The size and density of a 242Pu colloid has been measured by autocorrelation photon spectrometry. The sedimentation rate of the colloid was determined by ultraspeed centrifugation. From the concentration profiles of 242Pu in the centrifuged test tubes, a standard sedimentation formula was used to calculate the density; the size of the colloid was known from the light scattering experiments. The determined density of the 242Pu colloid was only slightly lower than the estimated density of crystalline PuO2.

Uranium, Th, Ra, Rn, Pb and Po radionuclide concentrations in ground waters from the Hanford Site indicate that U, Th, and Ra are highly jorbed. Relative to Rn, these radionuclides are low by factors of 10−3 to 10−6. Uranium sorption is likely due to its reduction from the +6 state, where it is introduced via surface waters, to the +4 state found in the confined aquifers. The distribution of radionuclides is very similar in all of the confined aquifers and significantly different from the distribution observed in the unconfined and surface waters. Barium correlates well with Ra over three orders of magnitude indicating that stable element analogs may be useful for inferring the behavior of radioactive waste radionuclides in this candidate geologic repository.

Previous analytical models for the steady-state radionuclide release rate through perforations in nuclear waste containers into the surrounding medium, are based on a zero wall thickness assumption. In this paper we investigate the effect of the wall thickness on the mass transfer rate through isolated circular holes. We solve the steady-state diffusion equation for the concentration field and derive a model based on the analytical solution. By direct comparison, we show that the zero wall thickness model overpredicts the mass transfer rate by about 1300 percent for a circular hole with 1 cm radius and a wall thickness of 10 cm. As expected, the zero-thickness model becomes even less accurate as the hole radius decreases; it predicts a greater release rate from a large number of small holes than the mass transfer rate from an uncontained waste form cylinder. In contrast, the results predicted by our model remain bounded for isolated holes and never exceed the mass transfer from an uncontained waste form.

Gas-phase carbon-14 dioxide released from spent fuel buried in unsaturated rocks will move upward by molecular diffusion and advection. The flux resulting from molecular diffusion will be determined by the concentration gradient of gaseous carbon-14 dioxide. Advection will be caused by density-driven flow of the pore gas. Such flow occurs under present conditions as a result of annual and diurnal temperature changes and differences in chemical composition between pore gas and air, and it will be enhanced when the spent fuel heats the rock. The advective flux will be determined by the gas-phase concentration of carbon-14 and by the velocity of convective flow of the pore gas. Gas-phase concentrations of carbon-14 dioxide will be strongly influenced by isotopic exchange with bicarbonate dissolved in water that is retained in the unsaturated zone by capillary forces. Isotopic equilibrium between these two phases will be attained very rapidly on repository time scales. If the liquid phase is saturated with calcium carbonate, precipitation of calcite may provide a sink for removal of carbon-14 from the mobile phases.

The dissolution rate of waste solids in a geologic repository is a complex function of waste form geometry, chemical reaction rate, exterior flow field, and chemical environment. We present here an analysis to determine the steady-state mass transfer rate, over the entire range of flow conditions relevant to geologic disposal of nuclear waste. The equations for steady-state mass transfer with a chemical-reaction-rate boundary condition are solved by three different mathematical techniques which supplement each other. This theory is illustrated with laboratory leach data for borosilicate-glass and a spherical spent-fuel waste form under typical repository conditions. For borosilicate glass waste in the temperature range of 57°C to 250°C, dissolution rate in a repository is determined for a wide range of chemical reaction rates and for Peclet numbers from zero to well over 100, far beyond any Peclet values expected in a repository. Spent-fuel dissolution in a repository is also investigated, based on the limited leach data now available.

