Abstract
To estimate the life-time of vitrified high level waste (HLW-glass) in geological disposal conditions in Boom Clay, the dissolution behaviour of waste glass has been studied in experiments in surface laboratories and in the HADES underground research facility of SCK•CEN since the 1980’s. The programme consists mainly of dissolution tests. The purpose of these tests is to understand the basic glass dissolution mechanisms, and to demonstrate realistic long-term dissolution rates. The main experimental variables are glass composition, environmental materials, temperature, and test duration. The studied glasses are the COGEMA glass R7T7, and the PAMELA glasses with SM539, SM527 and SM513 glass frit. The environmental materials comprise Boom Clay, metallic corrosion products and engineered barrier materials. Dissolution tests have been performed at temperatures from 40 to 190°C, for test durations from days to several years. The tests are performed with inactive glasses, which can be doped with radionuclides of interest. Because of the importance of silica sorption by the environmental materials, the dissolution test programme was extended with silica diffusion- and sorption tests in Boom Clay and FoCa clay. The interpretation of the experimental results is supported by geochemical and kinetic modeling. In the area of kinetic modeling, both analytical and Monte Carlo codes are applied.
The dissolution tests have demonstrated that, although the presence of Boom Clay initially increases the glass dissolution rate, the long-term dissolution rate decreases for diluted clay / clay water slurries. This decrease has not yet been demonstrated for the R7T7 glass in compact Boom Clay, but is expected to occur here also on the long term. The dissolution rate decreases faster after sufficient addition of glass powder to the medium. This was tested in experiments with the R7T7 glass at relatively high clay concentration (2000 g of humid Boom Clay per liter clay water, this is about half the solid/liquid ratio of compact Boom Clay), at 40 and 90°C. Linear interpolation of the long-term mass losses resulted in dissolution rates of ∼ 0.01 g.m−2.day−1. The statistical uncertainties on the dissolution test results did not allow to demonstrate smaller rates. The minimum statistically significant dissolution rate depends on the test conditions. Therefore, the present SCK•CEN programme includes dissolution tests at long-term near-field conditions (this is at 30°C, with compact Boom Clay and FoCa clay), which are considered more representee for the long-term situation. In view of the uncertainties on the experimental long-term dissolution rates and on the long-term dissolution mechanisms, rates smaller than 0.01 g.m−2.day1 (about 1 μ/year) should not be used as best estimate in the present performance assessment studies for disposal in Boom Clay. A constant dissolution rate of 0.01 g.m−2.day−1 would correspond to a dissolution time for a R7T7 glass package of approximately 150 000 years. The minimum dissolution time is of the order of 104 years.
Constraints on the Affinity Term for Modeling Long-Term Glass Dissolution Rates