Preliminary assessment of the controlled release of radionuclides from waste packages containing borosilicate waste glass

Abstract

PNL-7591 UC-802 previously reached a similar conclusion (NAS 1983), as have other national repository programs (K8S 1983; NAGRA 1985). Data from the WAPS test is largely irrelevant, therefore, to the release-rate performance at either the waste-packages or EBS level of analysis. Radionuclide solubility limits are key to establishing predictive reliability and safety for geologic disposal; obtaining such data should be one priority for future tests that are intended to support the performance assessment of defense high-level waste glass. Transport parameters, such as extremely low diffusion coefficients in partially saturated tuff, may dominate the release of radionuclides from waste packages, even for highly soluble radionuclides. WASTE GLASS ANALYSIS Glass is a metastable solid and will, in the presence of water or water vapor, alter to a more stable set of phases. These phases will impose solubility limits to the concentration of many radionuclides within the waste package. Mechanistically based models for predicting the long-term glass dissolution and release of soluble radionuclides as a function of environmental variables exist; empirical models cannot be defensibly extrapolated to the long periods relevant to repository performance. There is paucity of test data on the most current defense glass (SRL-202) suitable for modeling of long-term dissolution rate. The WAPS test was intended as a quality control test, not as a performance test suitable for repository licensing. An alternative WAPS test methodology could be designed that would satisfy the original intent as a quality control test, yet provide more relevant data for evaluating the long-term dissolution rate of the defense glass. Because environmental variables such as temperature and pore water chemistry will affect the dissolution behavior of glass, no single performance confirmation test is possible.

Yucca Mountain Project - Argonne National Laboratory annual progress report, FY 1994

“…In the repository the main release scenario will occur because of reaction between liquid water and pre-aged glass. Preliminary performance assessment studies [5] that compare the release of radionuclides for glass and spent fuel have concluded that glass can provide a significant proportion of the radionuclide release for the overall repository. To refine this assessment and to provide a bounding limit to radionuclide release, further testing of pre-aged glass must be performed.…”

Critical review of glass performance modeling

“…Related to this topic is the known tendency for some glasses to precipitate specific secondary phases in leach tests. For example, high alumina glasses will alter to form analcime and maintain lower levels of dissolved silica than glasses with lower alumina contents [119]. The lowered dissolved silica concentrations 44 will increase glass dissolution rates.…”

“…These tests should cover a wide range of surface area-to-volume ratios (SA/V). By using the pH stat technique, high SA/V tests should scale with low SAN tests when normalized using SA/V-t. High SA/V tests in closedsystem tests with free-drift pH do not scale because of early rapid ion exchange that causes high SA/V solutions to react at higher pH values than low SA/V solutions [119]. Using high SA/V as an accelerated test procedure is justified if all system parameteters are measured and the reaction mechanism does not change.…”

A Strategy to Conduct an Analysis of the Long-Term Performance of Low-Activity Waste Glass in a Shallow Subsurface Disposal System at Hanford

“…The ion-exchange process has been largely ignored because it has been thought to be a short duration, secondary, or tertiary process that had little or no bearing on long-term corrosion or radionuclide release rates from glasses. The only significant effect identified in the literature that is attributed to alkali ion exchange is an increase in solution pH in static laboratory tests conducted at high surface area-to-volume ratios (Strachan et al 1990;Bourcier and Feng 1993).…”