Results are presented of in-situ measurements, performed in a 300 m deep dry-drilled borehole, in the ASSEmine. Convergence measurements at ambient as well as elevated temperatures and pressure measurements at elevated temperatures are discussed. Creep equations derived from these experiments are used for the numerical analysis of the time dependent behaviour of a salt dome with a HLW repository. The analyses show that the total stresses in the salt remain compressive with deviatoric components smaller than 3 MPa.

The long term corrosion behaviour of the German HLW glass SM 58 LW 11, doped with 11 vt.% of inactive waste, was investigated in a simulated carnallite/rock salt environment. The experimental conditions were defined with respect to a hypothetical water entrance to a rock salt repository, i.e. temperatures up to 200°C and a pressure of 130 bar were applied; a specific brine from the quinary system MgCl2-MgSO4-NaCl-KCI-H2O (the so-called Q-brine [1] served as leachant. The effects of the corrosion were determined by weight loss measurements, by analyses of the leachates, and by the application of surface analytical techniques.

As a result of the heat producing high-level radioactive waste, volatile components which are in the host rock will be liberated and further gases will be generated by thermal cracking and radiolysis.

Six metal alloys were subjected to surface and creeprupture stress corrosion tests using NaCl- and Q-brines. Low-allou materials do not seem to be suitable as a barrier against salt brine attack due to the high surface corrosion rate. Creep-rupture strength corrosion was insignificant for the tested materials. Hastelloy C 4 showed no corrosion attack by salt brines in autoclaves at 200°C and 100 bar during 8 months of testing. Bentonite was shown to reduce the corrosion attack considerably.

The generation and behaviour of realcolloids and pseudocolloids of the actinides, 238 pu, 241Am and 244Cm, are investigated in four different Gorleben Groundwaters. Colloids are characterized by ultrafiltration at different pore sizes from 1 nm to 450 nm and various analytical instrumentations. Behaviour of pseudocolloids has been analysed by determining their effects on the solubility and the distribution coefficient of actinides between groundwater and geomatrix. Sorption of actinide ions on groundwater-colloids (generation of pseudocolloid) are examined and the migration of actinides as pseudocolloids is discussed.

Experiments concerning the sorption of radionuclides carried out on the site-specific unconsolidated sediments have been in progress in 6 laboratories since 1981 as part of the Gorleben site investigation program for the most relevant radionuclides. Sorption data are being determined for the site-specific conditions, and the influence of parameters, such as e.g. pH, Eh, complexing agents and NaCl concentrations is investicated. The data are necessary for the safety assessment of the repository. As the investigations have not yet been terminated, the results represent a preliminary state.

Pressurized brine reservoirs have been found in the Castile anhydrite which underlies the proposed Waste Isolation Pilot Plant. The major and minor component chemistry of the brines and the isotopic abundances of the brines and coexisting minerals have been determined and evaluated. Thermodynamic calculations indicate that the brines approach saturation with all major minerals in their host horizon. The brines therefore, appear to be in equilibrium with their geologic environment and do not have the capacity for substantially degrading the integrity of the site. When analyzed through solute versus bromide plots, the data indicate that the most likely origin for the brines is Permian seawater. As a result, the reservoirs appear to have been derived from waters no longer being introduced into the rock (i.e., they have a finite source).

The diffusion of plutonium (IV) and pertechnetate and the migration of colloidal gold in brine-saturated bentonite was measured. High ionic strength brine characterizes potentially intruding groundwater for radioactive waste repositories in salt. Plastic diffusion cells containing cylindrical diffusion columns were used for low density bentonite. Metal diffusion cells constructed entirely from Hastelloy C-276 were used for the diffusion of pertechnetate in highly compacted bentonite. Apparent distribution coefficients calculated from plutonium diffusion in two columns of low density bentonite are 2 m3/kg and 3 m3/kg. These values are in good agreement with the value of 3m3/kg that was calculated from previous batch sorption data. Pertechnetate anions migrated out of a brine-saturated high density bentonite diffusion column at rates that are too large to be rationalized with a simple diffusion theory. Additional measurements that take into account possible channeling of pertechnetate are required. Colloidal gold was excluded from low density brine-saturated bentonite, but colloidal gold may have channeled between the bentonite and the wall of the diffusion cell. These results support the effectiveness of brine-saturated bentonite as an engineered barrier to plutonium. The results also highlight the need for additional measurements of pertechnetate and gold transport in bentonite. Needs for transport measurements that take into account site-specific materials interactions are also described.

