Field experimental and analytical methods development is underway to define the hydraulic and transport properties of a thick saturated zone that underlies the planned high-level nuclear waste repository at Yucca Mountain, Nevada. The characterization strategy for the highly heterogeneous hydrology is that of hypothesis testing and confidence building. Three test wells, the UE-25c-holes, have been drilled and preliminary data have been collected. Hydro-mechanical analyses indicate formation fluid at depth is hydraulically connected to the water table. Preliminary hydraulic tests indicate highly localized, fracture-controlled transmissivity. Cross-hole seismic tomography is planned to assess the inter-borehole structure of fractures and faults. Multi-level cross-hole hydraulic interference and tracer tests are planned using up to 5 packed-off zones in each of the c-holes to assess the hydraulic conductivity and transport structure in a crude tomographic fashion. An equivalent discontinuum model conditioned with the observed hydraulic measurements will be applied to interpret the hydraulic test responses. As an approach to the scale problem the tests will be designed and analyzed to examine the hypothesis that the flow system may be represented by fractal geometry. 12 refs., 4 figs.

Performance assessment of the engineered barrier system for a nuclear waste repository combines information from relevant disciplines and predicts the net long-term performance of the EBS in unites of regulatory goals for performance. The performance assessment models are specific to the proposed Yucca Mountain, Nevada site. Early assessments are used for project planning and feedback. The EBS scenarios activity develops the scenarios and the consequent event sequences. Initial model development for single waste packages indicates that the radionuclide release rate performance is sensitive to the water flux, element solubilities, and/or the mode of water contact with the waste. The latter in turn depends on local scale hydrology and the modes of corrosion for the container material. For the release rate summed over waste packages, variations among waste packages and their near-field environments are anticipated. These variations place demands on data acquisition and modeling, as well as modulate the impact of localized changes of conditions. Sampling in uncertainty assessment is a subsequent step in examining the reliability of predictions made in the performance assessments. Advances made in sampling methods are referenced. 14 refs., 6 figs.

The importance of paleohydrology to the Yucca Mountain Site Characterization Project derives from the role water will play in radioactive-waste repository performance. Changes in hydrologic conditions during the lifetime of the repository may be estimated by investigating past hydrologic variations, including changes in the static water-level position. Based on the distribution of vitric and zeolitized tuffs and the structural history of the site, the highest water levels were reached and receded downward 11.6 to 12.8 myr ago. Since that time, the water level at central Yucca Mountain has probably not risen more than about 60 m above its present position. The history of the high potentiometric gradient running through northern Yucca Mountain may be partly elucidated by the study of tridymite distribution in rocks that have experienced saturated conditions for varying periods of time.

Many large simulations may be required to assess the performance of Yucca Mountain as a possible site for the nations first high level nuclear waste repository. A boundary integral equation method (BIEM) is described for numerical analysis of quasilinear steady unsaturated flow in homogeneous material. The applicability of the exponential model for the dependence of hydraulic conductivity on pressure head is discussed briefly. This constitutive assumption is at the heart of the quasilinear transformation. Materials which display a wide distribution in pore-size are described reasonably well by the exponential. For materials with a narrow range in pore-size, the exponential is suitable over more limited ranges in pressure head. The numerical implementation of the BIEM is used to investigate the infiltration from a strip source to a water table. The net infiltration of moisture into a finite-depth layer is well-described by results for a semi-infinite layer if {alpha}D > 4, where {alpha} is the sorptive number and D is the depth to the water table. the distribution of moisture exhibits a similar dependence on {alpha}D. 11 refs., 4 figs.,

To investigate the effects of heating on the mineralogical, geochemical, and mechanical properties of rock by high-level radioactive waste, cores are being examined from holes penetrating locations where electric heaters simulated the presence of a waste canister, and from holes penetration natural hydrothermal systems. Results to date indicate the localized mobility and deposition of uranium in an open fracture in heated granitic rock, the mobility of U in a breccia zone in an active hydrothermal system in tuff, and the presence of U in relatively high concentration in fracture-lining material in tuff. Mechanical -- property studies indicate that differences in compressional- and shear-wave parameters between heated and less heated rock can be attributed to differences in the density of microcracks. Emphasis has shifted from initial studies of granitic rock at Stripa, Sweden to current investigations of welded tuff at the Nevada Test Site. 7 refs., 8 figs.

