The U.S. Department of Energy (DOE) has released a Viability Assessment (VA) of the proposed high-level waste (HLW) repository at Yucca Mountain (YM), Nevada. The proposed standard for the repository requires the evaluation of the predicted —over 10,000 yr— performance of the system. In the VA, it is argued that long waste package (WP) lifetimes are responsible for significant containment and delay in release of the HLW. This paper examines the DOE WP lifetime prediction by combining the VA parameters modeling the corrosion behavior of the WP materials with the repository performance assessment code developed by the U.S. Nuclear Regulatory Commission (NRC) and the Center for Nuclear Waste Regulatory Analyses (CNWRA). The objective of the analysis is to exercise and strengthen the NRC review capabilities in preparation for receipt of the DOE license application for the proposed HLW repository. The results indicate that during the first 10,000 yr, radionuclide release depends almost entirely on the fraction of initially defective waste packages (IDWP). Radionuclide release resulting from corrosion of the WPs occurs only beyond 10,000 yr. The long WP lifetimes (>10,000 yr), reported in the VA and in this study, primarily depend on corrosion rates developed from an expert elicitation consensus. Given the importance in the estimation of the onset of radionuclide release and the lifetime of the WP materials, it is suggested the need for a stronger technical basis to assess the fraction of IDWP and to support the low corrosion rates of the WP materials.

Transuranic nuclear wastes being disposed of in the Waste Isolation Pilot Plant (WIPP) contain large amounts of organic material that may decay producing substantial quantities of undesirable CO2. Because of this possibility a MgO backfill has been included in the repository design. In addition to scavenging CO2, the backfill may also hydrate forming Mg(OH)2. Backfill hydration may provide a sink that may sorb much, or all, of the brine that enters the repository. The key to understanding the full implications of including the MgO backfill lies in knowing the rates of the various processes that may operate as the repository ages and evolves. Both carbonation and hydration reactions were found to occur rapidly enough to have beneficial impacts on repository performance.

Monitoring the vadose zone below buried waste provides an early warning of contaminate Advanced Tensiometer transport toward the groundwater. To quantify the transport mechanisms, vadose zone hydraulic characteristics and the physical variables need to be obtained. We have designed and implemented a Vadose Zone Monitoring System (VZMS) to monitor or sample the 3 state variables of the vadose zone, water potential, water content and chemical concentration. The state variables are monitored using an Advanced Tensiometer (AT), a borehole water content sensor (BWCS) and a vacuum lysimeter, respectively. This system was installed at the Savannah River Site (SRS) E-Area disposal site, where low level wastes have been disposed of in shallow trenches. The system has operated for The system has operated for several months providing nearly continuous water content and water potential data. The vacuum lysimeters were activated on a quarterly schedule. Installation details and an example data set are presented to illustrate the effectiveness of the VZMS, and demonstrate the utility of the VZMS as an indicator of contaminant transport.

The objective of this study is to optimize the storage system with respect to thermal constraints. Due to a large number of 3D calculations, a simple temperature model based on an Design Of Experiment method has been developed. The most significant parameters have been exhibited by an extensive parametric study. A satisfactory agreement is observed between polynomial model and 3D approach.

Bentonite clay has been proposed as buffer material in several concepts for nuclear high level waste repositories. The “Long Term Test of Buffer Material” (LOT) series aims at validating models and hypotheses concerning physical properties in a bentonite buffer material and of related processes regarding mineralogy, microbiology, cation transport, copper corrosion and gas transport under conditions similar to those in a the Swedish KBS3 repository.

The test series comprises 7 test parcels, which will be run for 1, 5 and 20 years. The testing principle is to place parcels containing heater, central copper tube, pre-compacted clay buffer, instruments, and parameter controlling equipment in vertical bore-holes in granitic rock. The parcels are equipped with heaters in order to simulate the decay power from spent nuclear fuel at standard KBS3 conditions (90°C) and adverse condition (130°C). Adverse conditions in this context refer also to high temperature gradients over the buffer, and additional accessory minerals leading to i.a. high pH and high potassium concentration in clay pore water. Temperature, total pressure, water pressure and water content are measured during the heating period. At test termination standard chemical and mineralogical analyses, and physical testing are made. The 2 pilot tests (1-year tests) have been completed and preliminary results concerning the bentonite analyses and tests are presented. The 5 test parcels in the main test series have been started during the fall, 1999.

