Development of the Conceptual Model of Unsaturated Zone Hydrology at Yucca Mountain, Nevada
Yucca Mountain is an arid site proposed for consideration as the nation's first underground high-level radioactive waste repository. Low rainfall and a thick unsaturated zone are important physical attributes of the site because the quantity of water likely to reach the waste and the paths and rates of movement of the water to the saturated zone under likely future climates will be major factors in estimating the concentrations and times of arrival of radionuclides at the surrounding accessible environment. The framework for understanding the hydrologic processes that occur at this site and that control how quickly water will penetrate through the unsaturated zone to the water table has evolved during the past 15 years. Early conceptual models assumed that very small volumes of water infiltrated into the bedrock, that much of the infiltrated water flowed laterally within the upper nonwelded units because of capillary barrier effects, and that the remaining water flowed down faults with a small amount flowing through the matrix of the lower welded, fractured rocks. When evidence accumulated indicating that infiltration rates were higher than initially estimated, and that mechanisms supporting lateral diversion did not apply at these higher fluxes, the flux calculated in the lower welded unit exceeded the conductivity of the matrix. This required water to flow vertically in the high-permeability fractures of the potential repository host rock.
U.S. Geological Survey, Sacramento, California
Lawrence Berkeley National Laboratory, Berkeley, California
Los Alamos National Laboratory, Los Alamos, New Mexico
The development of numerical modeling methods evolved concurrently with the conceptual model in order to account for the observations made at the site, particularly fracture flow deep in the unsaturated zone. This paper presents the history of the evolution of conceptual models of hydrology and numerical models of unsaturated zone flow at Yucca Mountain, Nevada.
On-land geologic disposal of high-level nuclear waste has been an issue in the United States for nearly half a century. In 1958, the National Academy of Sciences recommended considering geologic disposal of high-level nuclear waste (HLW). In 1959, concerns about the thermal effects of nuclear-waste disposal were added to the recommendation. In the early 1970s, Winograd (1972, 1974) proposed storing nuclear waste in the unsaturated zone, although it was not until the early 1980s that such a design was seriously considered. In 1976, the Director of the U.S. Geological Survey (USGS) suggested to the U.S. Energy Research and Development Administration [ERDA, the predecessor to the U.S. Department of Energy (DOE)] that a nuclear test site in Nevada [Nevada Test Site (NTS)] be examined for potential sites for HLW disposal. The major attributes of the NTS as a potential site for disposal are that it is in a remote location, it is a large contiguous block of land under federal ownership, there is much information on the unsaturated zone based on studies of the underground nuclear testing and the associated presence of radionuclides in the subsurface, rainfall is low, and there is a thick unsaturated zone with a variety of rock types (Winograd, 1971). Initially, however, use of the saturated zone as a nuclear waste repository was the prevailing choice (Roseboom, 1983; Hanks et al., 1999). By 1978, the first boreholes were being drilled on Yucca Mountain to explore the character of the saturated zone for disposal of nuclear waste. The high fracture transmissivity and elevated groundwater temperature of the saturated zone below Yucca Mountain made this zone undesirable as a repository site. In 1982, USGS scientists suggested to DOE that the unsaturated zone at Yucca Mountain be considered instead. Because Nuclear Regulatory Commission (NRC) draft regulations 10 CFR 60 “Disposal of High Level Waste in Geologic Repositories,” published in 1981, covered only repositories in the saturated zone, the USGS also suggested to the NRC that the regulations be modified to include the unsaturated zone, and Roseboom (1983) pointed out how such a repository would differ from one in the saturated zone. After extensive public comment and review, the final version of 10 CFR 60 that included the unsaturated zone was released in 1985.
The purpose of this paper is to describe and trace the evolution of the conceptual model of groundwater flow in the unsaturated zone at Yucca Mountain. For this discussion, a conceptual model is simply a relevant set of concepts that describe, in a qualitative way, the behavior of a natural system. Numerical models of the same system, which also will be discussed in this paper, are based on
the same set of concepts but describe the behavior of the system in a quantitative manner. It is important to note that the history of the characterization of Yucca Mountain or, in particular, the evolution of a conceptual model of groundwater flow at the Yucca Mountain site, cannot be accurately reconstructed solely on the basis of citable literature. To fully understand this history requires reference to unpublished or draft reports, memoranda, and rough notes. In addition, many of the concepts for the model were developed during discussions between DOE and such entities as the Nuclear Waste Technical Review Board (NWTRB), the Advisory Committee on Nuclear Waste (ACNW), or NRC technical interchanges. In many cases, ideas were developed and worked out during informal get-togethers and, as such, many important ideas and information used in the development are not readily available or directly citable.
YUCCA MOUNTAIN SITE DESCRIPTION
Yucca Mountain is located in southern Nevada about 145 km northwest of Las Vegas (Figure 2-1). The study area covers approximately 45 km2, of which approximately 5 km2 covers the potential repository site. Beneath the crest of Yucca Mountain, the water table ranges from approximately 350-750 m below land surface, with an average of 500 m. The potential repository host rock is the Topopah Spring Tuff of the Paintbrush Group, a densely welded and fractured tuff located in the unsaturated zone at an average depth of 300 m below land surface (Hanks et al., 1999).
Climate and Precipitation
An understanding of the response of the hydrologic system to current climatic conditions is a prerequisite for predicting the response of the system to potential future climatic conditions (Botkin et al., 1991). The climate in the Yucca Mountain area is arid to semiarid. Weather patterns vary seasonally. Summer precipitation comes primarily from the south and southeast. Winter winds bring moisture from the west, and hence the climate is subject to a regional rain shadow east of the Sierra Nevada and has been for the entire geologic history of Yucca Mountain, more than 13 million years. Topographic effects cause substantial variability in average annual precipitation in the Yucca Mountain area. Precipitation averages from less than 130 mm for lower elevation locations in the south to more than 280 mm for higher elevation locations in the north, with an estimate of 170 mm directly over the potential repository location (Hevesi and Flint, 1996).
