Impact of Eliminating Mercury Removal Pretreatment on the Performance of a High-Level Radioactive Waste Melter Offgas System

Abstract

Abstract

The Defense Waste Processing Facility (DWPF) at the Savannah River Site processes high-level radioactive waste from the processing of nuclear materials that contains dissolved and precipitated metals and radionuclides. Vitrification of this waste into borosilicate glass for ultimate disposal at a geologic repository involves chemically modifying the waste to make it compatible with the glass melter system. Pretreatment steps include removal of excess aluminum by dissolution and washing, and processing with formic and nitric acids to (1) adjust the reduction–oxidation (redox) potential in the glass melter to reduce radionuclide volatility and improve melt rate; (2) adjust feed rheology; and (3) reduce, by steam stripping, the amount of mercury that must be processed in the melter. Elimination of formic acid pretreatment has been proposed to eliminate the production of hydrogen in the pretreatment systems; alternative reductants would be used to control redox. However, elimination of formic acid would result in significantly more mercury in the melter feed; the current specification is no more than 0.45 wt%, whereas the maximum expected before pretreatment is about 2.5 wt%. An engineering study has been undertaken to estimate the effects of eliminating mercury removal on the melter offgas system performance. A homogeneous gas-phase oxidation model and an aqueous phase model were developed to study the speciation of mercury in the DWPF melter offgas system. The model was calibrated against available experimental data and then applied to DWPF conditions. The gas-phase model predicted the \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document}\begin{align*}{\rm Hg}_2^{2+} / {\rm Hg}^{2+}\end{align*}\end{document} ratio accurately, but some un-oxidized Hg⁰ remained. The aqueous model, with the addition of <1 mM Cl₂ showed that this remaining Hg⁰ would be oxidized such that the final \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document}\begin{align*}{\rm Hg}_2^{2+} / {\rm Hg}^{2+}\end{align*}\end{document} ratios matched the experimental data. Results of applying the model to DWPF show that due to excessive shortage of chloride, only 6% of the mercury fed is expected to be chlorinated, mostly as Hg₂Cl₂, whereas the remaining mercury would exist either as elemental mercury (90%) or HgO (4%).