A six-part, waste package materials field test was conducted in a halite horizon of a potash mine in southeastern New Mexico. The primary purposes of this test were to evaluate the thermophysical and geochemical performance of candidate HLW-package backfill materials emplaced in rock salt and the corrosion behavior of candidate waste canister or overpack alloys. This field test series also served as a precursor to forthcoming Waste Isolation Pilot Plant (WIPP) in situ waste package performance experiments on simulated defense high-level waste packages, serving to develop applicable testing, instrumentation, and sampling techniques. The backfill materials tested (individually, in one- to five-month tests) were: low-density bentonite clay; low-density bentonite (70 wt.%)-silica sand (30 wt.%) mixtures, both dry and brine-injected; high-density bentonite-sand annular compacts; trapped air; and finely-crushed WIPP salt. The in situ measured thermal conductivities (at a maximum canister-heater surface temperature of 150° or 250°C) for the backfills ranged from 0.25 W/mK for pure bentonite to about 1.25 W/mK for the high-density bentonite-sand. No significant backfill material degradation products were detected in post-test analyses. No appreciable corrosion of the titanium-, nickel-, or iron-based alloys embedded in the hot backfill was found; potentially significant pitting corrosion of 2 1/4 Cr-1 Mo steel and copper was detected.

The relative importance of material properties, geochemical processes, and environmental parameters affecting containment and release of radionuclides vary strongly as a function of time and spatial position within a nuclear waste repository located in basalt (NWRB). A simple matrix of spatial regions (Engineered-Barrier System, Disturbed-Rock Zone, Site System of the Controlled Access Zone) and time periods (Pre-emplacement, Containment, Isolation/Slow-release)can be used to identify the dominant processes that affect the containment, release, and migration of radionuclides. From this analysis, a directed and efficient program of field and laboratory studies can be conducted to obtain the data required to assess the feasibility of an NWRB and compliance with Federal regulatory criteria.

Geologic features are identified that may affect groundwater movement near a repository located in the Columbia River basalt beneath the Hanford Site, Washington State. These include pathways through flow interiors, flow contacts, and bedrock structural discontinuities. Four concepts are considered to describe groundwater movement in basalt. Differences between concepts principally focus on the degree of vertical leakage through basalt flow interiors. The existing geohydrologic data base is preliminary and insufficient to conclusively support one groundwater-flow concept over another. However, it is proposed that available data tend to support the existence of distinct groundwater-flow systems having relatively low vertical leakage under low hydraulic-head gradients except where structural discontinuities are present. A number of geohydrologic characterization activities are underway to reduce the Present uncertainty regarding potential groundwater pathways.

High-level nuclear Waste emplaced in a repository in basalt will lead to elevated temperatures and chemical reactions between the basalt and repository groundwater. The resultant changes in groundwater chemistry and the formation of secondary minerals will affect radionuclide release rates from the repository. In this study, Grande Ronde Basalt and synthetic groundwater were reacted at temperatures of 100°, 150°, and 300°C at a pressure of 30 MPa. Dickson-type sampling autoclaves were used to follow solution composition changes with time. Solution PH remained weakly alkaline, while steady state concentrations of F-, Cl-, SO4-2, and total Carbon were similar to or lower than initial values. Alteration assemblages at 300°C included silica, zeolites, potassium feldspar, and iron smectite. These assemblages are metastable, and prediction of alteration in the basalt and packing material depends on “accelerated” tests results and must be supplemented by study of metastable assemblages in natural analog systems and by tests in open systems with varying flow rates. Experimental data are consistent with rapid adjustment of Eh to reducing values; however, more work on reaction rates and buffering capacity is in progress.