The fate of O2 in a granitic repository has been addressed by an international project: The redox experiment in detailed scale (REX). The emphasis of the project was on a field experiment involving groundwater in contact with a fracture surface. To this aim a borehole, ≍20 cm in diameter, was drilled at 380 m depth in the tunnel of the Äspö Hard Rock Laboratory, Sweden. Injection pulses of molecular oxygen were performed at in situ temperature and pressure. Several microbial and chemical parameters were studied as a function of time: microbial counts, pH, O2-concentration, Eh (redox potential), etc.

The field study has been supported by laboratory experiments to determine O2 reaction rates and mechanisms. These laboratory studies have been performed with Äspö samples (both for inorganic and microbially mediated processes). A replica experiment has also been completed at CEA, France, with the other half of the fracture surface obtained in the drilling procedure of the field experiment. The aim of the replica experiment has been to duplicate as far as possible the conditions of the REX in situ experiment, for example by using groundwater sampled at the REX site in Sweden, shipped in special containers to France.

The data that has been collected from the O2 injection pulses in the REX field and replica experiments have been compared with the rates of O2 uptake determined in the laboratory experiments. These data allow an estimate of the life-times for oxygen uptake in fractures in granitoids, which is of consequence for performance assessment calculations.

A mathematical model for bentonite expansion through a narrow planar space has been developed based on Terzaghi's theory for clay deformation due to water intrusion. The bentonite expands in a radial direction through a horizontal planar gap, with a constant aperture, filled with water. The permeability and the compressibility of the bentonite are assumed functions of its void ratio. The resulting governing equation is a non-linear diffusion-like equation with void-ratio-dependent coefficients. Numerical solutions for the space-time-dependent void ratio in the expanding bentonite are obtained by applying the Finite Element Method. The finite element solution is combined with a predictor-corrector scheme for evaluations of the void ratio distribution and the location of the moving bentonite-tip boundary. A computer code has been developed for the numerical solutions. The numerical scheme is supported by comparing the results with an analytical solution.

Data from diffusion experiments using Na+, Ca2+, Sr2+, Cs+ and Ba2+ as tracers in a crystalline rock type (Äspö diorite, originating from the Äspö Hard Rock Laboratory) have been evaluated using a heterogeneity model. The model concept consists of diffusion in a variety of different channels with different porosities. The porosity distribution used has been obtained from independently performed measurements of porosity distributions using the 14C-polymethylmethacrylate impregnation method. Breakthrough curves in through-diffusion experiments as well as penetration profiles in the matrix have been evaluated using the porosity distributions. In the calculations only two parameters, the pore diffusivity (Dp) and the sorption distribution coefficient (Kd) have been varied in order to fit the experimental data to the proposed model. For the penetration profile of more strongly sorbing tracers, i.e., Cs+ and Ba2+, a significantly better explanation of the data is obtained using a heterogeneity model compared to using a uniform porosity distribution model. The data from the through-diffusion experiments gives a better explanation of the shape at the beginning of the breakthrough curve. The implication of the proposed diffusion model is discussed, both from an in situ sorption experiment application and a performance assessment application.

Reactive transport simulations are being used to evaluate the nature and extent of radionuclide contamination within alluvium surrounding an underground nuclear test at the Nevada Test Site (NTS). Simulations are focused on determining the abundance and chemical nature of radionuclides that are introduced into groundwater, as well as the rate and extent of radionuclide migration and reaction in groundwater surrounding the working point of the test. Transport simulations based upon a streamline-based numerical model are used to illustrate the nature of radionuclide elution out of the near-field environment and illustrate the conceptual modeling process. The numerical approach allowed for relatively complex flow and chemical reactions to be considered in a computationally efficient manner. The results are particularly sensitive to the rate of melt glass dissolution, distribution of reactive minerals in the alluvium, and overall groundwater flow configuration. They provide a rational basis from which defensible migration assessments can proceed.

