Development of Crystal-Tolerant High-Level Waste Glasses

Twenty five glasses were formulated. They were batched from HLW AZ-101 simulant or raw chemicals and melted and tested with a series of tests to elucidate the effect of spinel-forming components (Ni, Fe, Cr, Mn, and Zn), Al, and noble metals (Rh2O3 and RuO2) on the accumulation rate of spinel crystals in the glass discharge riser of the high-level waste (HLW) melter. In addition, the processing properties of glasses, such as the viscosity and TL, were measured as a function of temperature and composition. Furthermore, the settling of spinel crystals in transparent low-viscosity fluids was studied at room temperature to access the shape factor and hindered settling coefficient of spinel crystals in the Stokes equation. The experimental results suggest that Ni is the most troublesome component of all the studied spinel-forming components producing settling layers of up to 10.5 mm in just 20 days in Ni-rich glasses if noble metals or a higher concentration of Fe was not introduced in the glass. The layer of this thickness can potentially plug the bottom of the riser, preventing glass from being discharged from the melter. The noble metals, Fe, and Al were the components that significantly slowed down or stopped the accumulation of spinel at the bottom. Particles of Rh2O3 and RuO2, hematite and nepheline, acted as nucleation sites significantly increasing the number of crystals and therefore decreasing the average crystal size. The settling rate of ≤10-μm crystal size around the settling velocity of crystals was too low to produce thick layers. The experimental data for the thickness of settled layers in the glasses prepared from AZ-101 simulant were used to build a linear empirical model that can predict crystal accumulation in the riser of the melter as a function of concentration of spinel-forming components in glass. The developed model predicts the thicknesses of accumulated layers quite well, R2 = 0.985, and can be become an efficient tool for the formulation of the crystal-tolerant HLW glasses for higher waste loading. A physical modeling effort revealed that the Stokes and Richardson-Zaki equations can be used to adequately predict the accumulation rate of spinel crystals of different sizes and concentrations in the glass discharge riser of HLW melters. The determined shape factor for the glass beads was only 0.73% lower than the theoretical shape factor for a perfect sphere. The shape factor for the spinel crystals matched the theoretically predicted value to within 10% and was smaller than that of the beads, given the larger drag force caused by the larger surface area-to-volume ratio of the octahedral crystals. In the hindered settling experiments, both the glass bead and spinel suspensions were found to follow the predictions of the Richardson-Zaki equation with the exponent n = 3.6 and 2.9 for glass beads and spinel crystals, respectively