278078 Evaluation of the Integrated Solubility Model, a Graded Approach for Predicting Phase Distribution in Hanford Tank Waste

Thursday, November 1, 2012: 1:45 PM
305 (Convention Center )
Kayla L. Pierson, Kendra R. Seniow and Jeremy D. Belsher, System Planning & Modeling, Washington River Protection Solutions, Richland, WA

Evaluation of the Integrated Solubility Model, a graded approach for predicting phase distribution in Hanford tank waste

The Department of Energy (DOE) Hanford tank farms store approximately 56 million gallons of radioactive and chemically hazard waste, most of which originated during the creation of plutonium for defense. The mission of the DOE River Protection Project (RPP) is to retrieve and treat Hanford's tank waste and close the tank farms and associated facilities to protect the Columbia River. The RPP mission is modeled by the Hanford Tank Waste Operations Simulator (HTWOS), which is the basis of strategic planning decisions. HTWOS is a dynamic flowsheet simulator and mass balance model that calculates the flow of events occurring during waste retrieval from the tank farms, pretreatment, supplemental treatment, and vitrification into solid glass canisters. The quantity of waste in each glass canister is limited by multiple waste components, whose solubility during pretreatment processing impacts the amount of waste in the glass. Small changes in waste phase distribution can have large impacts on the quantities of waste glass produced, the mission duration, and the lifecycle cost.

Hanford tank waste consists of over 150 radioactive and non-radioactive components in solid, liquid and slurry phases that have been stored for an extended time and often at elevated temperatures. However, the HTWOS is not a thermodynamically-based software, and waste phase distributions have historically been approximated by zero-order dissolution split factors based on extrapolation of limited experimental data. The integrated solubility model (ISM) was developed to improve the chemistry basis in the HTWOS and better predict the outcome of the RPP mission.

In order to characterize such a complex system, the ISM uses a graded approach to focus on the components that have the greatest impact to the mission while building the infrastructure for continued future improvement and expansion. Components are grouped depending upon their relative solubility and impact to the RPP mission. The solubility of each group of components is characterized by sub-models of varying levels of complexity, ranging from simplified correlations to a set of Pitzer equations used for the minimization of Gibbs Energy.

The ISM is evaluated by comparing ISM predictions to experimentally determined component distributions in tank waste reports and studies. The ISM is reviewed against the current HTWOS approach of simple split factors to determine if the ISM predictions more closely match experimental data than does the current methodology. Overall the ISM is an improvement over the simple split factors, although further development is necessary. The need for continuous improvement was anticipated and the evaluation process is designed to facilitate this.

Based on the complex chemistry in tank waste, an ISM prediction within a factor of two of the experimental results is determined to be a good estimate. The ISM accurately predicts sulfate, oxalate, chloride, iron, potassium, hydroxide and sodium within a factor of two of the experimental results. The ISM predicts the molality of sodium in solution to within 20% of the experimental value, iron and sulfate to within 50%, chloride and hydroxide to within 60%, oxalate to within 70%, and potassium to within 80%. The accuracy of ISM predictions for fluoride, phosphate, carbonate, nitrate, and nitrite are dependent on the type of experimental study with the ISM predicting the molality anywhere from within 10% to more than 2,000% of the experimental molality.

The ISM molalities differ from the experimental molalities partially because the ISM predicts different solids than the experimental compositions contained. Specific problems occur with natrophosphate [Na7F(PO4)319H2O] and sodium fluoride sulfate (Na3FSO4). Sodium fluoride sulfate was over-predicted by the ISM while natrophosphate was not predicted when it was experimentally determined to be present. Although sulfate, fluoride and sodium had good molality predictions from the ISM, phosphate did not, and the inconsistencies in the solids formed lead to this problem. Double sodium salts (natrophosphate and sodium fluoride sulfate) are more common at Hanford than in other systems, so more experimental data would be helpful to improve the accuracy of the ISM predictions.

Chromium is also an area needing improvement in the ISM. The simple equation that is used for chromium is insufficient in predicting chromium solubility in tank waste, and some trends for chromium predicted by the ISM are the opposite of the trends in the experimental data. Chromium undergoes redox reactions during phase change and the enabling assumption of nitrate and nitrite in those redox reactions may affect the accuracy of the ISM predictions for those components. More investigation into redox reactions and chromium speciation in Hanford tank waste is recommended. The ISM predictions for aluminum follow the same trends as shown in the experimental data but the aluminum solubility is always under-predicted by the ISM. Previous studies have shown that aluminum is commonly under-predicted in Hanford tank waste for reasons that are not well understood. Aluminum speciation, slow kinetics and the complex chemistry in tank waste mostly likely cause the under-prediction of aluminum solubility.

 


Extended Abstract: File Uploaded