The results from the JSS phase II static glass leaching experiments are presented. The experiments have been performed in the presence of bentonite, granite and stainless steel corrosion products at 90°C for 28 days. The leaching behaviour of Cs has been analysed using gamma-spectrometry. The surface of the leached glass samples has been analysed by infrared reflection spectrometry.

The presence of uncompacted bentonite seems to have a very small, increasing, effect on the leach rate of the glass compared to when no bentonite is present. In the presence of compacted bentonite, a lower leach rate has beep founy than in the prgsencelof uncompacted bentonite. (For Cs 0.3 g × m−2 × d−1 and 0.7 g × m−2 × d−1 respectively.) The addition of stainless steel corrosion products seems to have a slightly increasing effect on the leach rate in the presence of compacted bentonite. The beneficial effect of granite on the leach rate seems to be overshadowed by the presence of uncompacted bentonite.

In nuclear waste repository design bentonite has been included as part of the backfill for its sorbtive capacity and low water permeability. Nevertheless, it cannot keep the waste form dry once intrusion of groundwater has occured [1]. Leach experiments [2], [3] with the radioactive nuclear waste form borosilicate glass JSS-A have been performed with and without bentonite present to provide a database which allows the long term stability of the glass in aqueous environment to be forecasted and the chemical effects of bentonite to be studied.

During the various processing activities in the production of plutonium for the U.S. national defense programs, waste by-products were generated. These wastes, which have been generated over about the past 40 years at sites such as Savannah River Plant in South Carolina and Hanford in Washington, have been stored in underground tanks. Plans for the final disposal of the Savannah River Plan waste have now been documented [1] and call for the decontaminated soluble fraction of the waste to be mixed with cementitious materials and disposed of in near-surface trenches [2]. The radionuclide fraction of the soluble waste and the insoluble fraction of the waste will be converted to glass and sent to a geologic repository. In order for the repository to accept the glass, certain performance criteria must be met. Although not fully defined, it is certain that these criteria will include some specification on the rate of release of nuclides from the glass waste form when contacted by water.

A corrosion test series was performed to clarify the role of reaction product layers on the corrosion of a simulated HLW borosilicate glass in a salt brine under hydrothermal conditions. The layers were unprotective at 200°C. At 120°C, slight protective effects ocurred when the leachant contained dissolved reaction products. The consequence for the long term behaviour between 120 and 200°C is a constant glass dissolution rate.

The French LWR reference glass was spiked successively with 237Np, 239Pu, 238Pu and 241Am to study the specific leaching behavior of each actinide, as part of a larqer program investigating the effects of a decay in this glass. This paper describes the preparation of the α doped glasses and presents the initial leaching results for neptunium, plutonium and americium.

Glass samples were leached for up to six months at 90°C under static conditions. Weight loss measurements were performed, and the leachate was analyzed by ICP for the inactive elements, and by a spectrometry for the actinides. The results provide a basis for comparing the retention capability of the glass for each actinide, and the alterability of the bulk glass.

Finally, a comparison between glass specimens doped with 239Pu and with 238Pu, having the same chemical composition but very different specific activities, allows the effects of radiolysis on leaching to be estimated.

Laboratory experiments have shown that groundwater conditions in a granite repository will be as reducing as those in a basalt repository. Chemical analysis of the reduced groundwaters confirmed that the Fe2+/Fe3+ couple controls the oxidation potential (Eh). The reducing groundwater conditions were found to decrease the time-dependent release of soluble elements (Li and B) from the waste glass. However, due to the lower solubility of multivalent elements released from the glass when the groundwaters are reducing, these elements have significantly lower concentrations in the leachates.

Gamma radiolysis reduced the oxidation potential of both granitic and basaltic groundwater in the absence of both waste glass and oxygen. This occurred in tests at atmospheric pressure where H2 could have escaped from the solution. The mechanism for this decrease in Eh is under investigation but appears related to the reactive amorphous precipitate in both groundwaters. The results of these tests suggest that radiolysis may not cause the groundwaters to become oxidizing in a crystalline repository when abundant Fe2+ species are present.

