Glass Macrocracking Determination in Prototypic Canisters Containing Lanthanide Borosilicate Glass

The Department of Energy Office of Environmental Management (DOE/EM) plans to conduct the Plutonium Disposition Project at the Savannah River Site (SRS) to disposition excess weapons-usable plutonium. Immobilization in a lanthanide borosilicate (LaBS) glass appears to be a viable option for the disposition of the plutonium. Several near-term data needs were identified to help increase confidence that the LaBS glass product is suitable for disposal in the Yucca Mountain Repository. One of these needs was to assess the degree of macroscopic cracking and/or voiding that occurs during processing of the Pu glass waste form and subsequent pouring of high level waste (HLW) glass. This data would then be utilized in repository modeling to provide a value for the actual glass surface area that would be available for leaching. This report summarizes testing and analyses that were completed on prototypic cans of LaBS glass to provide this data. Prototypic cans were filled with LaBS glass produced in the Cylindrical Induction Melter (CIM), then exposed to a thermal treatment to represent conditions that would be expected during pouring of HLW glass around the Pu glass product cans in the DWPF (Smith, 2000). The cans containing the LaBS glass were subjected to non-destructive evaluation (NDE) and destructive analysis techniques to assess the amount of cracking and/or voiding that occurred during processing. Testing conducted on large canisters or full-scale DWPF canisters have shown that thermal stresses and glass/canister interactions result in a 7 to 27 times increase in surface area as compared to the monolithic geometric surface area (Peters, 1981). Earlier testing conducted at SRS with simulated HLW glass and smaller diameter canisters (∼20 cm) resulted in a surface area increase of 5 times for canisters allowed to air cool (Kessler, 1982). Testing conducted under this task on stainless steel canisters of 2.87 inch inside diameter indicated thermal stresses and glass/canister interactions result in a 1.9 to 4.1 times increase in surface area due to glass cracking. Comparison of digital radiography images taken of the canisters prior to and after heat treatment suggests most of the cracking occurred as a result of stresses induced by canister deformation during the heat treatment. Dimensional constraints of the heat treatment furnace required the prototypic canisters be positioned horizontally during the heat treatment process, resulting in a slight ''flattening'' of the canisters, inducing stress in the LaBS glass upon cooling, resulting in cracking. Therefore, this additional cracking was an artifact of the test configuration and not the actual thermal treatment. Digital radiography images also show the presence of only very small voids within the LaBS glass prior to heat treatment, which decreased in number and size during the heat treatment. The data produced by testing conducted in this task is conservative in that the prototypic canisters were heat treated in a horizontal orientation, rather than vertically as will be the case in actual processing within the DWPF canisters. The cracking can also be reduced in the prototypic canisters by controlling the initial cooling rate as the LaBS glass is poured into the cans. Conservatism in the data is also realized in that the calculations of surface area increases due to glass cracking assume the cracks measured at the exposed surface of each can section extend completely through the length of the section