The US Department of Energy is responsible for the design, construction, operation, and decommission of a site for the deep geologic disposal of high-level radioactive waste (HLW). This involves site characterization and the use of performance assessment to demonstrate compliance with regulations for HLW disposal from the US Environmental Protection Agency (EPA) and the US Nuclear Regulatory Commission. The EPA standard states that a performance assessment should consider the associated uncertainties involved in estimating cumulative release of radionuclides to the accessible environment. To date, the majority of the efforts in uncertainty analysis have been directed toward data and parameter uncertainty, whereas little effort has been made to treat model uncertainty. Model uncertainty includes conceptual model uncertainty, mathematical model uncertainty, and any uncertainties derived from implementing the mathematical model in a computer code. Currently there is no systematic approach that is designed to address the uncertainty in conceptual models. The purpose of this investigation is to take a first step at addressing conceptual model uncertainty. This will be accomplished by assessing the relative impact of alternative conceptual models on the integrated release of radionuclides to the accessible environment for an HLW repository site located in unsaturated, fractured tuff. 4 refs., 2 …

Cation-ratio dating rock varnish is a recently developed technique for obtaining surface exposure ages of a wide variety of geomorphic surfaces. As originally proposed, the technique utilizes a ratio among minor cations [(K + Ca)/Ti] in rock varnish. Although this varnish cation ratio is related to the Ti concentration, it can also be affected by the presence of Ba that may be partially included in the analyzed concentration of Ti. During energy-dispersive spectroscopy (EDS), Ba L-alpha and L-beta peaks overlap with Ti K-alpha and K-beta peaks. We have compared the effect of Ba concentration on calculated varnish cation ratios using quantitative EDS with the scanning electron microscope (SEM), quantitative wavelength-dispersive spectroscopy (WDS) with a Cameca electron probe microanalyzer (EPM) and qualitative EDS with the SEM that does not decompose Ti and Ba lines. In this paper we document that, in fact, such separation of Ba from Ti is possible using both a quantitative (MICRO Q) and a semi-quantitative (SQ) Tracor Northern EDS analytical programs that decompose Ti and Ba lines. We also document that SEM EDS analyses may yield markedly similar results to wavelength dispersive spectrometer (WDS) analyses of the same varnish using an EPM. 6 refs., 6 figs., 2 …

Radioactive waste disposal in the unsaturated zone and its subsequent migration and decay calls for prediction with the aid of transport models. Usually, such models consist of a system of partial differential equations for the concentration C (function of space and time) that has to be solved with appropriate initial and boundary conditions. These equations are based on physical first principles, but contain a number of parameters that have to be determined experimentally. Hence, experiments are crucial for validating the equations and their solution and for determining parameter values. The Los Alamos National Laboratory (LANL) has initiated and performed a series of experiments at an intermediate scale, in caissons much larger than laboratory columns, of depth comparable with natural soil, but of limited horizontal extent. These experiments were conducted with both nonreactive and reactive solutes, and are described in a series of reports. Their main advantage is in expanding the scale on one hand, while permitting careful control and measurements on the other. 8 refs., 2 figs.

This paper presents preliminary transport calculations for radionuclide movement at Yucca Mountain. Several different realizations of spatially distributed sorption coefficients are used to study the sensitivity of radionuclide migration. These sorption coefficients are assumed to be functions of the mineralogic assemblages of the underlying rock. The simulations were run with TRACRN{sup 1}, a finite-difference porous flow and radionuclide transport code developed for the Yucca Mountain Project. Approximately 30,000 nodes are used to represent the unsaturated and saturated zones underlying the repository in three dimensions. Transport calculations for a representative radionuclide cation, {sup 135}Cs, and anion, {sup 99}Tc, are presented. Calculations such as these will be used to study the effectiveness of the site`s geochemical barriers at a mechanistic level and to help guide the geochemical site characterization program. The preliminary calculations should be viewed as a demonstration of the modeling methodology rather than as a study of the effectiveness of the geochemical barriers. The model provides a method for examining the integration of flow scenarios with transport and retardation processes as currently understood for the site. The effects on transport of many of the processes thought to be active at Yucca Mountain may be examined using this approach. 11 refs., …

