The development of novel, Radical Purex, reprocessing technologies, leadingto fission product waste streams with high levels of inert constituents, maymean that vitrification is no longer the optimum solution for theimmobilisation of highly active wastes at the back end of the nuclear fuelcycle. A ceramic phase assemblage is described which uses the inertconstituents of the waste stream as a functional component of the wasteform, permitting high waste loadings to be achieved and thus enabling wasteminimisation considerations to be satisfied. The initial development of thisphase assemblage is presented.

A kinetic model recently developed [1] for the radiolytically inducedoxidative dissolution of the spent fuel matrix is presented. This is basedon experimental studies on the generation and evolution of radiolyticproducts in a closed system containing fragments of PWR-fuel [2]. Theoutcome of this model is currently being integrated in the present PAexercise being prepared by SKB. The calibration of the model against variousexperimental information and its predictive capabilities for the long termperformance of the spent fuel matrix are presented.

Chlorine-36 has been identified as a potential source of radiological riskin the disposal of nuclear fuel waste. The radioisotope 36Cl (t1/2 = 3 × 1O5 a) is produced by neutronactivation of Cl impurities in UO2 fuel. The total average Climpurity level in four unirradiated CANDU UO2 fuel samples was2.3 ± 1.1 ppm. ORIGEN-S calculations using a 5 ppm Cl impurity in a CANDUfuel resulted in a 36Cl activity comparable to the activity of 129I and 14C produced in the fuel thus requiring 36Cl to be considered in disposal risk assessments. The“instant release” of 36Cl from the gap and grain boundary regionsof the fuel to solution was measured by leaching both clad fuel and fuelsamples crushed to grain-sized particles. The 36Cl concentrationwas measured by Accelerator Mass Spectrometry. The 36Cl releasesfrom fuel samples taken from 8 different fuel bundles ranged from 0.5% to20.4% of the total 3 Cl inventory over a leaching period of 32days. The 36Cl released was found to correlate with the stable Xegas release, the fuel burnup and the linear power rating (LPR). For atypical CANDU fuel with an LPR of -42 kW/m, the “instant release” of 36Cl would be about 5% of the total inventory.

A fuel leaching experiment has been in progress since 1977 to study thedissolution behaviour of used CANDU fuel in aerated aqueous solution. Theexperiment involves exposure of 50-mm clad segments of an outer element of aPickering fuel bundle (burnup 610 GJ/kg U; linear and peak power ratings 53and 58 kW/m, respectively), to deionized distilled water (DDH2O,∼2 mg/L carbonate) and tapwater (∼50 mg/L carbonate). In 1992, it wasobserved that the fuel in at least one of the leaching solutions showed somesigns of deterioration and, therefore, in 1993, parts of the fuel sampleswere sacrificed for a detailed analysis of the physical state of the fuel,using SEM and optical microscopy. Leaching results to date show that evenafter >6900 days only 5 to 7.7% of the total calculated inventory of 137Cs has leached out preferentially and that leach ratessuggest a development towards congruent dissolution. Total amounts of 137Cs and 90Sr leached are slightly larger intapwater than in DDH2O. SEM examinations of leached fuel surfacefragments indicate that the fuel surface exposed to DDH2O iscovered in a needle-like precipitate. The fuel surface exposed to tapwatershows evidence of leaching but no precipitate, likely because uranium iskept in solution by carbonate. Detailed optical and SEM microscopyexaminations on fuel cross sections suggest that grain-boundary dissolutionin DDH2O is not prevalent, and in tapwater appears to be limitedto the outer %0.5 mm (pellet/cladding) region of the fuel. Grain boundaryattack seems to be limited to microcracks at or near the surface of thefuel. It thus appears that grain-boundary attack occurs only near the fuelpellet surface and is prevalent only in the presence of carbonate insolution.