The Basalt Waste Isolation Project is conducting experiments to assess the stability of bentonite (sodium montmorillonite)and crushed basalt as waste package packing materials in a nuclear waste repository in basalt. The experiments are designed to identify changes in physical, chemical, and mineralogical properties that these materials could undergo in the repository environment. A series of bentonite dehy-dration experiments showed that after 1 year at 370°C the clay's structural and swelling properties were preserved and only reversible dehydration occurred. At 440°C, however, irreversible dehydration, collapse of the clay structure, and loss of swelling ability took place. Hydrothermal tests using bentonite, or an equal mixture of bentonite and basalt, along with synthetic groundwater, were also conducted. Abentonite + water experiment showed negligible structural alteration at 200°C, although some iron enrichment of the clay occurred. At 300°C, partial conversion of the mont-morillonite to an iron- and potassium-rich smectite occurred, along with the formation of secondary quartz and albite. An experiment in the basalt + bentonite + water system at 150°C resulted in partial etching of the basalt grains, but no detectable change in rock or clay mineralogy. At 300°C, the basalt was strongly etched. Furthermore, iron- and potassium-rich smectite apparently replaced montmorillonite as the dominant clay, and secondary silica, zeolites, and minor feldspar were formed. These mineral assemblages are thought to be metastable at 300°C, based on natural analogs. The kinetics and reaction paths of further transformations of packing materials are important areas for continued study.

Three low-carbon steels (AISI 1006, AISI 1020, and AISI 1025) were tested at 150° and 250°C in Hanford Grande Ronde Basalt groundwater (9.75 pH) under anoxic conditions (≤0.1 mg/L oxygen in water) in packing (75 wt% basalt + 25 wt% bentonite) for 1 and 2 weeks. Testing was initially performed in air-equilibrated groundwater (oxic conditions) without packing at 150°C. The average corrosion under anoxic conditions with packing was at least a factor of seven lower than that for oxic conditions at 150°C. The data indicated that carbon composition apparently affects the corrosion of low-carbon steel under oxic conditions. Under anoxic conditions with packing, corrosion was independent of the carbon composition of the steel as evidenced by almost equal weight losses for all three steels at 150° and 250°C. The weight loss at 250°C was about the same as that at 150°C due to the formation of a very adherent layer of iron-rich clay on the surface of the steel at 250°C, reducing the rate at which corrosion proceeds. Pitting was not detected in any of the specimens. Assuming linear corrosion kinetics, extrapolation of the 2-week data to 1,000 yr resulted in approximately 12 mm penetration of the low-carbon steel in Grande Ronde Basalt groundwater under anoxic conditions with packing.

The corrosion behavior of several iron-base and titanium-base alloys was studied in synthetic Grande Ronde Basalt groundwater at temperatures of 150°C to 2500°C and under irradiation dose rates to 2 × 106 rad/hr. The objective of these ongoing studies is to help select one or more materials for waste-package canisters that will maintain their integrity for time periods up to 1,000 yr in a nuclear waste repository constructed in basalt. The corrosion rates of iron-base alloys under irradiated conditions were generally 2 to 3 times as high as those obtained on similar materials under nonirradiated conditions. The titanium alloys exhibited low corrosion rates but absorbed significant amounts of hydrogen under irradiated conditions.

The release of polyvalent radionuclides from a nuclear waste repository located in basalt may be sensitively related to the redox potential (Eh) imposed by the basalt. A series of tests are reported here evaluating the effect of basalt on the concentrations of 99Tc released into solution from a borosilicate glass waste form. Crushed PNL 76-68 glass, doped with 0.7 mg 9 99Tc/g glass, was reacted with reference basalt groundwater under oxic hydrothermal conditions in a sampling autoclave, both alone and in the presence of crushed basalt. The steady state fluid concentrations of 99Tc and various stable species were determined from samples obtained at the test conditions of 200 °C, 30 MPa, and a initial solution to solid mass ratio of 10 for both tests.