Yucca Mountain is located within the Basin and Range physiographic province (Grayson, 1993). The linear mountains and valleys of this area that have a
distinct north to northwest trend define the Basin and Range physiography. Within the Basin and Range physiographic province, there are several topographic regions. Yucca Mountain is in the Death Valley region, which has the largest and most prominent desert basin in the Basin and Range physiographic province. The Death Valley region is primarily in the northern Mojave Desert; the region extends northward into the Great Basin Desert and lies in the rain shadow of the Sierra Nevada. Death Valley itself is the ground-water discharge area for a large part of the Death Valley region. The Death Valley region is composed largely of closed topographic basins that apparently coincide with several closed shallow groundwater flow systems (Winograd and Thordarson, 1975). Recharge in these systems is sparse, and is derived mostly from the higher altitudes and comes as infiltration of precipitation or the infiltration of ephemeral runoff. Discharge occurs primarily by spring flow and by evaporation and transpiration of shallow ground water from playas. The deepest part of the saturated flow system consists of extensive Paleozoic carbonate aquifers that connect the closed shallow groundwater systems at depth. Discharge from the system occurs in several intermediate areas that are geomorphically, stratigraphically, and structurally controlled; but ultimately, most groundwater flow discharges to Death Valley. The predominant direction of drainage for surface-water and groundwater flow in the Death Valley region is generally from north to south because of a decrease in the average altitude from north to south in the southern Basin and Range area.
Yucca Mountain consists of a 1-3-km-thick sequence of ash flow and ash fall tuffs erupted from Timber Mountain, a source caldera complex located directly to the north. The unsaturated zone at Yucca Mountain is about 500-750 m thick (Snyder and Carr 1982; Buesch et al., 1996), characterized by pyroclastic flows that consist of separate formations. From youngest to oldest, the formations are the Rainier Mesa Tuff (11.6 million years) of the Timber Mountain Group; the Tiva Canyon, Yucca Mountain, Pah Canyon, and Topopah Spring Tuffs of the Paintbrush Group (12.7 million years); the Calico Hills Formation (12.9 million years); and the Prow Pass, Bullfrog, and Tram Tuffs of the Crater Flat Group (13.5 million years) (Carr et al., 1986; Sawyer et al., 1994) (Figure 2-2). Interstratified with these formations are bedded tuffs that consist primarily of fallout tephra deposits and small amounts of pyroclastic flow deposits and reworked material (Moyer and Geslin, 1995; Buesch et al., 1996). The bottom and top of the Tiva Canyon and Topopah Spring Tuffs contain vitric, nonwelded to densely welded tuff; the interiors of the tuffs are thick, crystallized, and moderately to densely welded and fractured. Most of the lithostratigraphic units in the Tiva Canyon and Topopah Spring Tuffs are laterally continuous and stratiform (Scott and Bonk, 1984). The Yucca Mountain and Pah Canyon Tuffs are relatively thick to the north of the potential repository location near Yucca Wash and contain
both nonwelded and welded intervals. The welded intervals of these units, however, thin southward starting near Drill Hole Wash, and, therefore, only thin welded intervals occur in the center of the potential repository location. These welded intervals are absent altogether from the southern half of the repository location (Moyer et al., 1996). The nonwelded tuffs of the Paintbrush Group, including the nonwelded intervals of the Yucca Mountain and Pah Canyon Tuffs, the interstratified bedded tuffs, the nonwelded base of the Tiva Canyon Tuff, and the nonwelded top of the Topopah Spring Tuff collectively are commonly referred to as the Paintbrush nonwelded hydrologic unit (PTn). The Calico Hills Formation (CHn) is composed of nonwelded pyroclastic flow and fallout deposits (Moyer and Geslin, 1995). Tuffaceous rocks have been zeolitized (CHz) at the north end of Yucca Mountain, yet parts of the formation remain largely vitric
(CHv) towards the south end of the mountain. The Prow Pass Tuff is a compound cooling unit and consists of nonwelded to partially welded tuff at the top and bottom with intervals of welded tuff. The vitric parts of this unit are typically zeolitized in the north, but only in the southwestern part of Yucca Mountain does a significant part of this unit remain vitric, with partially to moderately welded, crystallized tuff in the interior of the unit.
The hydrology of Yucca Mountain has largely been influenced by interrelationships between tectonic and geomorphic processes. Faults and fault scarps, and erosional processes on the eastern sloping ridge, have defined the topography of the mountain, and have created a series of washes (Figure 2-1) that are downcut to varying degrees into different bedrock layers. The topography generally is controlled by high-angle faults that tilt the resistant volcanic strata eastward. Locally, slopes are steep on the west-facing escarpments of the Solitario Canyon Fault and in some of the valleys that cut into the more gentle eastward-facing dip slopes. Narrow valleys and ravines have been cut into the bedrock. Floors of wider valleys consist of alluvial deposits that have formed terraces into which intermittent streams have cut channels. Locally, small sandy fans flank the lower slopes and spread out on the valley floors. East of the crest of Yucca Mountain, drainage is into Fortymile Wash; west of the crest of the mountain, streams flow southwestward down fault-controlled canyons and discharge in Crater Flat. The study site area can be divided into two parts north and south of Drill Hole Wash. The washes in the southern area trend eastward, are relatively short (less than 2 km), and are defined by erosional channels that produce gently sloping sideslopes. The washes north of Drill Hole Wash are controlled by faults, are northwest trending, and are approximately 3-4 km long with steep sideslopes.