This investigation addresses the problem of prediction of the rate of migration of redox-sensitive solution species within packing and backfill materials under conditions of variable oxidation potential. Effects of changes of oxidation potential and precipitation of stable uranium compounds during diffusion of uranium from a region of high oxidation potential into a region of low oxidation potential were simulated numerically. Questions of particular interest addressed in the investigation were the existence of a moving “redox front” and the influence of precipitation-dissolution processes on uranium migration. The simulations showed that no expanding redox fronts existed at any simulated time up to 3.2×105 years (1013 s). In simulations where precipitation of stable solids was not allowed, variations of oxidation potential did not affect total uranium concentrations in solution. Concentration profiles could be predicted simply by diffusion of the (constant) source concentrations. In simulations where precipitation of stable solids was allowed, uraninite and calcium uranate accumulated at the source-transport domain interface, while coffinite penetrated further into the transport domain. Total uranium concentrations in regions of precipitation were determined by solubilities of the precipitated solids, and were six to seven orders of magnitude lower than those in the simulations without precipitation, throughout the domain of transport.

The Analytical Repository Source-Term (AREST) code has been developed for source-term evaluation of spent fuel as a final waste form in geologic repositories. AREST contains a set of analytical equations for the timedependent diffusional mass transport of both solubility-limited and inventory-limited radionuclides from a spent fuel in a failed container surrounded by a shell of packing or other porous material imbedded in a porous host rock. Three factors that affect release performance are examined: 1) congruent dissolution of the UO2 matrix, 2) chemical instability of the UO2 matrix, with precipitation of a more stable uranium phase within the waste package, and 3) the attenuation of release rate by distribution of containment failures with time.

For congruent matrix dissolution, the release rates of included nuclides are proportional to the product of solubility-limited release of uranium and the fractional abundance of the nuclide. For certain conditions, congruent release rates are calculated to be up to 10 orders of magnitude lower than release rates assuming individual solubility-limits. Precipitation of a more stable, lower solubility uranium phase within the waste package is shown to increase release rates from the UO2 matrix compared to the non-precipitation case, in agreement with previous calculations. During the first 300 to 1000 years after repository closure, the distribution of containment failures with time will act to attenuate the peak average release rates of soluble, longlived nuclides, such as iodine-129, to values smaller than release rates below regulatory limits. However, for soluble nuclides with short half-lives, such as cesium-137, a broader distribution of containment failure with constant mean time of failure can actually cause an increase In the peak average release rates.

The environmental assessment of the geological disposal of used nuclear fuel involves the assimilation of experimental results to produce an estimate of the potential impact of a disposal vault on man and the environment. In this paper we briefly outline the assessment methodology and expand upon one aspect of the data assimilation process, namely, the incorporation of research results on the performance of engineered barriers into assessment models.

The model to be discussed represents (i) failure modes for titaniumbased containers, including short-term failures due to undetected manufacturing defects and long-term failures due to uniform and local corrosion; (ii) release of radionuclides from used fuel, including relatively fast release of soluble fission products from gap and grain boundaries and slow, congruent release controlled by the dissolution of the fuel matrix itself; and (iii) mass transport of released radionuclides through the clay-based sealing materials surrounding the waste container, including diffusion, convection and radionuclide storage (sorption or retardation) effects.

Sensitivity surfaces have been developed to analyse the performance of engineered barriers of an underground nuclear fuel waste disposal vault. The major engineered barriers are the waste form, the containers and the buffer surrounding the containers. A modular computer code has been developed that calculates the release rate from each barrier. The modules are linked together through the use of time-series-management routines. By systematic variation of the model input parameters, sensitivity surfaces are constructed that illustrate the behaviour of an objective function such as maximum fractional release rate in response to the changes in the input parameters.

As part of the Nevada Nuclear Waste Storage Investigations (NNWSI) Project, a comprehensive geochemical/geophysical model is being compiled. This model incorporates the current and relevant stratigraphic, petrologic, hydrogeologic, geochemical, and material data associated with a candidate repository at Yucca Mountain, Nevada. In this report the known repository data are compiled and unknown parameter values are estimated based on the available data. More data are needed before the geochemical/geophysical model of Yucca Mountain can be regarded as satisfactory and a suitable base for multidimensional predictive flow and transport simulations. Recommendations for future studies concerning site characterization and data aquisition are presented.