The Total-System Analyzer is a modular computer program for probabilistic total-system performance calculations. The code employs stratified random sampling from model parameter distribution functions to generate multiple realizations of the system. The results of flow and transport calculations for each realization are combined into a probability distribution function of the system response as indicated by the performance measure. We give a detailed description of the code and present results for four example problems simulating the release of radionuclides from a proposed high-level-waste repository at Yucca Mountain, Nevada. The example simulations illustrate the impact of significant variation of percolation flux and sorption on radionuclide releases. We discuss the effects of numerical sampling error and of correlations among the model parameters. 20 refs., 7 figs., 2 tabs.

The TOUGH code developed at Lawrence Berkeley Laboratory (LBL) is being extensively used to numerically simulate the thermal and hydrologic environment around nuclear waste packages in the unsaturated zone for the Yucca Mountain Project. At the Lawrence Livermore National Laboratory (LLNL) we have rewritten approximately 80 percent of the TOUGH code to increase its speed and incorporate new options. The geometry of many requires large numbers of computational elements in order to realistically model detailed physical phenomena, and, as a result, large amounts of computer time are needed. In order to increase the speed of the code we have incorporated fast linear equation solvers, vectorization of substantial portions of code, improved automatic time stepping, and implementation of table look-up for the steam table properties. These enhancements have increased the speed of the code for typical problems by a factor of 20 on the Cray 2 computer. In addition to the increase in computational efficiency we have added several options: vapor pressure lowering; equivalent continuum treatment of fractures; energy and material volumetric, mass and flux accounting; and Stefan-Boltzmann radiative heat transfer. 5 refs.

The geometry of the asperities of contact between the two surfaces of a fracture and of the adjacent void spaces determines fluid flow through a fracture and the mechanical deformation across a fracture. Heuristically we have developed a stratified continuum percolation model to describe this geometry based on a fractal construction that includes scale invariance and correlation of void apertures. Deformation under stress is analyzed using conservation of rock volume to correct for asperity interpenetration. Single phase flow is analyzed using a critical path along which the principal resistance is a result of laminar flow across the critical neck in this path. Results show that flow decreases with apparent aperture raised to a variable power greater than cubic, as is observed in flow experiments on natural fractures. For two phases, flow of the non-wetting phase is likewise governed by the critical neck along the critical path of largest aperture but flow of the wetting phase is governed by tortuosity. 17 refs., 10 figs.

When an unsaturated porous medium is subjected to a temperature gradient and the temperature is sufficiently high, vadose water is heated and vaporizes. Vapor flows under its pressure gradient towards colder regions where it condenses. Vaporization and condensation produce a liquid saturation gradient, creating a capillary pressure gradient inside the porous medium. Condensate flows towards the hot end under the influence of a capillary pressure gradient. This is a heat pipe in an unsaturated porous medium. We study analytically the transport of gaseous species released from a spent-fuel waste package, as affected by a time-dependent heat pipe in an unsaturated rock. For parameter values typical of a potential repository in partially saturated fractured tuff at Yucca Mountain, we found that a heat pipe develops shortly after waste is buried, and the heat-pipe`s spatial extent is time-dependent. Water vapor movements produced by the heat pipe can significantly affect the migration of gaseous radionuclides. 12 refs., 6 figs., 1 tab.

