorporated in the abstracted uncertainty. Water contact mode (flow-through, batch, or drip) does not seem to be very important although the drip tests introduced larger variations. In developing a corrosion abstraction, all of the isotope measurements were considered equally. The distribution of release rates was used directly to develop the uncertainty. The mean corrosion rate was 1.8 × 10-4 fraction/year at 25°C (5%-95% range = 5.7 × 10-5 to 1.7 × 10-3). Using the derived corrosion rate for 25°C, and considering rapid axial splitting of the cladding, the CSNF fuel rod is expected to corrode in less than 2,000 years. The abstraction uses all the available experiments performed with water containing carbonates, silicon, or calcium and irradiated fuel to produce a corrosion rate distribution. Sensitivity studies using this corrosion rate abstraction in the TSPA-SR analysis show very small changes in dose (3%) in response to changes in the UO2 corrosion abstraction.

The UO2 in spent nuclear fuel is unstable in the oxidizing conditions within the volcanic tuffs at the proposed nuclear repository at Yucca Mountain, Nevada. Over time, the UO2 will oxidize and corrode, releasing actinides and fission products to the surrounding environment. However, uranyl (U6+) phosphates (autunite, phurcalite, sodium autunite, etc.) are stable in such an oxidizing environment. The mobility of released radionuclides may be greatly retarded if they can be incorporated into these naturally stable phosphate phases, while the complex structures, variable chemical compositions, and natural analogue occurrences of the uranyl phosphates suggests such a process is favorable. Current tests have focused on synthesizing such phases by reacting uranium oxynitrate or UO3 with a calcium, sodium, or potassium phosphate and a base (if necessary) in a Teflon reaction vessel. Excess water is added, and the solution is heated at 90°C for 7, 35, or 182 days. SEM analyses have confirmed that various uranium phosphate crystalline solids have formed. XRD results indicate that tests using two different calcium phosphorus source materials, Ca2P2O7 and Ca10(OH)2(PO4)6, have both created synthetic phosphuranylite, Ca(UO2)[(UO2)3(OH)2(PO4)2] 2*12H2O. The formation of this phase is appears to be kinetically favored over other similar phases. Results utilizing sodium phosphates, NaH2(PO4)*H2O and Na2HPO4, have produced sodium autunite (NaUO2PO4*2H2O), but other phases are probably also present. Test results utilizing potassium phosphate, K3PO4, were inconclusive. Experiments using surrogate radionuclides are currently being performed in order to determine whether radionuclides, such as 239Pu, 137Cs, and 99Tc, released from corroded spent nuclear fuel can become incorporated into the crystalline structure of specific uranium phosphate phases, effectively limiting any further migration.

Alloy 22 (UNS N06022) was selected to fabricate the corrosion resistant outer barrier of a two-layer nuclear waste package container. This paper reviews the main corrosion degradation modes that are predicted for the outer layer of the container. Current results show that the containers would perform well under general corrosion, localized corrosion and environmentally assisted cracking (EAC). For example, the general corrosion rate is expected to be below 100 nm/year and the container is predicted to be outside the range of potential for localized corrosion and environmentally assisted cracking.