@ This document was printed on recycled paper. (9197)t r SummaryThis report describes two bench-scale column tests conducted to demonstrate the removal of Tc-99 from actual alkaline high salt radioactive waste. The waste used as feed for these tests was obtained fromlhe Hanford double shell tank AW-101 , which contains double shell slurry feed (DSSF). The tank sample was diluted to approximately 5 M Na with water, and most of the Cs-137 was removed using crystalline silicotitanates. The tests were conducted with two small columns connected in series, containing 10 mL of either a sorbent, ABEC 5000 (Eichrom Industries, Inc.), or an anion exchanger Reillex?-HPQ (Reilly Industries, Inc.). Both materials are selective for pertechnetate anion (TcO4-). The process steps generally followed those expected in a full-scale process and included 1) resin conditioning, 2) loading, 3) caustic wash to remove residual feed and prevent the precipitation of Al(OH)3, and 4) elution. A small amount of Tc-99m tracer was added as ammonium pertechnetate to the feed and a portable GEA counter was used to closely monitor the process. Analyses of the Tc-99 in the waste was performed using ICP-MS .with spot checks using radiochemical analysis. Technetium x-ray absorption spectroscopy ( U S ) spectra of 6 samples were also collected to determine the prevalence of non-pertechnetate species [e.g. Tc(IV)].Some of the relevant flow test parameters are given in Table S The 1525% of the Tc-99 that was not extracted corresponds to a Tc-99 level in the effluent of 1.65 X 10-2 Cum3 to 2.75 X 10-2 Ci/m3. This is 5.5% to 9.2% of the NRC Class A limit of 0.3 Ci/m3. The concentration in the final waste form would likely be higher.The relatively high fraction of non-pertechnetate Tc-99 in the AW-101 waste suggests that this phenomenon may be common in the tank wastes. It was expected that significant fractions of non-pertechnetate Tc-99 would be restricted to tank wastes with a high organic content. The DSSF from AW-101 contains a relatively low amount of organic ( approx. 2.5 g TOC/L in tank).As indicated by the So values, the effective Tc-99 capacity of the Reillexm-HPQ is greater than that for the ABEC 5000 material. The difference between the h50 values for the Tc-99m tracer data and the actual Tc-99 data is attributed to the non-pertechnetate fraction in the waste.Both materials are efficiently eluted, but the ABEC 5000 material is eluted with deionized water which would contribute a relatively small amount of other components to a final waste form.Few operational problems were experienced with either material, although a small portion of the ReillexT"-HPQ bed on the second column was observed to float in solution and preparation of the ABEC 5000 column was tedious because of the difficulty in removing entrained air from the bed. It is suggested that a larger particle size would alleviate this problem.--

LEGAL NOTICE l%is report was preparedby BattelleMemorialInstitute (Battelle)as an accountof sponsoredresearchactivities. Neither Clientnor Battelle nor any person actingon behalf of eitheM AKEs ANY WARIUNTY OR REPRESENTATION, EXPRESS OR IMPLIED, with respect to the accuracy, completeness, or usefulness of the informationcontainedin this repom or that the use of any information,apparatus, process, or compositiondisclosed in this report may not iniiinge privately om"ed rights; or Assumes any liabilitieswith respectto the use ofi or for damagesresultingfromthe use ot any information,apparatus,process,or composition disclosedin this report.

.Sixteen 1/25-scale sludge mobilization experiments were conducted in fiscal year (FY) 1993. The results of this testing are presented in this document. The ability of a single, centrally-located, scale model mixer pump to resuspend a layer of simulated tank sludge was evaluated for five different simulant types. The resistance of these simulants to the mobilizing action of the mixer pump jets was not found to adequately correlate with simulant vane shear strength. The data indicate that the simulant cohesion, as quantified by tensile strength, may provide a good measure of mobilization resistance.A single test was done to evaluate whether indexed mixer pump rotation is significantly more effective than the planned continuous oscillation. No significant difference was found in the sludge mobilization caused by these two modes of operation.Two tests were conducted using a clay-based sludge simulant that contained approximately 5 wt% soluble solids. The distance to which the mixer pump jets were effective for this simulant was approximately 50% greater than on similar simulants that did not contain soluble solids. The implication is that sludge dissolution el€& may significantly enhance the performance of mixer pumps in some tanks. The development of a means to correlate the magnitude of this effect with waste properties is a direction for future work.

SummaryThe mission of the Retrieval Process Development and Enhancements (RPD&E) project is to develop an understanding of retrieval processes, including emerging and existing technologies, gather data on these technologies, and relate the data to specific tank problems such that end-users have the requisite technical bases to make retrieval and closure decisions. The development of waste simulants is an integral part of this effort. The work of the RPD&E simulant-development task is described in this document. The key FY95 accomplishments of the RPD&E simulant-development task are summarized below. Development of Saltcake Simulants of Various Porous MediaA number of properties were examined to provide information regarding the correlation between mechanical strength and material structure of the baseline saltcake simulants. This investigation was conducted to obtain a better understanding of the large-scale production of saltcake simulant. The FY95 results of this work are shown in Table S. 1.In general, potassium magnesium sulfate (simulated saltcake) forms a cement-like material via a hydration reaction. Addition of different amounts of water (1 0 wt% to 30 wt%) changes not only the , quantity of reacted starting material but also the material-packing density, which directly affects the pore diameter. The conclusions from this study are as follows: -Water content controls the quantity of salt reacted in hydration reactions but not the degree of hydration reaction. These results have been confirmed by x-ray diffraction study and thermal analysis. 'Compressive strength of the saltcake simulant varies with water content due to the effect of water content on simulant porosity. During the formation of simulated saltcake, partial dissolution of the starting material affects the particle size and shape, which results in significant changes in porosity, pore diameter, and mechanical strength.Beyond a water content of about 14 to 16 wt?!, even though the total reacted water content is not changing according to the thermal analysis, the compressive strength decreases because the porosity gradually increases while the pore size remains relatively unchanged. Progress in GAAT-TS Sludge Simulant Development for RPD&E TestingA preliminary qualitative assessment of the physical properties of GAAT tank sludge was made based on the videotaped extrusion and testing of samples W07H302 through W07H304 from tank W-7. Two qualitatively different types of sludge were visible in these samples. In the lower regions of the tank, apparently there was a brown, cohesive sludge. Above the brown sludge was a yellow/orange sludge layer with noticeably different physical properties. Relevant physical and rheological data are not currently available for these wastes; thus, simulated ORNL wastes cannot be developed based on quantitative waste property data. Rheological measurements are needed to help develop simulants to support the development of mining path and end-effector orientation strategies. Also, operational A range of sludge simulants...

No abstract