(626g) Synthesis of Non-Radioactive Slurries to Simulate the Processing Behavior of Particles in Radioactive Waste Slurries

Legacy nuclear waste generated at the Savannah River Site (SRS) during Cold War production of enriched uranium and plutonium is currently being processed into a stable borosilicate glass waste form for long term storage. A majority of the legacy waste is stored as a mixture of insoluble hydroxides and hydrous oxides in large cylindrical storage tanks at SRS. Over 90% of the waste solids are non-radioactive chemical byproducts derived from fuel rod targets and purification chemistry. Uranium and plutonium are two of the major radionuclides by mass in the balance of insoluble material. The 3.5-4.5 thousand cubic meter (900,000-1,200,000 gallon) carbon steel waste storage tanks also contain 5-7M sodium solutions rich in hydroxide, nitrate, and nitrite anions that are contaminated with soluble radioactive isotopes of cesium and strontium.

The majority of the laboratory and pilot scale process development work being performed at the Savannah River National Laboratory (SRNL) in support of nuclear waste processing is done with non-radioactive simulated waste slurries, or ?simulants?. These match the bulk chemical properties of the non-radioactive components in the insoluble and aqueous phases of the actual radioactive wastes. Limited proof of concept testing is done in special facilities using radioactive samples from the actual waste tanks. Cost is one of the major factors driving this approach. There is roughly a ten-fold cost advantage to performing the preliminary testing with non-radioactive analogs to the real waste. The success of this strategy is dictated in part by how well the processing properties of the non-radioactive simulant slurries match those of the actual waste slurries.

Actual waste slurries were produced at SRS from about 1955-1990. Many have sat relatively undisturbed in the storage tanks since they were originally precipitated by adding caustic to the acidic waste streams. The freshly precipitated slurry particles are not in thermodynamic equilibrium. Atomic scale processes, such as Ostwald ripening, act to change the chemical and physical properties of the particles over time. For example, aluminum precipitates primarily as Gibbsite, Al(OH)₃, but a significant fraction converts to Boehmite, AlOOH, over time. Similar changes are likely occurring with the amorphous hydrous ferric oxide precipitate, Fe₂O₃?xH₂O. The percent crystallinity of the precipitated solids is low when examined by X-ray diffraction, and this is particularly true of the ferric oxide. The waste solids are very difficult to quantify with respect to speciation, crystal forms, etc. due to the presence of over fifty elements. The solids are believed to include double salts as well as simpler compounds involving single cations. The waste is, however, relatively free of any organic species (there is a small amount of oxalate).

Several different approaches have been taken historically toward preparing simulated waste slurries. All of the approaches used in the past dozen years involve some precipitation of the species using similar chemistry to that which formed the radioactive waste. All of the approaches add certain species as commercially available insoluble solid compounds. The number introduced in this manner, however, varies widely. All of the approaches make the simulated aqueous phase by adding the appropriate ratios of various sodium salts. The simulant system generally starts the preparation with an acidic pH and ends up with a caustic pH (typically in the 10-12 range).

Presently, all recipes first precipitate MnO₂ from Mn(NO₃)₂ and KMnO₄ by combining two aqueous solutions. Following MnO₂ precipitation, hydrous ferric oxide and nickel hydroxide are precipitated in parallel from a solution of their nitrates in the presence of the MnO₂ solids. Nitrates of many other elements may or may not be present and precipitated at this point including Al, Ba, Ca, Cu, Cr, Gd, La, Mg, Pb, Sn, Zn, and Zr. Alternatively, these species may be added as mixtures of oxides, hydroxides, sulfates, oxalates, carbonates, etc. as indicated by thermodynamic modeling and the bulk chemical distribution of the elements. Excess nitrate is removed by adding inhibited water (0.001 M NaOH, 0.001 M NaNO₂) and allowing the insoluble solids to settle, then decanting the aqueous supernate. Certain species, such as silica and TiO₂, are added as insoluble solids based on data indicating that the proper form or speciation would not be obtained from a nitrate precipitation. Finally, soluble sodium salts are added to adjust the composition of the aqueous phase.

Significant differences in physical properties of the simulated wastes are noted depending on the preparation route selected. The rheological properties of the slurries tend to be more viscous when more species are precipitated. Batch precipitations tend to produce more viscous slurries than continuous precipitations. Continuous precipitations tend to produce a narrower range of particle sizes than batch precipitations. The particle size distribution of the MnO₂ in the first-step precipitation depends on the presence or absence of the nitrates of the species being formed in the second-step precipitation. Precipitation of iron, aluminum, and other waste hydroxides from the nitrate solution using NaOH proceeded relatively unaffected by the particle size distribution of the MnO₂ particles formed in the first stage precipitation. The conclusion was that the MnO₂ particles were not acting as seed crystals for the other precipitated solids.

The various steps involved in producing the final simulant (two precipitations, a carbonate strike, batch washing, and final trimming) although lasting only about three weeks exhibited a gradually evolving particle size distribution. One disturbing trend was the gradual accumulation of fines as the number of processing steps increased. This was linked to undesirable increases in the apparent viscosity of the final simulants.

Particle size distributions of the commercial insoluble solids had a significant impact on the rheological properties of simulants where the precipitated species make up less than half of the total insoluble mass. Thermal treatments of the freshly precipitated slurries have caused minor changes in the physical properties. High shear mixers are being used to increase the apparent viscosity of simulants that are within a factor of two to three of the target rheological behavior.

The radioactive waste simulants being produced today mimic the chemical processing behavior of real waste essentially as well as can be quantified and with appropriate allowances for the absence of the radioactive species. Differences still remain in physical behavior, however. Currently, a program to improve physical properties of simulants is underway. One goal is to investigate how to produce particles that are foamier under boiling conditions. Real waste tends to foam when boiled, while simulants do not. Research at the Illinois Institute of Technology has established that biphyllic particles are stabilizing slurry foams. Simulants form such particles during the acidification phase of waste processing (acidified simulants are foamy during gas generation and boiling), but apparently do not have enough biphyllic particles as made.