Plutonium Solubility in Simulated Savannah River Site Waste Solutions

Abstract

To address the accelerated disposition of the supernate and salt portions of Savannah River Site (SRS) high level waste (HLW), solubility experiments were performed to develop a predictive capability for plutonium (Pu) solubility. A statistically designed experiment was used to measure the solubility of Pu in simulated solutions with salt concentrations and temperatures which bounded those observed in SRS HLW solutions. Constituents of the simulated waste solutions included: hydroxide (OH⁻), aluminate , sulfate , carbonate , nitrate , and nitrite anions. Each anion was added to the waste solution in the sodium form. The solubilities were measured at 25 and 80°C. Five sets of samples were analyzed over a six month period and a partial sample set was analyzed after nominally fifteen months of equilibration.

No discernable time dependence of the measured Pu concentrations was observed except for two salt solutions equilibrated at 80°C which contained OH⁻ concentrations >5 mol/L. In these solutions, the Pu solubility increased with time. This observation was attributed to the air oxidation of a portion of the Pu from Pu(IV) to the more soluble Pu(V) or Pu(VI) valence states. A data driven approach was subsequently used to develop a modified response surface model for Pu solubility. Solubility data from this study and historical data from the literature were used to fit the model. The model predicted the Pu solubility of the solutions from this study within the 95% confidence interval for individual predictions and the analysis of variance indicated no statistically significant lack of fit.

The Savannah River National Laboratory (SRNL) model was compared with predicted values from the Aqueous Electrolyte (AQ) model developed by OLI Systems, Inc. and a solubility prediction equation developed by Delegard and Gallagher for Hanford tank waste. The agreement between measured or values predicted by the SRNL model and values predicted by the OLI AG model was very poor. The much higher predicted concentrations by the OLI AQ model appears to be the result of the model predicting the predominate Pu oxidation state is Pu(V) which is reported as unstable below sodium hydroxide (NaOH) concentrations of 6 M. There was very good agreement between the predicted Pu concentrations using the SRNL model and the model developed by Delegard and Gallagher with the exception of solutions that had very high OH⁻ (15 M) concentrations. The lower Pu solubilities in these solutions were attributed to the presence of and which limit the oxidation of Pu(IV) to Pu(V).

Keywords:

• alkaline high level waste salt solution
• plutonium
• Savannah River Site
• solubility