Laboratory tests were conducted to validate a radionuclide mass transfer model. During the experiments, cesium and uranium releases from a simulated waste form embedded in an isotropic medium of quartz sand were measured. A 0.01 M NaH2PO4 buffer solution flowed past the waste form. Downstream concentrations obtained from computer simulations were compared with experimentally measured concentrations of cesium and uranium. Uranium release was found to be controlled by solubility-limited mass transfer, while cesium release was controlled by waste form dissolution kinetics. It was also found that the effects of dissolution on groundwater chemistry must be coupled with solubility-limited mass transfer models to defensibly predict radionuclide release rates under realistic repository conditions.

A code(PER8MIGR) for the finite element method (F.E.M.) using 8-node point isoparametric elements has been developed for solving three-dimensional groundwater flow and advection diffusion equations with a decay chain of radioactive materials. Galerkin's method is applied to discretize the equations, and asymmetric band matrix method is used to solve the equation. Some simple problems for groundwater flow and migration of nuclides were solved by the code to compare the results for groundwater flow with the one- and three-dimensional analytical solutions. The results for migration of nuclides were also compared with the results obtained by some existing available codes. Our results and those by the other codes were in good agreement. The present work makes it possible to solve the groundwater flow and diffusion problems more accurately for radioactive waste disposal in complex geometry.

A mathematical model of waste dissolution that includes the formation of secondary solid precipitate(s) explicitly is formulated. For a static type of experiment, the model indicates that the concentration of the highly soluble species compared to the concentration of the main component of the matrix is equal to their corresponding density in the matrix (i.e., congruent release). For low solubility nuclides, the solubility limit of precipitate(s) may be exceeded (depending on the waste composition) and a supersaturated solution is temporarily achieved. From the model, the concentration of radionuclides and mass flux at the waste form surface in a geologic repository is obtained. Numerical results (Sr-90) indicate that the liquid-surface concentration rises very rapidly after dissolution starts. In the first week, the solubility limit of Sr precipitate(s) is exceeded and Sr solid precipitate(s) will be formed. Subsequently the concentration decreases and becomes almost equal to the solubility limit and remains so for up to 1 year. It decreases later due to radioactive decay. After the first week, the surface concentration and surface mass flux are independent of the retardation coefficient throughout the leaching process. The mass transfer rate is determined by the matrix dissolution rate and not by the solubility of the precipitate.

Equipment for sampling and analyses of groundwater has been developed and operated. The results from the chemical characterization is used to group the waters into different classes mainly with respect to origin and residence time. The most important process affecting the chemistry of the groundwater is the carbon dioxide weathering.

Cohassett basalt was reacted with synthetic Hanford groundwater in a Dickson rocking autoclave at 300°C and 30 MPa for ca. 7600 hours. During the experiment fresh solution was twice reinjected into the reaction cell, raising the nominal water:rock mass ratio from 50:1 to 140:1. The reinjection experiment is intermediate between closed-system Dickson experiments and flow-through tests in that the fluid and solids can remain in contact for extended periods of time before the solution is replaced. This allows more time for equilibrium to be approached. Data from this experiment suggest:

(1) After reinjection many solution parameters quickly (hrs-days) returned to near preinjection values. (2) The redox buffer capacity of the basalt was not exceeded, i.e. fO2 values remained near magnetite-hematite values, although the nominal water:rock mass ratio was raised to ca. 140 during the experiment. (3) After reinjection the stable high-temperature pH value was only slightly less than the initial pH value. (4) The silica concentration stabilized near apparent quartz saturation rather than the cristobalite saturation value found in closed-system experiments. (5) Short-term relationships between cations appear to be controlled by ionexchange between the solutions and clay minerals.

Reaction products identified from the experiment include: Fesmectite, illite, hematite, minor cristobalite and possibly Ti-maghemite.

The radiation chemistry of nitrogen/oxygen/water systems is reviewed. General radiolytic effects in dry nitrogen/oxygen systems are relatively well characterized. Irradiation results in the formation of steady state concentrations of ozone, nitrous oxide and nitrogen dioxide. In closed systems, the concentration observed depends on the total dose, temperature and initial gas composition. Only three studies have been published that focus on the radiation chemistry of nitrogen/oxygen/water homogeneous gas systems. Mixed phase work that is relevant to the gaseous system is also summarized. The presence of water vapor results in the formation of nitric acid and significantly changes the chemistry observed in dry air systems. Mechanistic evidence from the studies reviewed are summarized and discussed in relation to characterizing the gas phase during the containment period of a repository in tuff.