The alteration behavior of UO2 pellets following their reactionunder unsaturated drip-test conditions, at 90°C, for time periods of up to10 years has been examined by solid phase and leachate analyses. Samplereactions were characterized by preferential dissolution of grain boundariesbetween the original press-sintered UO2 granules comprising thesamples, development of a polygonal network of open channels along theintergrain boundaries, and spallation of surface granules that had undergonesevere grain boundary corrosion. The development of a dense mat ofalteration phases after two years of reaction trapped loose granules,resulting in reduced rates of particulate uranium release. The parageneticsequence of alteration phases that formed on the present samples was similarto that observed in surficial weathering zones of natural uraninite (UO2) deposits, with alkali and alkaline earth uranylsilicates representing the long-term solubility-limiting phases for uraniumin both systems.

The release fractions of the five elements in the ε-phase (99Tc, 97Mo, Ru, Rh, and Pd) as well as that of 238U arereported for the reaction of two oxide fuels (ATM-103 and ATM-106) inunsaturated tests under oxidizing conditions. The 99Tc releasefractions provide a lower limit for the magnitude of the spent fuelreaction. The 99Tc release fractions indicate that a surfacereaction might be the rate controlling mechanism for fuel reaction underunsaturated conditions and the oxidant is possibly H2O2, a product of alpha radiolysis of water.

The dissolution rate of unirradiated UO2 (s) has been studied asa function of hydrogen carbonate concentration at three differenttemperatures (298.15 K, 313.15 K and 333.15 K) under oxidizing conditions ina continuous flow-through reactor with a thin layer of solid particles(particle size from 100 to 300 μm). From the results of these experiments,two different rate laws have been determined. At high temperature (313.15 Kand 333.15 K), we obtained a dissolution rate proportional to hydrogencarbonate concentration while at 298.15 K, the rate almost depends on thesquare root of the hydrogen carbonate concentration. This indicates adifferent reaction mechanism depending on temperature which can be relatedto the oxidation step of the overall process. The apparent activation energyobtained was 41 kJ mo1−1.

Extensive Research is performed in many countries in order to evaluate thespent fuel behaviour under repository conditions. Several aspects as thecontrol of the oxidative spent fuel dissolution by secondary phasesformation are not yet clear.

Coprecipitation experiments from SIMFUEL solutions are performed to study ifminor elements will influence the formation of secondary phases. Therefore,coprecipitation studies from SIMFUEL solutions aims at identification ofstable phases of significant simulated fission products. These experimentsprovide upper limits for solution concentration and distribution ratios ofsimulate fission products at several pH values. SIMFUEL pellets, whichsimulate an irradiated fuel with burnup of 50 GWd/tU were provided by AECLResearch Laboratories, Canada. Experiments were carried out by addition ofan aliquot of the initial SIMFUEL solution to 5 m NaCI free of carbonatessolution. The selected pH was maintained constant during the experiments.The pH range considered was from 5.5 to 9.3. Analyses of the solutions wereperformed for uranium by Laser fluorescence and for the minor elements byICP-MS. Solid phases formed at pH 5.5 were dissolved and analysed by ICP-MS.Results of the evolution in solution vs. pH of simulated fission productsconcentrations are shown in this paper.

The solubility behavior of unirradiated UO2 pellets was studiedunder oxic, air-saturated conditions in deionized water, in NaHCO3 solutions and in two types of synthetic groundwater(25°C). The Allard groundwater represents natural fresh groundwaterconditions at great depths in granite bedrock and bentonite groundwatersimulates the effects of bentonite on granitic fresh groundwater. Therelease of uranium was measured during static batch dissolution experimentsof long duration (6 years). A comparison was made with the theoreticalsolubility data calculated with the geochemical code, EQ3/6, in order toevaluate solubility (steady state) limiting factors. Various hypotheses forredox control (redox potential of the bulk solution or redox potential atthe surface) were tested in the modeling calculations.