Construction of well-posed scenarios for the range of conditions possible at any proposed repository site is a critical first step to assessing total system performance. Event tree construction is the method that is being used to develop potential failure scenarios for the proposed nuclear waste repository at Yucca Mountain. An event tree begins with an initial event or condition. Subsequent events are listed in a sequence, leading eventually to release of radionuclides to the accessible environment. Ensuring the validity of the scenarios requires iteration between problems constructed using scenarios contained in the event tree sequence, experimental results, and numerical analyses. Details not adequately captured within the tree initially may become more apparent as a result of analyses. To illustrate this process, the authors discuss the iterations used to develop numerical analyses for PACE-90 (Performance Assessment Calculational Exercises) using basaltic igneous activity and human-intrusion event trees.

This study examines the roughness profiles and aperture distributions of fractures and faults by using concepts from fractal geometry. Simple models of flow of fluid in rough fractures are also discussed. A deterministic fractal representation of the roughness profile is presented which is shown to have many distinct advantages over other numerical methods, such as information compression, uniqueness and repeatability of surface simulation, retention of statistical information, and self-similarity over many scales. Also the fractal representation enables an isotropic surface and an aperture distribution to be simulated by examining a measured profile. Saturated fluid flow in fractures is then computed using a combined Navier-Stokes and Darcy equation. 14 refs., 5 figs.

In the proposed Yucca Mountain repository, packages of high-level waste are separated from surrounding unsaturated rock by a 2-cm air-gap annulus. If a waste package contacts the surrounding rock, if the annulus becomes filled with rock and rubble, or if the rock becomes saturated, there can exist pathways for release of dissolved radionuclides by liquid diffusion. We have applied the analytical solutions for time-dependent mass transfer of dissolved species through porous rock to predict the time-dependent release rates of key radionuclides. The expected ground-water velocities are so low that mass-transfer is predicted to occur predominantly by molecular diffusion in pore liquid. Conservatively, we neglect mass-transfer resistance from fuel cladding and failed metal containers. 6 refs., 3 figs.

The basis for identifying site-specific data needs for the Yucca Mountain site was a thorough review and evaluation of the regulatory requirements for a geologic repository. The US Nuclear Regulatory Commission`s (NRC) regulation governing disposal of high-level waste in a geologic repository utilizes a multiple barrier approach that includes reliance on both engineered and natural elements. Given an understanding of the regulatory requirements, the next step was to identify the natural features of the site and the components of the engineered barriers considered most likely to make significant contributions to isolation of waste at Yucca Mountain. The intent was to develop a site testing program that focused on understanding those natural site characteristics and processes most likely to contribute to waste isolation; and providing site data necessary to design and assess the performance of engineered barriers. 3 refs., 3 figs.

The potential nuclear waste repository at Yucca Mountain is to be in partially saturated rock. Released radioactive gases such as {sup 14}CO{sub 2} could have a direct pathway to the biosphere. We study the release of {sup 14}C released as {sup 14}CO{sub 2} from partly failed nuclear waste containers by analyzing the flow of gas into and out of a container. We analyze the transport of released {sup 14}CO{sub 2} in an unsaturated, fractured, porous medium with gas-phase advection and dispersion. We calculate the inhalation dose to a maximally exposed individual above ground, based on some conservative assumptions about release from containers. For the assumed parameter values, a simple atmospheric diffusion model gives very small doses when compared to background radiation doses. 12 refs., 4 figs., 1 tab.

A waste package consisting of a container and high-level nuclear waste is being developed for the permanent disposal of radioactive waste. Yucca Mountain, Nevada, is being studied as a potential site for the underground high-level nuclear waste repository. A major consideration for choosing Yucca Mountain is the presence of zeolite in tertiary ash-flow tuffs. The presence of zeolites could provide geological barriers to radionuclide migration. The suitability of the tuffaceous rocks at Yucca Mountain for the repository is being investigated since the properties of the environment around a waste site must be well characterized to reliably predict performance. The results of experiments at Lawrence Livermore National Laboratory (LLNL) to assess the possibility of imaging water in Nevada Test Site welded tuff samples showed that nuclear magnetic resonance imaging is not viable. This leaves neutron tomography and high-frequency electromagnetic geotomography as possibilities for the practical imaging of distribution and flow of fluids in rock, including tuff specimens. Water tracers are needed in electromagnetic tomography techniques since the contrast for detecting water in cracks of tuff is lower than in granite because of the higher porosity in tuff. The results of preliminary testing with geotomography by LLNL indicates relatively low spatial resolution. …