Alluvial deposits in the valley floors and washes include fluvial sediments and debris-flow deposits. Soil development and thickness of the alluvial deposits are variable, and the soils are gravelly in texture. The deposits range from 100 m thick in the valleys to less than 30 m thick in the mouths of the washes. Midway up the washes, most alluvial fill is less than 15 m deep in the center of the wash. Many of these deposits have developed cemented calcium carbonate layers (Flint and Flint, 1995).
DEVELOPMENT OF INITIAL CONCEPTUAL AND NUMERICAL MODELS (1983-1990)
A conceptual model describes the physical processes that are part of an environment, how they relate to each other, and which processes dominate the system. It describes the physical framework within which the processes can be understood and numerical relations can be developed.
Conceptual Model Issues at Yucca Mountain
Many current and historical issues are relevant to the discussion of the conceptual model of unsaturated zone hydrology at Yucca Mountain. We discuss the most significant issues and how the associated components of the conceptual model developed or were modified as the conceptual model changed. A simplified schematic that highlights these issues is presented in Figure 2-3.
In general, the major components of the conceptual model include the following processes and features: (1) surface infiltration rates and their distributions in space and time, (2) lateral flow in the nonwelded PTn, (3) lateral flow at the vitric-zeolitic interface in the matrix of the deep nonwelded tuffs (CHv and CHz), (4) the role of faults as conduits or barriers to flow, (5) the occurrence and stability of perched water, (6) the distribution and significance of fast pathways, and (7) the flux between fractures and matrix in unsaturated rock. Most conceptual models for Yucca Mountain include these components, but advances in our scientific understanding of these processes and features have greatly influenced the way the conceptual and numerical models have developed over the years.
Initial Data Collection
Initial data collection at Yucca Mountain consisted of mapping the bedrock surface and drilling boreholes to describe the geology and water table depths at the site. By 1986, more than 100 boreholes had been drilled at or near Yucca
Mountain. Many of the early holes drilled during the late 1970s extended to the water table, yielding important data such as potentiometric surfaces, groundwater chemistry, detailed lithostratigraphy, and matrix properties (Anderson, 1981; Rush et al., 1983; Weeks and Wilson, 1984; Whitfield et al., 1984; Flint and Flint, 1990). These were followed in 1984 by an extensive series of shallow boreholes that were drilled to investigate shallow infiltration processes (Hammermeister et al., 1985). Studies of the surface geology at Yucca Mountain had already been ongoing for more than 10 years (Byers et al., 1976; Scott et al., 1983; Scott and Bonk, 1984) prior to the time that investigations of infiltration and percolation processes began in earnest.
A stop-work order for most site characterization activities was issued by the DOE in early 1986 because of concerns related to the quality assurance of data collection; however, selected surface and laboratory investigations (those considered to be collecting irretrievable data) were allowed to continue in order to characterize the geology, faults, and matrix and fracture properties of Yucca Mountain (Klavetter and Peters, 1987; Istok et al., 1994; Flint et al., 1996b). After the stop-work order was lifted in late 1991, a series of shallow, cored neutron boreholes were drilled to study infiltration processes (Flint and Flint, 1995). On completion of the neutron boreholes, deep boreholes were drilled for long-term monitoring and geotechnical boreholes were drilled along the surface projection of the underground Exploratory Studies Facility (ESF) prior to its construction (Rousseau et al., 1998) to provide design information for the construction of the ESF. The results of core analysis and borehole instrumentation and geophysics, which measured subsurface conditions, have aided in the development of the conceptual model used to help understand infiltration and percolation rates and processes at Yucca Mountain. The measurements and analyses have provided detailed data sets needed for the development and testing of the site-scale numerical flow model.
Early Conceptual Models of Hydrology at Yucca Mountain
The earliest detailed conceptual model of the unsaturated zone at Yucca Mountain was published by Scott et al. (1983) (Figure 2-4). Their conceptual model of hydrology at Yucca Mountain is a component of a larger geologic/hydrologic framework model that is presented in a very straightforward manner. First, they identified the problem and stated that groundwater was one of the most critical parameters for nuclear waste isolation. Second, they described the stratigraphic, structural, and hydrologic framework that is the basis of their geologic/hydrologic framework model by presenting the detailed geologic setting. Third, they identified the relevant hydrologic processes needed to describe the hydrology for their geologic/hydrologic framework model. Finally, using these processes and applying them to their model, they described the hydrologic consequences of groundwater flow.
Although Scott et al. (1983) acknowledged that there were uncertainties in their estimates, their conceptual model of the hydrologic system was as follows: approximately 200 mm/yr precipitation falls on Yucca Mountain, of which about 3 percent enters the hydrologic system (6 mm/yr net infiltration). This infiltrating water moves vertically through fractures in the welded Tiva Canyon Tuff (TCw; in 1983 the accepted nomenclature was Tiva Canyon Member), vertically as matrix flow in the nonwelded PTn (although they believed some potential existed for lateral flow at the TCw-PTn interface), vertically through fractures in the welded Topopah Spring Tuff (TSw) and through the matrix in the lower vitric nonwelded tuffs, laterally at the vitric-zeolitic interface of the nonwelded tuffs, and vertically in lower zeolitic nonwelded tuffs. They also believed that lateral flow occurred along the tilted strata and that the potential existed for perched water to accumulate on the upgradient side of faults. Although they believed that faults could be either sealed or open to flow, they cautioned that faults should be assumed to be open in the absence of evidence to the contrary. This first and simple conceptual model is perhaps closer to the current (2000) conceptual model for Yucca Mountain than any of those formulated in the intervening years.