The measured concentrations for uranium at steady-state in deionized waterwere equal to the solubility of schoepite (PO2 = 0.2 atm). In NaHCO3 solutions with lower carbonate concentrations (0.98 –1.96 mmol/1) and in Allard groundwater they were close to the calculatedsolubilities of U at the U3O7/U3O8 redox potential. In bentonitegroundwater, the results suggest the formation of a secondary phase with alower solubility. Only uranium oxide with a crystal structure of uraninite (UO2 - U3O7) was identified in allwaters, when analyzing particulate material in the solutions after contactwith UO2 pellets.

The purpose of this work has been to measure and model the intrinsicdissolution rates of uranium oxides under a variety of well-controlledconditions that are relevant to a geologic repository. When exposed to airat elevated temperature, spent fuel may form the stable phase U3O8. Dehydrated schoepite, UO3H2O, has been shown to exist in drip tests onspent fuel.

Equivalent sets of U3O8 and UO3H2 dissolution experiments allowed a systematicexamination of the effects of temperature (25–75°C), pH (8–10) and carbonate (2–200×10−4 molar) concentrations at atmospheric oxygenconditions.

Results indicate that UO3H2O has a much higherdissolution rate (at least ten-fold) than U3O8 underthe same conditions. The intrinsic dissolution rate of unirradiated U3O8 is about twice that of UO2.Dissolution of both U3O8 and UO3.H2O shows a very high sensitivity to carbonateconcentration. Present results show a 25 to 50-fold increase inroom-temperature UO3H2O dissolution rates between thehighest and lowest carbonate concentrations.

As with the UO2 dissolution data the classical observed chemicalkinetic rate law was used to model the U3O8dissolution rate data. The pH did not have much effect on the models, inagreement with the earlier analysis of the UO2 and spent fueldissolution data,. However, carbonate concentration, not temperature, hadthe strongest effect on the U3O8 dissolution rate. The U3O8 dissolution activation energy was about 6000cal/mol, compared with 7300 and 8000 cal/mol for spent fuel and UO2 respectively.

The thermodynamic and kinetic dissolution properties of a syntheticallyobtained soddyite have been determined at different bicarbonateconcentrations. This uranium-silicate is expected to be a secondary solidphase of the oxidative alteration pathway of uranium dioxide in waters withlow phosphate content and, consequently, it is likely to constitute one ofthe long-term uranium solubility limiting solid phases.

The experimental data obtained at the end of the experiments correspondfairly well to the theoretical model calculated with a log K0S0 of 3.9±0.7.

On the other hand, the general trend of the total uranium in solutionmeasured in the experiments as a function of time has been fitted by using akinetic equation obtained from the principle of detailed balancing of thedissolution reaction. In addition, the EQ3/6 code has also been used tomodel the uranium concentrations as a function of time. In both modelingexercises comparable results were obtained. The dissolution rate, normalizedto the total surface area used in the experiments as measured with the BETmethod, gave an average value of 6.8 (±4.4) 10−14 mol cm−2 s−1.

Spent nuclear fuel will, by the radiation, split nearby water into oxidizingand reducing compounds. The reducing compounds are mostly hydrogen that willdiffuse away. The remaining oxidizing compounds can oxidize the uraniumoxide of the fuel and make it more soluble. The oxidised uranium willdissolve and diffuse away. The nuclides previously incorporated in the spentfuel matrix can then be released and also migrate away from the fuel.

A model is proposed where the produced oxidizing species compete forreaction with the fuel and for escaping out of the system. The chemicalreaction rate of oxygen and fuel is taken from literature values based onexperiments. The escape rate of oxidants to a receding redox front in thebackfill is modelled assuming a redox reaction of oxidizing component andreducing component in the surrounding. The rate of movement of the redoxfront is determined from the rate of production of oxidants. This isestimated using a previously devised model that has been calibrated to insitu observed radiolysis.

Three cases are modelled. In the first case it is assumed that the reducingcompound is insoluble and that the reaction between oxygen and reducingmineral is very fast. In the second case it is assumed that the reducingcomponent has a known solubility and that it can migrate to meet the oxygenand quickly react. In a third case a finite reaction rate is modelledbetween the oxygen and the reducing species.