This study presents a generic methodology that can be used to evaluate the uncertainty in the calculated accidental offsite doses at the Yucca Mountain repository during the preclosure period. For demonstration purposes, this methodology is applied to two specific accident scenarios: the first involves a crane dropping an open container with consolidated fuel rods, the second involves container failure during emplacement or removal operations. The uncertainties of thirteen parameters are quantified by various types of probability distributions. The Latin Hypercube Sampling method is used to evaluate the uncertainty of the offsite dose. For the crane-drop scenario with concurrent filter failure, the doses due to the release of airborne fuel particles are calculated to be 0.019, 0.32, and 2.8 rem at confidence levels of 10%, 50%, and 90%, respectively. For the container failure scenario with concurrent filter failure, the 90% confidence-level dose is 0.21 rem. 20 refs., 4 figs., 3 tabs.

Corrosion resistant materials are being considered for the metallic barrier of the Yucca Mountain Project`s high-level radioactive waste disposal containers. Nickel-chromium-molybdenum alloys and titanium alloys have good corrosion resistance properties and are considered good candidates for the metallic barrier. The localized corrosion phenomena, pitting and crevice corrosion, are considered as potentially limiting for the barrier lifetime. An understanding of the mechanisms of localized corrosion and of how various parameters affect it will be necessary for adequate performance assessment of candidate container materials. Examples of some of the concerns involving localized corrosion are discussed. The effects of various parameters, such as temperature and concentration of halide species, on localized corrosion are given. In addition, concerns about aging of the protective oxide layer in the expected service temperature range (50 to 250{degree}C) are presented. Also some mechanistic considerations of localized corrosion are given. 45 refs., 1 tab.

Yucca Mountain is being studied to determine its suitability as a location for a high-level nuclear waste repository. Amter and Ross developed a model called TGIF (Topographic Induced Flow) to simulate gas flow under Yucca Mountain. The TGIF model differs significantly from previous gas flow models. It uses a governing equation that is based on the concept of freshwater head, thus avoiding the numerical problems associated with the near-cancellation of the forces due to gravity and the pressure gradient. Unlike most other models, dipping, layered media can be simulated. This paper describes a systematic sensitivity study that was designed to test several aspects of the TGIF model when used to simulate gas flow under Yucca Mountain. Values of three important inputs to the model were systematically varied to form a matrix of 80 runs. The matrix consisted of five values of permeability contrast between a bedded tuff layer and surrounding welded units (in all cases, bulk permeabilities were used to represent the combined effect of both fractures and matrix permeability), four temperature profiles representing different stages of repository cooldown, and four finite-difference grids.

There is an immediate need to begin to validate models that can be used for assessing the performance of waste package materials in an unsaturated repository environment. This paper examines available testing information and testing approaches that could support validation of models for engineering barrier system (EBS) radionuclide release. The content is presented in the context of the general methodology that has been proposed for validating performance assessment models. Available experimental observations are used to test some of the EBS release rate modeling premises. These observations include evidence of fluid film formation on waste glass surfaces in isothermal humid environments, accelerated waste glass reaction rates under repository service conditions of large glass surface area to water volume ratio, and mobilization of radionuclides as solutes and colloids. It is concluded that some important modeling premises may not be consistent with available experimental information. However, it is also concluded that future laboratory testing, which simulates the integrated waste package systems, is needed to evaluate the significance of these inconsistencies and to test the system level models. A small-scale apparatus which was developed and tested to examine the feasibility of laboratory analog testing for the unsaturated Yucca Mountain repository environment is described. 16 …

The US Department of Energy (DOE) is committed to keeping the citizens of Nevada informed about activities that relate to the high-level nuclear waste repository program. This paper presents an overview of the Yucca Mountain Project`s public interaction philosophy, objectives, activities and experiences during the two years since Congress directed the DOE to conduct site characterization activities only for the Yucca Mountain site.