PNNL

Abstract

Product Consistency Tests (PCT) and Vapor Hydration Tests (VHT) were performed on a radioactive glass made from Hanford tank AP-107 waste (i.e., R-AP-107) and its non-radioactive equivalent glass made from simulant tank waste (i.e., S-AP-107). The radioactive and non-radioactive experiments were conducted in parallel with identical equipment, protocols, and staff to remove additional experimental uncertainties. PCT and VHT results are reported for all samples and are shown in Table S.1. Average normalized mass loss of B (NLB) and Na (NLNa) calculated from the three replicate test results for each glass with their respective standard deviations (SD) are presented for PCT results while individual VHT rates (rVHT) with their propagated errors (s_(r_prop )) from individual thickness measurements on each sample are presented for the VHT results. Some VHT samples experienced more than 50% evaporation loss during the experiments. However, no visible alteration was detected on any VHT samples regardless on the amount of evaporation loss detected as shown in Table S.1. Table S.1. Average NLB and NLNa and standard deviations for the three replicates of R-AP-107 and S-AP-107, and individual VHT rates (rVHT) with propagated errors (s_(r_prop )) from individual thickness measurements. The DOE limits for NLB, NLNa, and rVHT are provided for comparison. R-AP-107 S-AP-107 Average NLB ± SD (g·m 2) 0.51 ± 0.05 0.37 ± 0.01 Average NLNa ± SD (g·m 2) 0.61 ± 0.04 0.42 ± 0.01 NLB and NLNa DOE Limits (g·m 2) 2 2 VHT Replicate 1: rVHT ± s_(r_prop ) (g m-2 d-1) 3 ± 5 -1 ± 3 VHT Replicate 2: rVHT ± s_(r_prop ) (g m-2 d-1) 2 ± 4 0 ± 3 VHT Replicate 3: rVHT ± s_(r_prop ) (g m-2 d-1) 4 ± 7 3 ± 8 rVHT DOE limit 50 50 Underlined values indicate the test had more than 50% evaporation loss Statistical analyses were performed on the results and indicated that the test responses were within experimental uncertainty for all PCT and VHT experiments. The experimental uncertainties were estimated using pooled standard deviations determined in Vienna et al. (2022) which incorporated estimates of SDs from several replicate sets of other low-activity waste (LAW) glasses to include uncertainties due to fabricating glasses, performing experiments, and sample measurements (e.g., leachate analysis). The pooled standard deviations are a better representation than simple SD determined from three replicates for one glass. Statistical analysis was also performed using simple SD and propagated errors for the VHT samples for comparison. The VHT samples were still within experimental uncertainty regardless of how the experimental uncertainty was represented. These results help to confirm the Waste Treatment and Immobilization Plant composition control methods that predict PCT and VHT responses using models fit primarily to non-radioactive simulant glasses are valid for radioactive tank waste glasses.