The sample calculations show that if the reducing mineral has to be suppliedfrom the backfill a large fraction of the spent fuel could be oxidised. Ifthe corrosion products of a degraded steel canister can supply the reducingspecies and the redox reaction is fast, very small amounts of the fuel couldbe oxidised. Literature data indicate that the redox reaction rate may notbe so fast that it can be considered instantaneous and then a considerablefraction of the fuel could be oxidised. The model gives a means of exploringwhich mechanisms and data may be of most importance for radiolytic fueldissolution, but the realism of the data and the model must be testedfurther. There is a lack of understanding and data on reaction rates,heterogeneous as well as homogeneous. This is crucial to the results.

Spent nuclear fuel (SNF) is unstable under oxidizing conditions. Althoughrecent studies have determined the paragenetic sequence for uranium phasesthat result from the corrosion of SNF, there are only limited data on thepotential of alteration phases for the incorporation of transuraniumelements. The crystal chemical characteristics of transuranic elements (TUE)are to a certain extent similar to uranium; thus TUE incorporation into thesheets of uranyl oxide hydrate structures can be assessed by examination ofthe structural details of the β-U3O8 sheet type.

The sheets of uranyl polyhedra observed in the crystal structure of β-U3O8 also occur in the mineral billietite (Ba[(UO2)3O2(OH)3]2(H2O)4),where they alternate with α-U3O8 type sheets.Preliminary crystal structure determinations for the minerals ianthinite, ([U24+(HO2)4O6(HO)4(H2O)4](H2O)5),and “wyartite II” (mineral name not approved by IMA committee on mineralnames), {CaCo3}[U4+(UO2)2O3(OH)2](H2O)4,indicate that these phases also contain β-U3O8 typesheets. The β-U3O8sheet anion topology containstriangular, rhombic, and pentagonal sites in the proportions 2: 1:2. In allstructures containing β-U3O8 type sheets, thetriangular sites are vacant. The pentagonal sites are filled with U6+O2 forming pentagonal bipyramids. The rhombicdipyramids filling the rhombic sites contain U6+O2 inbillietite, U4+O2 in β-U3O8U4+(H2O)2 inianthinite, and U4+O3 in “wyartite-II” (in which oneapical anion is replaced by two O atoms forming a shared edge with acarbonate triangle of the interlayer). Interlayer species include: H2O (billietite, “wyartite II”, and ianthinite), Ba2+ (billietite) Ca2+ (”wyartite II”), and CO3−2 (”wyartite II”); there is no interlayer in β-U3O8. The similarity of known TUE coordinationpolyhedra with those of U suggests that the β-U3O8sheet will accommodate TUE substitution coupled with variations in apicalanion configuration and interlayer population providing the required chargebalance.

To solve the problem of silts and soils that are contaminated withradioactive and toxic substances, the following method has been developed atSIA RADON. The material is mixed with limestone and other components,including up to 70 % (mass) of dried residue of liquid radioactive waste.The mixture is heated at 800 to 1000 °C, shredded, and used to form cement.This cementation process may be used to treat radioactive or other chemicalwaste.

The work demonstrated that in the case of silts, for example, the productvolume is reduced by a factor of 1.5 to 3 compared to the initial siltsvolume. A fast hardening, durable product is obtained, the quality of whichis not inferior to that obtained by using the traditional binders. For someparameters, e.g., hardening rate and macroelement leaching, the currentmethod is significantly better than the traditional cementation process. Ithas also been demonstrated that, over a wide range of parameters, when dryresidue of liquid radioactive waste is used in the initial mixture, the partof NOxCan be reduced to nitrogen, thereby reducing NOxemissions.

The gamma-ray initiated oxidation of dicyclohexano-18-crown-6 (DCH-6) wasstudied in two-phase system under agitation: nitric acid, aqueous solutionand organic solvent.