The characteristics and abundance of reactive surfaces in aquifer media have long been recognized as key factors controlling the migration of contaminants and other dissolved constituents in groundwater. The authors have shown previously that the effective reactive surface area of a heterogeneous aquifer is a complex function of groundwater advective velocity and the correlation structures of the physical and chemical heterogeneities. Although in principle, the available surface area within an aquifer could be estimated using geostatistical techniques and laboratory BET surface area determination for individual samples, this approach is fraught with difficulties associated with inadequate sample coverage and lack of appropriate methods for scaling laboratory surface area measurements. An alternative approach (multiple-tracer approach) to the difficult process of estimating available aquifer surface area relies on the use of conservative and slightly reactive tracers to assess the integrated effective reactive surface area along flow paths. This multiple-tracer approach sacrifices detailed understanding of the fine-scale heterogeneity but can provide integrated large-scale estimates of effective reactive surface area useful for the prediction of reactive transport. The approach is demonstrated by the analysis of the breakthrough curves for paired tracers in heterogeneous media.

Uranium and thorium-series disequilibrium in nature permits the determination of many in-situ physico-chemical, geologic and hydrologic variables that control the long-term migration of radionuclides in geologic systems. It also provides site-specific, natural analog information valuable to the assessment of geologic disposal of nuclear wastes. In this study, a model that relates the decay-series radioisotope distributions among solution, sorbed and solid phases in water-rock systems to processes of water transport, sorption-desorption, dissolution-precipitation, radioactive ingrowth-decay, and α recoil is discussed and applied to a basaltic aquifer at the Idaho National Engineering and Environmental Laboratory (INEEL), Idaho.

A number of Hanford tanks have leaked high level radioactive wastes (HLW) into the surrounding unconsolidated sediments. The disequilibrium between atmospheric CO2or silica-rich soils and the highly caustic (pH > 13) fluids is a driving force for numerous reactions. Hazardous dissolved components such as 133Cs, 79Se, 99Tc may be adsorbed or sequestered by alteration phases, or released in the vadose zone for further transport by surface water. Additionally, it is likely that precipitation and alteration reactions will change the soil permeability and consequently the fluid flow path in the sediments. In order to ascertain the location and mobility/immobility of the radionuclides from leaked solutions within the vadose zone, we are currently studying the chemical reactions between: 1) tank simulant solutions and Hanford soil fill minerals; and 2) tank simulant solutions and CO2.

We are investigating soil-solution reactions at: 1) elevated temperatures (60 - 200 °C) to simulate reactions which occur immediately adjacent a radiogenically heated tank; and 2) ambient temperature (25 °C) to simulate reactions which take place further from the tanks. Our studies show that reactions at elevated temperature result in dissolution of silicate minerals and precipitation of zeolitic phases. At 25 °C, silicate dissolution is not significant except where smectite clays are involved. However, at this temperature CO2 uptake by the solution results in precipitation of Al(OH)3 (bayerite). In these studies, radionuclide analogues (Cs, Se and Re – for Tc) were partially removed from the test solutions both during high-temperature fluid-soil interactions and during room temperature bayerite precipitation. Altered soils would permanently retain a fraction of the Cs but essentially all of the Se and Re would be released once the plume was past and normal groundwater came in contact with the contaminated soil. Bayerite, however, will retain significant amounts of all three radionuclides.

