PNNL

Abstract

This laboratory-scale investigation is focused on decreasing mobility of uranium in subsurface contaminated sediments in the vadose zone by in situ geochemical manipulation at low water content. This geochemical manipulation of the sediment surface phases included reduction, pH change (acidic and alkaline), and additions of chemicals (phosphate, ferric iron) to form specific precipitates. Reactants were advected into 1-D columns packed with Hanford 200 area U-contaminated sediment as a reactive gas (for CO2, NH3, H2S, SO2), with a 0.1% water content mist (for NaOH, Fe(III), HCl, PO4) and with a 1% water content foam (for PO4). Uranium is present in the sediment in multiple phases that include (in decreasing mobility): aqueous U(VI) complexes, adsorbed U, reduced U(IV) precipitates, rind-carbonates, total carbonates, oxides, silicates, phosphates, and in vanadate minerals. Geochemical changes were evaluated in the ability to change the mixture of surface U phases to less mobile forms, as defined by a series of liquid extractions that dissolve progressively less soluble phases. Although liquid extractions provide some useful information as to the generalized uranium surface phases (and are considered operational definitions of extracted phases), positive identification (by x-ray diffraction, electron microprobe, other techniques) was also used to positively identify U phases and effects of treatment. Some of the changes in U mobility directly involve U phases, whereas other changes result in precipitate coatings on U surface phases. The long-term implication of the U surface phase changes to alter U mass mobility in the vadose zone was then investigated using simulations of 1-D infiltration and downward migration of six U phases to the water table. In terms of the short-term decrease in U mobility (in decreasing order), NH3, NaOH mist, CO2, HCl mist, and Fe(III) mist showed 20% to 35% change in U surface phases. Phosphate addition (mist or foam advected) showed inconsistent change in aqueous and adsorbed U, but significant coating (likely phosphates) on U-carbonates. The two reductive gas treatments (H2S and SO2) showed little change. For long-term decrease in U reduction, mineral phases created that had low solubility (phosphates, silicates) were desired, so NH3, phosphates (mist and foam delivered), and NaOH mist showed the greatest formation of these minerals. In addition, simulations showed the greatest decrease in U mass transport time to reach groundwater (and concentration) for these silicate/phosphate minerals. Advection of reactive gasses was the easiest to implement at the laboratory scale (and presumably field scale). Both mist and foam advection show promise and need further development, but current implementation move reactants shorter distances relative to reactive gasses. Overall, the ammonia and carbon dioxide gas had the greatest overall geochemical performance and ability to implement at field scale. Corresponding mist-delivered technologies (NaOH mist for ammonia and HCl mist for carbon dioxide) performed as well or better geochemically, but are not as easily upscaled. Phosphate delivery by mist was rated slightly higher than by foam delivery simply due to the complexity of foam injection and unknown effect of U mobility by the presence of the surfactant.