A somewhat more simplified conceptual model was proposed by Roseboom (1983) as a generic case for considering the unsaturated zone in the arid southwest for HLW disposal. As explained in the Introduction, the NRC's existing draft regulations assumed that all repositories would be in the saturated zone and the regulations would have to be changed to accommodate unsaturated sites. Thus, Roseboom presented the first diagrammatic representations of a repository in unsaturated rock (Figure 2-5) and pointed out fundamental differences between repositories in the saturated and the unsaturated zone. Although the model was generic for the most part, specific discussions in Roseboom's paper of Yucca Mountain presented a conceptual model similar to that of Scott et al. (1983). Assuming that precipitation is 127 mm/yr and that 3 percent of this precipitation becomes net infiltration, Roseboom estimated a recharge rate of 3.8 mm/yr to the local water table; he said that the flow path was downward, with no lateral component, and that flow was rapid through the fractured zones. The amount of flux through the repository, therefore, would be directly related to the average precipitation. Roseboom also believed that even higher infiltration rates for short-term climate change could be easily drained from the stratigraphic horizon in which the repository is located because of the high fracture permeability. Rapid drainage of water through fractures would minimize contact time between water and the waste canisters and therefore repositories in the unsaturated zone would be preferable to repositories in the saturated zone if the host rock had high fracture permeability.
In 1984, Montazer and Wilson (1984) presented a detailed conceptual model of Yucca Mountain (Figure 2-6) that was based on soil physics; this model was similar to the model of Scott et al. (1983) but with several exceptions. Montazer and Wilson estimated that the average annual precipitation at Yucca Mountain is
approximately 150 mm. Using the same 3 percent rate for net infiltration as Scott et al., they calculated a net infiltration of 4.5 mm/yr but suggested that recharge ranged from 0.0-4.5 mm/yr. They believed that water probably infiltrated either directly into fractures within bedrock exposures or as surface runoff seeping into alluvium beneath the channels of washes. As water saturated the walls of the fractures, it could continue flowing to greater depths without increasing the saturation of interior matrix blocks.
Despite their belief that complete matrix saturation was not necessary to induce and propagate fracture flow, they introduced a relation between effective (bulk) permeability and matrix potential for a combined fracture-matrix medium (composite porosity, see Figure 2-10, which is discussed later) that required high matrix saturations before fractures would conduct water. Figure 2-7 shows their analysis of flow through a single fracture (curve 2) transecting a porous matrix (curve 1) with much lower effective permeabilities. At very low levels of saturation, flow due to gravity or “drainage ” is limited to the matrix and follows curve 1 to lower levels of permeability and saturation. As the water content increases to higher levels of partial saturation, this curve crosses the curve for the fracture. At higher saturation levels, fracture flow dominates the drainage. Scott et al. (1983) suggest that with rapid wetting from periods of intense precipitation, less complete saturation of the matrix might occur. Curves 1a and 1b illustrate the “hysteresis effect ” of more rapid wetting, which effectively moves curve 1 to lower values because air becomes trapped in the matrix. This would reduce the permeability to lower values as the wetting rate increases. Thus with more rapid wetting, the fractures become more dominant in draining the rock mass. The combined uppermost parts of a fracture curve and a matrix curve can be combined to represent a single medium with composite porosity, and is termed an “equivalent continuum.”
Montazer and Wilson (1984) maintained that most of the infiltrating water (4 mm/yr) was diverted laterally within the Paintbrush nonwelded unit (PTn) and that the remaining water (0.5 mm/yr) moved downward through the matrix of the Topopah Spring welded unit (TSw). They cited work by Weeks and Wilson (1984), who estimated a downward flux of between 0.003 and 0.2 mm/yr in the matrix of the TSw, as support for the matrix flow component of their conceptual model. Montazer and Wilson suggested that lateral flow would lead to perched water on the upgradient side of faults, but that the water would ultimately move into the fault and recharge the saturated zone. Noting the sharp contrast in pore size distributions of the TSw compared with those of the underlying nonwelded and bedded vitric tuffs, they believed that the vitric tuffs with their much larger pores would act as a capillary barrier to vertical flow, leading to lateral flow in the TSw, whereas zeolitic tuff, which has much smaller pores, would not divert flow. This perspective differs from the one held by Scott et al. (1983), who proposed that the zeolitic boundary would be a barrier because of the low conductivity of this unit. The most significant differences between the conceptual model by
Montazer and Wilson compared with the earlier model by Scott et al. were a lack of fracture flow in the TSw and the prevalence of lateral flow above and within the PTn. The views of Montazer and Wilson greatly influenced the direction of subsequent hydrologic research at Yucca Mountain for several years.
In a general review of the hydrology in and around Yucca Mountain, Waddell et al. (1984) suggested that recharge to the water table probably was less than 5 mm/yr, although they also cited the work of Sass and Lachenbruch (1982), who estimated 8 mm/yr flux through the TSw on the basis of an analysis of temperature profiles (analyses that would be redone in the 1990s). Similar to Montazer and Wilson (1984), they believed that infiltration occurred predominantly in the washes or by direct entry into fractures exposed at the surface.
Early Numerical Process Models
Four major concepts would strongly influence further development of the conceptual model of Yucca Mountain and would control the thinking of most Yucca Mountain hydrologists and modelers for the next 10 years. These concepts were (1) that the matrix must be fully saturated before fracture flow could be initiated or sustained (Wang and Narasimhan, 1985), (2) that overall, flux is low (Scott et al., 1983; Montazer and Wilson, 1984), (3) that only matrix flow (i.e., fluxes less than 0.5 mm/yr) is assumed to occur in the TSw (Montazer and Wilson, 1984), and (4) that lateral diversion of most of the net infiltration occurs within or above the PTn.
Wang and Narasimhan (1985) presented a conceptual and a numerical theory of fracture flow and fracture-matrix interaction in fractured rock that showed that the matrix must be fully saturated before fracture flow could be initiated or sustained. They suggested that there is virtually no downward flow in fractures beneath the PTn and that flow between adjacent matrix blocks would occur across fractures only at asperities (Figure 2-8). They proposed that fracture flow occurs only when the water potential of the fracture is in equilibrium with the water potential of the matrix, which, in the unsaturated zone, occurs only during periods of surface flooding when there is a transition between saturated and unsaturated fracture conditions. They numerically modeled flow through a matrix system with discrete fractures and suggested that, for the most part, unsaturated fractured systems could be simulated numerically without taking fractures explicitly into account.
