Wednesday, November 7, 2007 - 5:30 PM
493g

Simulant Development for Savannah River Site High Level Waste

Michael Stone, R. E. Eibling, D. C. Koopman, D. P. Lambert, and P. R. Burket. Savannah River National Laboratory, Aiken, SC 29808

The Defense Waste Processing Facility (DWPF) at the Savannah River Site vitrifies High Level Waste (HLW) for repository internment. The process consists of three major steps: waste pretreatment, vitrification, and canister decontamination/sealing. The HLW consists of insoluble metal hydroxides (primarily iron, aluminum, magnesium, manganese, and uranium) and soluble sodium salts (carbonate, hydroxide, nitrite, nitrate, and sulfate). The HLW is processed in large batches through DWPF; DWPF has recently completed processing Sludge Batch 3 (SB3) and is transitioning to Sludge Batch 4 (SB4).

Simulants are utilized to perform tests of the pretreatment process and vitrification process. The sludge is non-Newtonian and is typically described as a Bingham Plastic. The yield stress of actual waste is typically higher than the simulants prepared by the current preparation technique.

The original simulant preparation process was tested to determine if changes could be made to better match the actual rheological properties with the simulants. This sludge preparation process is performed in four steps: 1) Manganese nitrate precipitation by addition of potassium permanganate 2) Ferric and nickel nitrate precipitation by addition of sodium hydroxide 3) Solids washing to remove soluble sodium and potassium nitrate 4) Trim chemical addition to add remaining metals (Al, Ba, Ca, Cr, Mg, Na, etc.) and anions (nitrite, sulfate, oxalate). The yield stress from this recipe is sensitive to the particle size of the trim chemicals and the precipitation parameters.

A revised process was developed that added additional metal nitrates to the iron and nickel during the hydroxide precipitation. Trim chemical additions were limited to base-reactive species (silica and titanium oxide). This process resulted in simulants with slightly higher yield stresses than actual waste when tested in small batches. Scale-up of the process resulted in significantly higher yield stresses in the simulants, with some simulants too viscous to process through the lab-scale test apparatus. A small-scale continuous process was developed to allow large batches to be produced with the scale-up issues noted during the batch process.