Hydrothermal Experiments (300 Degrees Celcius) Conducted Using Bannock Rhyolite Tuff (Montana) And Conway Granite (New Hampshire): Implications For High-Level Nuclear Waste Disposal And Comparisons With Geothermal Solution Data

Abstract

Holocrystelline Conway granite, and 95X glass Bannock Rhyolite Tuff were reacted with 3.4x10·3M NaCl solution in Dickson rocking autoclave experiments conducted at 300°C. Granite and tuff experiment data were compared against similar basalt data to determine the effects of rock type (granite, glassy tuff, and 10·40X glass Basalt) and crystallinity in containing high-level nuclear waste. These data were evaluated to determine the degree to which short-term autoclave data correspond with the long-term chemistry of geothermal solutions. Redox comparisons of 300°C autoclave solutions indicate that the stable log fO2 produced by reactions with rhyolitic glass are 3-5 units more oxidizing than those with granite or basalts. More oxidized redox potential may result in higher waste container corrosion rates and solubility of UO2 for spent fuel rods, at a tuff repository site (i.e., Yucca Mountain, Nevada). The redox buffering capacity (based on FeO content) is also lowest for the glassy tuff, followed by granite and basalts, respectively. The exhaustion of highly soluble rhyolitic glass by oxidizing repository solutions may result in the more rapid acceleration of container corrosion· and U-complex release rates at a tuff repository (compared to granite and basalt). In these tuff and granite experiments Fe-Mg poor, Al-rich dioctahedral beidellitic smectites were formed. In comparison Fe·Mg rich, Al-poor trioctahedral smectite (saponite) was formed in 300°c basalt autoclave solutions. The precipitation of beidellitic smectites in the tuff and granite solutions resulted in comparatively acidic solutions (compared to basalt). For tuff and granite, lower repository solution pH may result in comparatively higher container corrosion rates, and greater solubility for UO2 in spent fuel rods. The illite/smectite mass ratio is higher for the secondary clay minerals removed from the tuff and granite experiments (compared to basalt). The formation of comparatively larger amounts of sorptive smectites in a basalt repository should result in comparatively lower radionuclide mobility (compared to tuff end granite). Lesser amounts of secondary minerals were formed in the granite experiments. The formation of less secondary minerals in granite repository solutions would result in comparatively higher radionuclide mobility at such a site (compared to basalt and tuff). High concentrations of Fin glass-influenced tuff repository solutions may result in higher solubility of uo2 for spent fuel rods. These geochemical criteria suggest that basalt may be the most suitable of the three repository host rock types compared (followed by granite and tuff, respectively). Experimental solution geothermometer, cation/proton mass ratio, and secondary mineral data from our high SiO2-rock:water experiments (holocrystalline granite and 95X-glass tuff), Temple basalt (10-40X glass) experiments and geothermal fields, when compared indicate that the granite autoclave solution came closest to emulating long-­term geothermal fluid chemistry, followed by the basalt and tuff solutions, respectively. These data indicate that the time interval necessary for short-term autoclave experiments to predict long-term geothermal fluid chemistry is a function of glass content. Natural geothermal field data is considered to be useful in predicting the long-term chemistry of high-level nuclear waste repository solutions. Our results indicate that rock crystallinity should be considered when selecting geothermal analog data for such purposes.