Reports by Sinnock et al. (1984, 1987), Klavetter and Peters (1986), and Peters and Klavetter (1988) continued to support the assumption that net infiltration was low (4 mm/yr), that lateral movement occurred at the base of the PTn, and that although no perched water was expected in the PTn, most of the water flowed downward through the faults, with less than 0.5 mm/yr of flux flowing through the TSw. Klavetter and Peters (1986) and Peters and Klavetter (1988) agreed, on the basis of the work of Wang and Narasimhan (1985), that the matrix
must be saturated in order for fracture flow to occur. In the results of the first Total System Performance Assessment (TSPA) for the Yucca Mountain site, Sinnock et al. (1984) argued that distributed fracture flow would not occur if fluxes were less than 0.5 mm/yr and would result in a relatively dry TSw (less than −2 MPa) despite the fact that Weeks and Wilson (1984) had estimated water potentials to be between −0.06 and −0.26 MPa on the basis of experimental data. Sinnock et al. (1984, 1987), Klavetter and Peters (1986), and Peters and Klavetter (1988) based their numerical analyses on the conceptual model of the U.S. Department of Energy (1984) (Figure 2-9), which essentially was a simplified version of the Montazer and Wilson (1984) model.
The first numerical two-dimensional flow model of the unsaturated zone at Yucca Mountain was presented by Rulon et al. (1986) and supported Montazer and Wilson's (1984) concept of lateral flow in the PTn at low fluxes with as much
as 50 percent of infiltration being laterally diverted for fluxes of less than 1 mm/yr. However, at a higher flux of 4.5 mm/yr, the model predicted only about 1 mm/yr of lateral flow; this prediction was based partly on the assumed high permeability of the faults. Using the combined fracture-matrix characteristic curve (see Figure 2-7) from Montazer and Wilson (1984), which essentially was an equivalent continuum model, Rulon et al. (1986) defined a critical saturation above which fracture flow would dominate over matrix flow. [The shape of this curve was demonstrated experimentally by Peters et al. (1984).] The value for critical saturation varied depending on the assumed matrix properties. For their two sets of fracture-matrix simulations, the critical saturation was either 0.3 or near 1; the latter represented the concept of Wang and Narasimhan (1985).
Rockhold et al. (1990) produced the first three-dimensional numerical model, but only for a small areal extent around the potential repository location. They modeled fracture flow in the TCw but matrix flow in the TSw and the PTn, where the PTn diverted most of the vertical flow laterally.
Early TSPA models for Yucca Mountain investigated only one-dimensional columns (Barnard et al., 1992). Their analysis suggested there was less than a 1 percent chance that flux through the TSw was more than 3 mm/yr and an 80 percent chance that the flux was less than 1 mm/yr. During this same time frame, in an effort to evaluate the possibility of significant fracture flow in the TSw, Gauthier et al. (1992) developed a “weeps” model that represented only fractures; their model rapidly drained the fractured TSw and accommodated higher infiltration fluxes without causing significant lateral diversion in the PTn.
Altman et al. (1996) presented a general view of the conceptual models of flow between matrix and fractures and the resultant relative permeability curves (assuming Richards' equation-type flow) that are typically required by numerical models of flow through fractured rock. Figure 2-10 demonstrates the evolution of numerical modeling from the early composite porosity used by Rulon et al. (1986), but first presented by Montazer and Wilson (1984), to the more sophisticated dual-permeability models used currently; the figure gives common terminology for further discussion.
DEVELOPMENT OF CURRENT CONCEPTUAL AND NUMERICAL MODELS (1991-2000)
Three-Dimensional Site-Scale Numerical Model
The basis for the current site-scale model of the unsaturated zone at Yucca Mountain was established during meetings in 1991 between the USGS and Lawrence Berkeley National Laboratory (LBNL); these meetings resulted in the first three-dimensional numerical site-scale model (Wittwer et al., 1992). Although the model was designed to allow for the spatial distribution of infiltration and for discrete faults (Figure 2-11), it also allowed for the use of the equivalent
continuum concepts of Montazer and Wilson (1984). Initially, simulations using the three-dimensional model used a uniform (i.e., not spatially distributed) surface infiltration rate of 0.1 mm/yr. This model was used to explore the role of faults and the redistribution of moisture at depth (Wittwer et al., 1995).
Development of Conceptual Model of Spatially Distributed Infiltration
Hevesi et al. (1992) used a geostatistical approach to construct the first detailed map of precipitation for the Yucca Mountain site; they estimated an average annual precipitation of 170 mm/yr. This map was used with the Maxey-
Eakin method (Maxey and Eakin, 1950), which postulates a simple correlation between precipitation and recharge on a regional scale, to estimate the spatial distribution of recharge for the Yucca Mountain area and the Death Valley region. Recognizing that there is a large variability in the spatial and temporal distribution of precipitation in the study area, it was noted that spatial and temporal distributions of infiltration could be the most important factors influencing the distribution and rate of percolation in the unsaturated zone (U.S. Department of Energy, 1992).
With the ongoing development of a three-dimensional numerical site-scale model, it became clear that quantitative estimates of the spatial distribution of infiltration would be needed to define the upper boundary condition for the flow model. The flow model of Wittwer et al. (1992) had incorporated three topographic zones that served as the initial estimated zones of infiltration. The topographic zones were chosen on the basis of field observations and preliminary analyses of neutron borehole logs that indicated a qualitative correlation between topographic zones and depth of alluvium. Distinctly different changes in subsurface water content with time and depth of water penetration following precipitation characterized the different infiltration zones (Flint et al., 1993; Flint and Flint, 1994; Flint and Flint, 1995).