Performance assessment calculations by the U.S. Department of Energy and the Nuclear Regulatory Commission indicate that Tc-99 is a major contributor to dose to a hypothetical receptor group 20 km downgradient of a proposed high-level nuclear waste repository at Yucca Mountain, Nevada, within the first 10,000 yr after permanent closure. This result is due in large part to the high solubility and low retardation of Tc under oxidizing conditions in the Yucca Mountain environment. Recent site characterization data on the chemistry of saturated zone groundwater at Yucca Mountain and vicinity indicate the presence of locally reducing geochemical conditions, which could decrease the solubility and enhance the sorption and retardation of Tc-99. In this study, a preliminary assessment of the potential effects of reducing conditions on the transport and release of Tc-99 was conducted. Sensitivity analyses using the NRC/CNWRA Total-system Performance Assessment code (TPA Version 3.2) indicate that decreased Tc solubility and increased Tc sorption due to reduction of Tc(7+) to Tc(4+) can significantly delay the arrival of Tc-99 at the receptor group location. Decreased Tc solubility can decrease the Tc-99 dose by three orders of magnitude relative to the TPA 3.2 base case. Enhanced Tc retardation in the tuff aquifer only does not greatly decrease the calculated Tc-99 peak dose, whereas increased Tc retardation in the alluvial aquifer alone prevents Tc-99 from reaching the receptor group in 50,000 yr. The release and transport of other redox-sensitive radioelements could be affected in a manner similar to Tc. Thus, reduced groundwater conditions could significantly enhance the performance of the geologic barrier system and reduce the dose to the receptor group.

Analytical pore water compositions from Yucca Mountain were evaluated for internal thermodynamic consistency. Significant ionic charge imbalances, unequilibrated aqueous speciation relationships, and erratic variability with depth in some species concentrations were found. Thermodynamic consistency was restored by introducing measured CO2 gas pressure as a constraint, imposing equilibrium aqueous speciation, and adjusting pH to achieve charge balance. Reinterpreted water chemistry data were used to evaluate and interpret vertical and lateral variations in water chemistry, differences between unsaturated zone pore and perched water compositions, and water-rock equilibria.

Calculations of the consequences of the geological disposal of solid radioactive waste require models that apply a simplified description of the evolution of a repository. In work carried out to date for United Kingdom Nirex Limited (Nirex), models of repository performance have assumed that the contents of the repository are homogeneous. There is no direct representation in the models, for example, of the fact that all the waste will be contained initially within waste packages, with backfill surrounding the packages.

It is important that the assumption of a homogeneous repository is demonstrated to be cautious (i.e. the calculations of radiological consequence, based on a homogeneous model, should err in the direction of slightly overestimating any release of radionuclides). This paper introduces models that simulate the impact of specific heterogeneities on radionuclide release. These have been constructed and applied, in order to assess the appropriateness of the assumption of homogeneity.

Calculations using the NAMMU computer program have examined the release of radionuclides from within waste packaging. These calculations suggest that recent performance assessments overestimate the release of short-lived radionuclides. Calculations, including the use of the INHOMOG computer program, have examined the relative placement of different categories of waste packages within repository vaults (for the Nirex disposal concept). It is concluded from the calculations that, for a few radionuclides, consideration of the distribution of some waste packages may be appropriate in future performance assessment studies.

Bentonite may be used, in conjunction with cementitious materials, in repository designs for TRU wastes in Japan. Therefore, possible effects of the interaction of highly alkaline water with bentonite are of interest. In this study, the sorption of plutonium and uranium has been measured on a bentonite sample that had been altered by hydrothermal exposure to a high pH solution.

Samples of Kunipia-F® bentonite were exposed to a pH 14 solution of sodium, potassium and calcium hydroxides for 20 days at 130°C, 160°C and 200°C under air. X-Ray analysis showed that alteration at 160°C was similar to alteration at 200°C.

Batch sorption experiments were carried out at room temperature under nitrogen by contacting 200°C-altered bentonite with synthetic equilibrated water containing both uranium and plutonium in the presence of sodium dithionite as a reducing agent. The experiments were equilibrated for three months and then sampled after centrifugation, 0.45 gm filtration, and 10,000 nominal molecular weight cut-off (NMWCO) filtration. Uranium and plutonium remaining in solution were determined. The final redox potentials were between -500 to -51OmV vs SHE and the final pH values were 12.2.

Sorption of both actinides showed some dependence on the solid:liquid separation method. The average RD value for uranium sorption after centrifugation was 2 × 103cm3g−1 which increased to 3 × 104 cm3g−1 after 0.45 µm filtration. However, the effect of a further decrease in the pore size of the filter had little additional effect; 10,000 NMWCO filtration gave an average RD of 5 × 104 cm3g−1. The sorption of plutonium after solid:liquid separation by centrifugation was very similar to uranium with an average RD of 6 × 103 cm3g−1. After 0.45 µm filtration the increase in RD to 2 × 105 cm3g−1 was more marked than for uranium. There was no significant increase in RD value after 10,000 NMWCO filtration compared to 0.45 μm filtration.