The distribution coefficients of radionuclides, hydrodynamic characteristics(phase separation and interface surface tension), and the radiolysisproducts after irradiation at a dose rate (60Co) of 1.7 Gy/s weredetermined. The maximum total dose was 3.O×105 Gy (3.0×107R).

The possibilities for using various technologies for liquid radioactivewaste (LRW) management were considered. Technologies based on theapplication of sorbents highly selective to radionuclides were shown to havegood prospects. Possible ways of increasing selective sorbent efficiency incleaning radionuclide-polluted waters of various types were examined. Theresults of studying new high-selective sorbents produced in the Institute ofChemistry FEDRAS are given. The use of these sorbents enables one to treatLRW of various types in a single stage. The results of testing a sorptionfilter installation for management of LRW produced during operation, repairand disposal of nuclear reactors are presented. Use of this installationresolves a major portion of LRW disposal problems.

A reprocessing of uranium hexafluoride waste, which was accumulated in thenuclear industry for many years, is a very vital question now. There are alot of gas tanks filled with UF6 that have been exposed for along time and are posing a serious ecological danger. A plasma chemicaltechnique is proposed here to transform the uranium hexafluoride depletedwith isotope 235U (less than 0.1%) into the safe storage form -powdered uranium oxides. The second product is HF (70–80% aqueous solution)which may be easily reprocessed by means of the rectification column up topure anhydrous hydrogen fluoride.

The designed technological scheme and the basic technical characteristics ofthe plasma chemical converter are presented. The proposed technique isefficient, easy controlled and free of any harmful by-products.

A new repository waste package (WP) concept for spent nuclear fuel (SNF) isbeing investigated. The WP uses depleted uranium (DU) to improve performanceand reduce the uncertainties of geological disposal of SNF. The WP would beloaded with SNF. Void spaces would then be filled with DU (∼0.2 wt % 235U) dioxide (UO2) or DU silicate-glass beads.

Fission products and actinides can not escape the SNF UO2crystals until the UO2 dissolves or is transformed into otherchemical species. After WP failure, the DU fill material slows dissolutionby three mechanisms: (1) saturation of WP groundwater with DU andsuppression of SNF dissolution, (2) maintenance of chemically reducingconditions in the WP that minimize SNF solubility by sacrificial oxidationof DU from the +4 valence state, and (3) evolution of DU to lower-densityhydrated uranium silicates. The fill expansion minimizes water flow in thedegraded WP. The DU also isotopically exchanges with SNF uranium as the SNFdegrades to reduce long-term nuclear-criticality concerns.

We have studied a co-precipitation method to remove actinides fromradioactive liquid waste. Lanthanum phosphate was selected as aco-precipitation material because actinides and lanthanides are among thehomology series in the periodic table. In this report, the conditions, underwhich the lanthanum compound is precipitated was looked at, as well as, waysto widen the pH range in which the precipitation occurrs in sodium nitratesolution. The properties of the resulting precipitates were alsoinvestigated. Further, we investigated the incorporation ratio of amerícium(decontamination factor), one of the actinides, into the precipitate fromsodium nitrate solution. Lanthanum phosphate was found to be more effectivecompared to ferric compounds as the co-precipitation material to removeamerícium. The incorporation conditions for strontium using lanthanumphosphate were also investigated. The removal of strontium, by this method,was more effective when the pH is above 7.0.

In a simplified (SiO2, B2O3,Na2O, Al2O3, ZrO2) glass,corresponding to the basic matrix for the French nuclear waste containmentglass, the authors developed a molecular dynamics simulation of the atomdisplacement cascades resulting from a disintegration of the actinides.After simulating the secondary cascades resulting from the first atomsdisplaced by a collision with a recoil nucleus, an actinide (U) was added tothe model and the cascade produced by accelerating this type of atom wasexplicitly investigated, notably to observe the morphological evolution ofthe cascades and the resulting changes in the glass structure.