Lead-iron phosphate glasses have recently been proposed as a new primary disposal medium for both commercial nuclear reactor wastes and some types of high-level radioactive U.S. defense wastes [1–41. The initial work on the characterization of lead-iron phosphate nuclear waste glasses, hereafter named phosphate glasses, concentrated on the preparation, thermal properties, and devitrification behavior (both during and after preparation) of these glasses. Aqueous corrosion rates were measured at 90°C in distilled water, acidic and basic solutions, low eH distilled water, and in the “reference’ natural ground water appropriate to disposal in a tuffaceous formation (i.e., J-13 well water [5]). The results of these initial investigations were extremely encouraging and indicated that homogeneous, highly leach resistant (at 90° C) phosphate glasses loaded with either commercial or high-level U.S. defense wastes could be prepared at relatively low temperatures (process temp.:≤ 1050°C). The chemical stability of the phosphate glasses was found to be primarily due to the stabilizing effect of iron on the structure of lead phosphate glass. This effect is illustrated in Fig. 1 where the leachate conductivity at 90°C is plotted versus time for phosphate glasses with varying iron concentrations. The addition of 9 wt.% iron oxide to lead metaphosphate glass [Pb(PO3)2] increases the chemical durability of the glass in water by a factor of about 104. The tendency of pure lead phosphate glasses to crystallize is also greatly suppressed by the addition of iron oxide [3,4]. Lead metaphosphate glass can be completely crystallized by heating the glass in air at 30O°C for several hours. Lead-iron phosphate glasses, however, can be heated in air for several hundred hours at a temperature as high as 500°C without exhibiting any evidence of crystallization.

A preliminary assessment of the long-term durability of concrete in a repository sited in clay is presented. The assessment is based on recorded experience of concrete structures and both field and laboratory studies. It is also supported by results of the examination of a concrete sample which had been buried in clay for 43 years.

The enoineering lifetime of a 1 m thick reinforced concrete slab, with one face in contact with clay, and the way in which pH in the repository as a whole is likely to vary with time have both been estimated from available data. The estimates indicate that engineering lifetimes of about 103 years are expected (providing that sulphate resisting cement is used) and that pH is likely to remain above 10.5 for about 106 years.

Extensive studies have been made on the interactions between a waste glass and repository materials under static conditions. One of the PNC reference glasses was leached in the solution prepared from water in contact with crushed granite, tuff, diabase and backfill materials such as bentonite and zeolite. The leachant solutions except for some bentonite solutions reduced the glass leach rate compared with that measured in distilled water. The extent of the reduction was a function of silicon concentration in solution. The bentonite solutions enhanced the glass dissolution rate by a factor of 2 to 3 at low bentonite/water ratios but the effect was found to be less important at high bentonite/water ratios and in the long-term experiment. Addition of granite and zeolite to the bentonite solutions decreased the leach rate below the value measured in distilled water.

The crystallization behavior of a ferri-silicate α-waste glass was studied by means of several experimental techniques. The main crystal phase showed a plate-like morphology and was identified as a monoclinic pyroxene. The Johnson-Mehl-Avrami equation was used for the determination of the kinetic parameters of the process. Following both isothermal and non-isothermal techniques, the Avrami exponent, n, was found to be nearly one. The activation energy for crystal growth is Eg = 356 kJ mole−1. The crystallization process is governed by instantaneous nucleation and diffusion controlled two dimensional growth.

Laboratory experiments alone cannot be used to verify models for the long-term release of radionuclides from nuclear waste glasses. Basalt glasses have been proposed as an appropriate natural analogue for the long-term validation of release models [1]. Their analogous behavior has been demonstrated in laboratory experiments in which both types of glasses display similar reaction rates, alteration products and surface layer morphologies [2,3]. This paper illustrates how empirical data from natural occurrences of basalt glass can be interpreted by a model which is developed to describe the reaction progress of the corrosion of nuclear waste form borosilicate glass [4,5].

The reaction of basalt glasses with natural waters were proposed as analog systems to the long-term alteration of nuclear waste glasses in the geosphere [1].

High resolution electron microscopy, analytical electron microscopy and selected area electron diffraction are used to establish the crystallochemical mechanisms by which simulated high level waste enters Synroc. It is shown that waste species are incorporated largely as a non-continuous solid solution. In other words, simple isomorphic substitution of waste elements in place of matrix cations, i.e. the classical continuous solid solution, does not operate (or is only significant at very low waste levels of radwaste). Usually, substitution is accompanied by the appearance of coherent extended defects which can take the form of antiphase boundaries, unit cell twinning (both mimetic and polysynthetic), or cation ordering. In this manner, a numb~er of closely related phases appear in response to variations in waste composition and loading. It is believed that the mechanism of structural modification will endow Synroc with considerable flexibility to respond to chemical changes in the wastestream. Other ceramic materials currently being evaluated as possible wasteforms (viz. sphene glass-ceramics and APO4 phosphates), are also likely to possess this capacity.