The initial approach taken to spatially distribute surface infiltration on a site scale is explained in Flint and Flint (1994). For that study, they relied on maps of surface bedrock geology and alluvium (Scott and Bonk, 1984) to distribute material properties and applied the Darcy flux approach to neutron log water-content data for shallow boreholes and core property data. According to this approach, the depth at which the water content does not change seasonally is assumed to represent steady-state moisture conditions, and the effective hydraulic conductivity at that moisture content was assumed to be equal to flux. A site-scale map was then prepared of the bedrock units at the steady-state depths. These units were classified into five representative categories for which water content and hydraulic properties were used to calculate a flux (Plate 1), which then was distributed among the site-scale grid blocks. Fluxes ranged from 0.02 mm/yr for areas exposed with TCw to 13.4 mm/yr for areas exposed with PTn, with a site-scale mean of 1.4 mm/yr. Fluxes were highest in the washes and along the exposure of the permeable PTn in the Solitario Canyon escarpment.
Hudson and Flint (1995) refined this approach using estimates from precipitation maps (Hevesi and Flint, 1996), soil thickness, topographic position, and calculations of flux in boreholes to develop correlation matrices and to statistically distribute infiltration for a larger area than the area of the site-scale model (Figure 2-12). Infiltration rates ranged from 0-45 mm/yr, with a mean of 11.6 mm/yr in the study area. They suggest that most of the infiltration probably enters the system on ridgetops and sideslopes where soil cover is thin.
Estimates of the spatial distribution of infiltration indicated that some areas of the site can have high rates of infiltration while other areas can have little to no
infiltration. Initial efforts to use the infiltration maps of Flint and Flint (1994) and Hudson and Flint (1995) in numerical models resulted in a saturated matrix and sublayers, which was inconsistent with measured saturation data (Wittwer et al., 1995; Ho et al., 1995) because of invalid coupled matrix-fracture assumptions and an equivalent continuum approach. In many areas of the models, fluxes exceeded the saturated conductivity of the matrix, which indicated the necessity to modify the models to allow for fracture flow in units with a low-permeability matrix.
The final iteration of approaches to estimate infiltration was done by incorporating a detailed conceptual model of infiltration (Flint et al., 1993; Flint and Flint, 1995) into a numerical model (Flint et al., 1996a). The numerical model was driven by stochastically modeled precipitation, modeled solar radiation and evapotranspiration, soil depth, and bulk bedrock properties and was calibrated to measurements of water content changes in neutron boreholes. The numerical model allowed for the investigation of infiltration under different climate scenarios. Under present-day climatic conditions, infiltration rates ranged from 0-80 mm/yr, with a mean of about 4.5 mm/yr across the general area of the repository (Plate 2). This model was a modified water-balance model (Flint et al., 1996a) and not based on Richards ' equation and, therefore, did not require fracturematrix equilibrium to initiate fracture flow.
Modeling with High Infiltration Rates: the Conversion to Fracture Flow
Lateral Diversion in the PTn
By 1995, most project hydrologists had accepted that the conceptual model of unsaturated zone flow must include fracture flow in the TSw and that numerical models must incorporate approaches other than the equivalent continuum model if they were to accommodate high fluxes through the TCw. However, some hydrologists still believed that lateral diversion in and above the PTn may reduce the volume of water that would penetrate the TSw. Although there was no significant or conclusive field evidence to support lateral diversion, a number of modeling exercises were done to test if the measured properties and geometry of the PTn could support or induce diversion (Ho, 1995; Altman et al., 1996; Moyer et al., 1996; Wilson, 1996). Quite typically in these simulations, lateral diversion could be attained but, consistent with an earlier study by Rulon et al. (1986), the percentage of water diverted was greatly reduced at high fluxes. At lower fluxes diverted water may have been an artifact of the model's simplified geometry, unrealistic hydrologic properties [extremely high air-entry pressure and high matrix permeability (Peters et al., 1984) are incompatible], idealized stratigraphic contacts as linear features or a misrepresentation of the gradational nature of the TCw-PTn transition (Moyer et al., 1996; Flint, 1998), or the downdip boundary conditions (e.g., the capillary properties assumed for the faults). In any case, on
the basis of capillary barrier theory (Montazer and Wilson, 1984; Ross, 1990) and modeling results, lateral diversion could not be counted on to divert high fluxes from the repository horizon.
Evidence of Fast Fracture Flow
During this period, Fabryka-Martin et al. (1993, 1994), Fabryka-Martin and Liu (1995), and Yang et al. (1996) reported bomb-pulse concentrations of isotopes in surface-based boreholes; these pulses indicated possible fast pathways through the TCw and deep into the TSw. Measurements of 36Cl and calculations of chloride mass balance supported the high rates of the spatially distributed estimates of infiltration and the existence of fracture flow. Secondary calcite deposits observed in fractures throughout the ESF indicated that considerable fracture flow had occurred (Paces et al., 1996). The results of several studies that had evaluated how water flows through fractures contradicted the traditional, but perhaps outdated, conceptual model of sheet flow in fractures (Rulon et al., 1986) and supported fast fracture flow by demonstrating channeling or fingering (Pruess and Tsang, 1990; Glass, 1993; Glass et al., 1994; Pruess, 1998). Nevertheless, modelers were still using either uniform low infiltration rates in their models (Wittwer et al., 1995) or spatially distributed rates that had been proportionally reduced to be consistent with previously low annual average infiltration values (Altman et al., 1996). Modelers justified their use of reduced distribution rates estimated from the infiltration maps of Hudson and Flint (1995) or Flint and Flint (1994) because their equivalent continuum models predicted unrealistic saturated matrix conditions and because they could not force enough water through the profiles to accommodate the high fluxes. Robinson et al. (1995) normalized the infiltration values from the Hudson and Flint (1995) map to a site-scale average of 1 mm/yr but noted that there was too much evidence for high infiltration rates and that the process modelers should now consider these higher rates. They proposed that nonequilibrium fracture flow could occur even if the matrix was not completely saturated. The method of Robinson et al. (1995) used the dualpermeability approach (Figure 2-10) to reduce the potential for saturation of the matrix under conditions of fracture flow by changing the parameters of the relative permeability curves of the fractures.