Two different types of experiments have shown that alpha and gamma radiation each increase the redox potential of salt brines. Other experiments have shown that the decrease in potential indicated by platinum electrodes in irradiated brines is due to the response of the electrodes to the hydrogen generated from the resulting water decomposition. Platinum electrodes indicated no change in redox potential of solutions containing equimolar concentrations of K3Fe(CN)5 and K4Fe(CN)5 when they were gamma irradiated in quartz containers; when stainless steel containers were used, the solutions reacted with the stainless steel but gamma radiation had no significant effect on that reaction or the solution composition. A gamma radiation dose rate of 2.4 × 104 rad/h measurably increased the uniform corrosion rate of mild steel in salt brines at 75°C, but at 150°C the increased thermal corrosion rate obscured any effect of even higher gamma radiation dose rates.

Because of its ability to retard the movement of radionuclides, sodium bentonite is being considered as a packing and backfilling material in a number of nuclear waste isolation programs. Few studies, however, have considered how the chemical properties of the bentonite backfill will vary with distance from the canister and change in time as groundwater constituents diffuse through the bentonite. Of particular importance is the conversion of sodium bentonite to calcium bentonite by ion exchange, which will in turn affect the migration of cationic radionuclides, such as Cs+ and Sr2+, leaking from a waste canister.

Experiments measuring the movement of trace amounts of radionuclides through compacted bentonite have typically used unaltered bentonite. Models based on experiments such as these may not lead to accurate predictions of the migration through altered or partially altered bentonite of radionuclides that undergo ion exchange. To address this problem, we have modified an existing transport code to include ion exchange and aqueous complexation reactions. The code is thus able to simulate the diffusion of major ionic groundwater components through bentonite and reactions between the bentonite and groundwater. Numerical simulations have been made to investigate the conversion of sodium bentonite to calcium bentonite for a reference groundwater characteristic of deep granitic formations.

Perturbations to the natural chemistry of the geology surrounding a nuclear waste repository may affect the overall nuclide release rates through properties such as their solubilities and sorption characteristics. In this paper, we develop some of the necessary techniques to construct a detailed research model of the evolution of the groundwater under the influence of the backfill material. These techniques involve various extensions to the coupled ionic migration and chemical equilibria code, CHEQMATE and are used in the first stages of such a model. Although the design and location of a deep low- and intermediate-level nuclear waste repository have not yet been finalised in the U.K., we have considered a concrete backfill material embedded in a clay geology. Preliminary results suggest that over timescales of about a thousand years the natural buffering action of the clay against changes in pH has a significant effect on the scale of perturbation by the ingress of highly alkaline porewater. It seems likely therefore that this type of modelling will have considerable relevance to the safety assessment models.

Experiments have been conducted by the Basalt Waste Isolation Project (BWIP) to assess changes in solution chemistry in the near-field environment of a nuclear waste repository in basalt. These Dickson autoclave experiments were carried out using Grande Ronde basalt ± bentonite and synthetic groundwater or deionized water at 300°C, 30 MPa, and solution-to-solid mass ratio of 10 for up to two years. Groundwater solution changes during reaction of the basalt and basalt/bentonite included initial decreases in pH and sodium concentration presumably due to smectite formation. This initial trend subsequently reversed in the basalt system with pH rising to ca. 7.5 and sodium increasing to the starting value. Steady state pH values for the basalt/bentonite system were ca. 6.4. The basalt + deionized water test exhibited a constant rise in pH to ca. 7.9 and release of sodium to solution in response to basalt dissolution. Slightly oxidizing conditions characterized the early part of all of the experiments followed by a decrease in fO2 to 10−31 to 10−32 These results are consistent with other work at similar and lower temperatures, suggesting that the packing material will react in the waste package environment to produce slightly alkaline, reducing conditions.