Rheological Properties of an Alkoxide Derived HLW Slurry Feed

  • PDF / 354,740 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 106 Downloads / 189 Views

DOWNLOAD

REPORT


RHEOLOGICAL PROPERTIES OF AN ALKOX1DE DERIVED HLW SLURRY FEEDt L. H. CADOFF, D. SMIITH-MAGOWAN, D. E. HARRISON Westinghouse R&D Center, Pittsburgh, PA 15235 and J. M. POPE West Valley Nuclear Services Co., Inc., West Valley, NY 14171-0191

ABSTRACT Coagulation of HLW feed slurry can result from the interaction between acid metal cations and soluble silicates/silica. Feed slurry viscosity depend on a number of factors: (a) concentration and composition of the waste slurry electrolytes, (b) type and concentration of silica and silicate constituents in the glass former, (c) pH of the slurry feed, and (d) degree and order of mixing of the glass former with the waste sludge. Coagulation can be avoided by mixing to achieve a good dispersal of glass former globules in the waste slurry even when pH and electrolyte content favor it.

INTRODUCTION The reference process for immobilizing high level nuclear waste (ILW) in borosilicate glass at West Valley is to melt a slurry of glass frit and waste sludge in a liquid fed ceramic melter. An alternative to glass frit is to utilize a dispersion of glass forming additives for blending with the waste sludge[l],[2]. A method for doing this based on the hydrolysis of metal alkoxides demonstrated that it had several potential advantages over the glass frit method. These include: (a) intimate mixing of the glass forming components with waste particles to achieve more rapid melting into a more homogeneous, defect-free waste glass, and (b) a feed slurry which is non-abrasive, resistant to segregation, and readily transportable by air lifting. Success of the reactive mixing technique depends on producing a suspension of glass formers that will readily blend with neutralized waste sludge to form a weak coagulum. Bench scale experience was that weak promoters in dilute solutions would hydrolyze a boron/silicon alkoxide sufficiently to yield a suspension of borosiloxane that would produce a weak coagulum when blended with neutralized waste sludge. Thus, Na(NO) or NaCl gave a better dispersion of glass formers than NaOH. However, compositions based on either nitrates or hydroxides made better glasses than did chlorides. For these reasons a sodium and lithium nitrate glass composition was selected for the first continuous liquid fed melting trials at PNL. When this glass additive composition was mixed with neutralized Purex waste sludge, the result was a stable, readily air liftable feed slurry which could be concentrated up to -200 g/1 while retaining adequate flow over the surface of the melt. However, it proved to be an unsatisfactory liquid melter feed because of the formation of an insulating bubble layer at the melt/cold cap interface, apparently due to the decomposition of the nitrates. To overcome this problem, the glass additive composition was altered by replacing the sodium nitrate with sodium silicate and the lithium nitrate with lithium hydroxide. This new composition formed a stable, readily air liftable feed slurry which melted without foaming. However, at oxide loadings above 15