To reiterate the concepts that maintained long-term viability, it is pointed out here that the original proposed existence of nonequilibrium fracture flow came back full circle in the late 1990s to pre-Wang and Narasimhan (1985), who introduced the necessity of having a saturated matrix before fractures would flow. In the first conceptual model, Scott et al. (1983) suggested steady-state matrix saturation was necessary for fracture flow, but Montazer and Wilson (1984) maintained that a saturated matrix was not necessary to induce deep fracture flow, only wetted walls were required. Despite the recognition that fracture and matrix might not be in equilibrium, Montazer and Wilson (1984) sug
gested the use of the combined matrix-fracture moisture characteristic curve, or equivalent continuum approach, which necessitated high matrix saturations, except in cases of rapid wetting (Figure 2-7). This approach had survived for more than a decade.
Dual-Continuum Modeling Methods
Dual-permeability models (see Figure 2-10 for conceptual and numerical representations) were being used with the low infiltration rates estimated for Yucca Mountain in the early 1990s. To accommodate the high infiltration rates, Ho et al. (1995) decoupled fracture flow from matrix flow by reducing the simulated contact area between the fracture and matrix continua (dual-permeability model, Figure 2-10) by two orders of magnitude. Ho (1997) later suggested a reduction of at least four orders of magnitude in the contact area. This decoupling enabled the simulation of high rates of infiltration indicative of young isotopic ages and the fast transport of bomb-pulse nuclides that had been observed without causing the matrix to be unrealistically wet. In 1996, Wu et al. used the same technique presented by Robinson et al. (1995) to modify the equivalent continuum model by introducing different hydraulic properties for the fractures and the matrix. This was done to produce a dual-permeability-type approach that could accommodate 5 mm/yr of flux in two-dimensional cross sections without saturating the profiles, but also without decoupling the fracture-matrix interaction as Ho et al. (1995) had done.
Evidence supporting the high infiltration rates continued to accumulate. Analyses of thermal gradients in boreholes corroborated the estimated infiltration rates of 5-10 mm/yr for most of the repository (Bodvarsson and Bandurraga, 1996; Kwicklis and Rousseau, 1999). Estimates of 5-10 mm/yr also were calculated (Bodvarsson and Bandurraga, 1996) from analyses of perched water (Patterson, 1999).
Dual-continua modeling methods allowed the high fluxes to pass through the TSw, which, in turn, allowed simulations of young water in the CHn and in perched water. Model predictions of perched water at the vitric-zeolitic contact within the CHn or Prow Pass Tuff were supported by field observations (Figure 2-13a). An isopach surface (Figure 2-13b) of the top of the zeolitized units illustrates how water is envisioned as flowing across the top of the zeolites and into the water table either directly or through the faults (Bodvarsson and Bandurraga, 1996). Lateral flow at this contact also is supported by thermal gradient analyses, which indicate a decreased vertical percolation rate below this contact in the CHn and Prow Pass Tuff (Bodvarsson and Bandurraga, 1996). To account for the perched water, the site-scale model currently treats faults as permeability barriers to lateral flow, an assumption substantiated by pneumatic evidence (LeCain, 1997) and by the sharp drop in the potentiometric surface across the Solitario Canyon Fault (Tucci and Burkhardt, 1995). By 1996, the
conceptual model of hydrology for Yucca Mountain had evolved to include pervasive fracture flow in the TSw (Figure 2-14).
Development of Conceptual Model for Fast Pathways
Isotopic studies at the site began to support a conceptual model in which pervasive fracture flow and fast pathways through the unsaturated zone altered
the long-held conceptual model of flow in an arid environment (Winograd, 1981; Roseboom, 1983; Montazer and Wilson, 1984). Bomb-pulse concentrations of tritium and 36Cl and high 14C activities were detected in some of the several hundred unsaturated samples from boreholes and walls in the ESF (Yang et al., 1996, 1998; Fabryka-Martin et al., 1998). Tritium data also indicated that the groundwater was young at the top of the saturated zone in a fault zone, illustrating a mechanism for fast flow that may be important for local recharge. With the construction in 1995 of the ESF tunnel through the repository horizon, detailed in situ studies could be done to evaluate percolation flux through the TSw. An
unexpectedly large number of fast pathways were discovered using feature-based and systematic sampling of porewater to measure 36Cl (Figure 2-15), and a detailed conceptual model was developed to account for the distribution of the fast pathways (Fabryka-Martin et al., 1996, 1997; Sweetkind et al., 1997; Wolfsberg et al., 1999). Simulation results of this model indicate that the fastest pathways from the land surface to the repository horizon initiate as near-surface fracture flow beneath shallow soils that propagate along continuous fracture paths through the TCw, through faults in the PTn that allows water to bypass the rock matrix, and through continuous fracture paths in the TSw. Flow is required to remain in the fault only through the PTn. The concept of deep, rapid infiltration into fractured tuffs exposed at the surface or buried under shallow soils is supported by the presence of bomb-pulse 36Cl and tritium at depth, as well as by low concentrations of chloride, which reflect fairly high infiltration rates (Fabryka-Martin et al., 1998).