In the glass + groundwater test, the 99Tc concentration rose rapidly to about 50 mg/L after only 200 hr of run time and remained at a value between 50 and 60 mg/L throughout the duration of the test. For the basalt + glass + groundwater test, the 99Tc concentration rose to an initial value of about 2.5 mg/L. At about 700 hr, the 99Tc concentration began to drop rapidly until a value near the analytical detection limit (approximately 0.005 mg/L) was reached after a test duration of 1,400 hr. It is concluded that the presence of basalt in these hydrothermal experimtents reduces the concentration of 99Tc in solution by nearly four orders of magnitude, probably by control of solution Eh and subsequent precipitation of a solid containing a reduced form of technetium. Reaction mechanisms are discussed that can account for these observations.

The Grande Ronde Basalts underlying the Hanford Site are being evaluated as a possible site for a high-level nuclear waste repository. Experiments, in which basalt from the Umtanun flow of the Grande Ronde Basalt and basalt with simulated spent fuel were reacted with synthetic Hanford groundwater, were conducted to determine steady state concentrations which can be used in radionuclide release-rate models. Tests were performed at temperatures of 100°, 200°, and 300°C; 30 MPa pressure, and a solution:solid mass ratio of 10:1 for durations up to 7,000 hr. Solution aliquots were extracted periodically during the experiments for analysis. The pH was measured at 250°C and recalculated to higher temperatures. In the basalt-water system the stable high-temperature pH values achieved were 7.2 (100°C), 7.5 (200°C), and 7.6 (300°C). Solution composition variations are due to mesostasis (glass) dissolution and precipitation of secondary phases. Solution measurements indicate a redox potential (Eh) of about -0.7 volts at 300°C. Secondary phases produced include silica, potassium feldspar, iron oxides, clays, scapolite, and zeolites. Tests in the basalt + simulated spent fuel + water systen show that calculated pH values stabilized near 7.6 (100°C), 7.2 (200°C), and 7.7 (300°C). At higher temperatures, solution concentrations were controlled by secondary phases similar to those found in basalt-water tests. Less than 1% of uranium, thorium, samarium, rhenium, cerium, and palladium were released to solution while somewhat higher amounts of iodine, molybdenum, and cesium were released. The UO2 component was unreactive; however, other components (e.g., cesium-bearing phases) were almost completely dissolved. Secondary phases incorporating radionuclide-analog elements include clays, palladium sulfide, powellite, coffinite, and a potassium-uranium silicate.

Gamma radiolysis of reference Grande Ronde Basalt groundwater containing 700 mg/L methane produces a milky liquid that is a suspension of fine particles of organic polymers similar to polyethylene. The amount and nature of these polymers does not vary much over the range of conditions tested. Differences in the radiolytic products that are formed may occur when host rock and engineered-barrier materials are present. This possibility is currently being investigated as is the potential for interactions between radionuclides and the products of radiolysis.

The design, testing, and calibration of metallic Eh and zirconia pH sensors for use at elevated temperatures and pressures are discussed.

This paper describes a method of stochastically analyzing groundwater traveltime. The method uses a Monte Carlo technique to generate a suite of random spatial fields that are subsequently input to the groundwater flow and groundwater traveltime equations. Stochastic inputs to these equations can be (1) transmissivity (or hydraulic conductivity), (2) effective thickness (or effective porosity), or (3) boundary conditions. In a transient problem, storage coefficient (or specific storage) could also be stochastically treated. Spatial correlation in the random input fields is accounted for by means of a multivariate random-number generator, which requires the first two statistical moments of these fields to be specified. The output from the Monte Carlo analysis is a suite of groundwater traveltime realizations that can be used to derive exceedance probabilities. These probabilities provide a measure of the degree of confidence in meeting set criteria.

A preliminary application of this method using data from the deep basalts beneath the Hanford Site is also presented. The results illustrate how this method can be used to evaluate compliance with a technical criterion relating to groundwater traveltime.