High resolution electron microscopy and scanning electron microscopy were used to determine the distribution of intergranular films and microvoids in Synroc C. Diffraction contrast derived from these films, which were 1–3 nm thick, showed then to be ill-defined crystallographically, and they may be described as glassy. Pores were usually several micrometers in extent and occurred principally in rutile-rich areas. The chemical composition of these structural features was obtained using analytical transmission electron microscopy, secondary ion mass spectrometry, x-ray photoelectron spectroscopy and scanning Auger microscopy. Within intergranular films, elemental enhancement of cesium, sodium, potassium and aluminium, and possibly silicon and molybdenum was observed. Enhancement of cesium, sodium, phosphorous, aluminium and silicon was found at triple points. Fracture faces preferentially expose grain boundaries, and dissolution of these surfaces proceeds rapidly at ambient temperatures. During the fabrication of Synroc C, microvoids trap cesium vapour, and after cooling this condenses onto pore surfaces. Recognition of the (simulated) waste species which segregate at grain boundaries and pores, permitted the reinterpretation of published leach data for monolithic and crushed Synroc C.

The redox potential of the couple UO22+ /U4+ has been determined at 25°C in various HC1O4 media by direct potentiometric determinations in 0.25, 0.5, 1, 2 and 3 M HC1O4 media. Some spectrophotometric measurements of the equilibrium: Cu(s) + UO2+ = U4+ + Cu2+ + H2O.

The oxidation of CANDU fuel (UO2) by the alpha-radiolysis products of water has been investigated using electrochemical and X-ray photoelectron spectroscopic-experiments. Experiments with O2 and H2O2, two of the expected products of radiolysis of water, indicate that the rate of oxidation of UO2 by H2O2 is about 200 times faster than by dissolved oxygen. Oxidation by both H2O2 and O2 shows pH dependence. Possible reaction paths for the oxidation of UO2 by radiolysis products are discussed.

Samples of used UO2 fuel exposed to air-saturated water at 25°C for eight years have been examined using electron microscopy and X-ray diffraction techniques. The results, in conjunction with solution analysis uata, show evidence for a UO3.2H2O precipitate on the fuel surface, and confirm the importance of grain-boundary leaching in controlling Cs-137 release. Studies of useo-fuel dissolution under both oxidizing and reducing conditions at 150°C show significantly lower radionuclide release to solution for reducing conditions, illustrating the potential importance of reoox chemistry in determining the long-term stability of used fuel after disposal.

Leaching of high burnup BWR fuel for up to 3 years showed that both U and Pu attain saturation rapidly at pH 8.1, giving values of 1–2 mg/l and 1 μg/l respectively. The leaching rate for Sr-90 decreased from about 10−5/d to 10−7/d but was always higher than the rates for U, Pu, Cm, Ce, Eu and Ru. Congruent dissolution was only attained at pH values of about 4.

The hypothesis that alpha radiolytic decomposition of water is a driving force for UO2 corrosion even under reducing conditions has been examined in leaching tests on low burnup (low alpha dose-rate) fuel. No significant effect of alpha radiolysis under the experimental conditions was detected.

Results are presented for the dissolution of Turkey Point pressurized water reactor (PWR) spent fuel in J-13 well water at ambient hot cell temperatures. These results are compared with those previously obtained on Turkey Point fuel in deionized water, on H. B. Robinson PWR fuel in J-13 water, and by other workers using various fuels in dilute bicarbonate groundwaters. A model is presented that represents the conditions under which maximum dissolution of spent fuel could occur in a repository sited at Yucca Mountain, Nevada. Using an experimentally determined upper limit of 5 mg/l for uranium solubility in J-13 water, a fractional release rate of 6.4 × 10−8 per year is obtained by assuming that all water entering the repository carries away the maximum amount of uranium.

Sphene (CaTiSio5 ) is obtained as the main crystalline phase in Na2O-A12O3-CaO-TiO2-SiO2 glass-ceramics containing up to ∼ 15 wt% of simulated nuclear fuel recycle waste, by maintaining the precursor glass at ∼ 1050°C for sufficient time to allow nucleation and crystal growth. The implications for production of an engineering-scale waste form are discussed.