Suspensions of colloids obtained from a Spanish bentonite were studied by potentiometric acid-base titrations and electrophoresis in order to analyse their surface chemical properties, which are responsible of their stability behaviour and that are very important to consider in radionuclides sorption modelling.

“Fast” titrations and “batch - back” titrations techniques were used to determine the contribution of the pH - dependent charge and the difference in the results obtained are discussed. Experimental data obtained by acid/base titrations were interpreted according the EDL theory. The model prediction agreed satisfactorily with the experimental data in the alkaline pH range. Protonation / deprotonation reactions of surface functional groups (SOH) appeared to be the main surface charge - determining mechanism in the alkaline pH range whereas ion - exchange type reactions, had to be taken into account over the acidic pH range. Surface potentials were calculated for different salt concentrations, from experimental data and taking into account both layer and edge sites charge contributions.

In a high-level waste repository, a carbon steel overpack will corrode after the repository is closed. This will create a reducing environment in the vicinity of the repository. Reducing conditions are expected to retard the migration of redox-sensitive radionuclides such as plutonium.

The apparent diffusion coefficients of plutonium were measured in compacted bentonites (Kunigel V1® and Kunipia F®, Japan) in contact with carbon steel and its corrosion products under a reducing condition and, for comparison, without carbon steel under an oxidizing condition. Gas bubbles were observed in some bentonite specimens which had low dry densities after contact with carbon steel for approximately two years. This observation suggests hydrogen generation during corrosion of the carbon steel. The apparent diffusion coefficients measured were approximately 10−14 m2/s under the reducing condition and less than 1015 m2s under the oxidizing condition. There was a significant effect of redox conditions on the apparent diffusion coefficients. The effects of dry density (0.8 to 2.0 Mg/m3) and montmorillonite contents (50% for Kunigel V1 or 100% for Kunipia F), however, were not observed clearly. The chemical species of plutonium were expected to be PuOH2+ for the reducing condition and Pu(OH)4 for the oxidizing condition, respectively.

Effective diffusion coefficients(De) in bentonite were measured as a function of ionic charge to evaluate the degree of surface diffusion and anion exclusion. The De measurements for Ni2+, Sm3+ and Se32− were carried out for 1.8Mg•m−3 by through-diffusion method. Sodium bentonite, Kunigel-V1® was used. The order of obtained De values was Sm3+>Ni2+>SeO32−. These De values were compared with those reported to date. Consequently, the order of De values was Cs+>Sm3+>HTO>Ni2+>anions(I−, Cl−, CO32−, SeO32−, TcO4−, NpO2CO3−, UO2(CO3)34−), showing a tendency of cations>HTO>anions. The reason that the De of Ni2+ was lower than that of HTO may be because the free water diffusion coefficient(Do) of Ni2+ is about 1/3 of that of HTO. The formation factors(FF) were in the order, Sm3+>Cs+>Ni2+>HTO>anions, indicating a possibility of surface diffusion in cations and of anion exclusion in anions. In this case, the FF of Sm3+ was approximately 5 times greater than that of HTO. However, since the Do of Sm3+ is about 1/3 of that of HTO, the De of Sm3+ may have been a little higher than that of HTO. Based on this, it is presumed that surface diffusive effect on De in bentonite is insignificant.

Proton titration experiments have been conducted at the Idaho National Engineering and Environmental Laboratory on synthetic goethite and soil in an effort to develop adsorption parameters that will help predict migration of radioactive liquid waste. This is the initial step in a reactive transport project to understand contaminant migration in a system characterized by strong chemical gradients. For this stage, two levels of pretreatment were applied to the soil to remove carbonate minerals and soluble salts to focus on the remaining mineral fraction. Without some sort of treatment or conditioning, native soil has a large buffer capacity that interferes with proton titration experiments. In this report, results are presented from the initial stages of the project.