Corrosion Testing of Low-Activity Waste Glasses Fiscal Year 1998 Summary Report

Analytical results are presented on the chemical composition and other physical properties of a glass, given the identification BNFL-A-S98, made at Pacific Northwest National Laboratory' that is representative of the low-activity waste glass composition proposed by BNFL, Inc.* for immobilization of envelope A double-shell tank wastes at the Hanford Site. This glass was prepared for use in a testing program to be conducted at Pacific Northwest National Laboratory and at Argonne National Laboratory for the purpose of characterizing its long-term corrosion behavior. Detailed examination of the glass microstructure using transmission electron microscopy showed structural features indicative of amorphous phase separation. A remelt was performed on a smaller batch (100 g) to ensure rapid cooling. The glass microstructure was reexamined and showed no evidence of phase separation. Selected long-term (some to 860 d) product consistency tests were terminated, and the leachates were analyzed on tests with three other representative low-activity waste glass formulations (L8- 1, L8-3, and L8-7). The results showed no evidence of corrosion rate acceleration at three times the duration of tests where another well-studied glass, LD6-5412, had been completely altered under identical test conditions. These tests (and others not discussed in this report) provide clear evidence that low-activity waste glasses with at least 20 mass% Na20 can be made that have excellent long-term corrosion resistance. However, glass composition has a large impact on long-term behavior and so careful experiments with several different techniques are essential to ensuring that a particular glass will have good long-term corrosion resistance