Isotopic studies of the porewater salts in the underground ESF support a conceptual model for fast pathways in which faulting in the overlying PTn is the dominant factor controlling the spatial distribution of fast pathways to the repository horizon (Fabryka-Martin et al., 1997). Three conditions must be present to transmit bomb-pulse 36Cl to the sampled depth (about 300 m below the surface) within 50 years. First, a continuous fracture path must extend from land surface, which requires the presence of faults that cut the PTn and increase its fracture conductivity and connectivity. Second, surface infiltration rates must be high enough to initiate and sustain at least a small component of fracture flow along the connected fracture path. Transport simulations indicate that the threshold rate
is probably about 1 to 2 mm/yr (Wolfsberg et al., 1999). Third, the residence time of water in the soil cover must be less than 50 years, i.e., the soil thickness must be less than 3 m so that the bomb pulse is not retained within the soil profile. The ESF studies also support the theory that the capacity of the unfaulted PTn to buffer and redistribute infiltration appears to decrease in the south of the repository where the unit thins to 25 m (Wolfsberg et al., 1999).
Current Conceptual Model of Flow Through the Unsaturated Zone at Yucca Mountain
The current (2000) conceptual model of flow through the unsaturated zone at Yucca Mountain (Bodvarsson et al., 1998; Flint et al., in press) as presented here reflects those processes invoked and supported by the majority of research participants on the Yucca Mountain Project, as well as those concepts that are most consistent with most of the measured data and observations. The existing sitescale numerical flow models effectively encompass and integrate hydrologic processes that operate at multiple scales at Yucca Mountain. The ability of these numerical models to provide simulations that are consistent with a broad suite of characterization data and observations collected independently is substantial corroboration of the viability of the conceptual model on which the numerical models are based. The numerical gridding has become more detailed for the area of the potential repository (Figure 2-16) and currently contains approximately 10,000 surface grid nodes that extend vertically through approximately 28 hydrogeologic unit layers from the land surface to the water table (Bodvarsson et al., 1998).
The four most important features of the current conceptual model are (1) the existence of relatively high spatially and temporally variable infiltration rates that virtually eliminate (2) large-scale lateral diversion of water above and within the PTn and that force (3) the pervasive flow of water through fractures in densely welded tuff units, despite nonequilibrium water potentials between fractures and the adjacent matrix, and (4) vertical flow in the CHv and extensive lateral flow and perching of water at the zeolitic boundary abutting faults. This simplified version of the conceptual model is very similar to the earliest conceptual models for Yucca Mountain, particularly that of Scott et al. (1983). These four features basically control most of the remaining details of the hydrologic processes represented schematically in Figure 2-17.
The infiltration processes are governed primarily by the distribution and timing of precipitation, the properties of the surface soils and bedrock, and the components controlling evapotranspiration. Average annual rates of infiltration range from 0 to more than 80 mm/yr and average approximately 5-10 mm/yr across the repository block area, or 3 to 6 percent of the average annual precipitation of 170 mm/yr. Most water that becomes net infiltration infiltrates from the ridgetops and sideslopes where fractures are present and soils are shallow to depths below the effects of evapotranspiration. Net infiltration is negligible in
deep soils because of the large storage component and evapotranspiration, except in the deep soils of channels fed by large volumes of runoff following extreme periods of precipitation.
Most of the infiltrating water passes quickly through the fractures of the TCw to be slowed during transition to matrix flow in the PTn except where
faults or broken zones disrupt the PTn, providing fast pathways for a small component of the flow. A small percentage of water in the TCw (less than 0.5 mm/yr) is lost to the atmosphere by way of upward vapor flow (barometric pumping; vapor diffusion; and convective, buoyancy-driven gas flow). Most flow is vertical and slow through the PTn matrix with possible local-scale lateral diversion just above the altered, nonwelded base of the Tiva Canyon Tuff, at linear contacts, or above low-permeability layers. Zones at the TCw-PTn and PTn-TSw contacts are nearly saturated but do not constitute perched layers. Water enters the TSw through faults or through localized broken zones at a low-conductivity vitrophyre at the top of the TSw and, to a lesser extent, through microfractures within the vitrophyre. The transition from highly porous tuffs to densely welded rocks occurs across a very short vertical distance, resulting in high saturations in this part of the flow system. Relatively pervasive broken-up areas that probably formed as the vitrophyre cooled provide ready access for entry of water into the underlying vapor-phase corroded nonlithophysal rocks of the TSw.
Once through the PTn, the bulk of the percolating water transitions back to vertical flow through the fractures and the matrix in the upper TSw and flows dominantly through fractures in the middle of the TSw. Fracture flow apparently occurs primarily under conditions of disequilibrium with the surrounding matrix
when averaged for relatively large matrix blocks and may occur by channeling or focused flow. In many locations, particularly in the northern part of the site, the densely welded basal vitrophyre of the TSw coincides with the vitric-zeolitic boundary and therefore serves as a permeability barrier to vertical flow, which results in perched water at this stratigraphic location. Where the vitric-zeolitic boundary does not extend upward as far as the base of the vitrophyre, perching occurs locally at the vitric-zeolitic contact in the CHn. It is surmised that the sloping alteration boundary promotes lateral flow within the perched layers in which transport velocities may be quite high. This mechanism functions similarly in the vitric-zeolitic contacts within the Prow Pass Tuff.
The role of faults in the deep unsaturated zone is not yet fully understood. In the shallow unsaturated zone, faults are highly permeable to air flow and may be major conduits for rapid water flow in the PTn. Fast flow in the TSw may be more dispersed because of its high fracture permeability. Fast-flow pathways persist through the PTn where faults disrupt the generally matrix-flow-dominated nonwelded tuffs, but water moving along such paths probably is only a small fraction of the total flux because of matrix imbibition. Fault zones may have the capacity to conduct substantial volumes of water through the entire unsaturated zone to the water table; but without a supply from lateral flow, major faults may conduct little water. Faults may be locally impermeable to lateral flow, resulting in perching, where the fault has offset a permeable bed opposite a less permeable bed. In general, faults are vertical conduits for both air and water flow.
This paper has described and traced the evolution of the conceptual model of hydrology for the unsaturated zone at Yucca Mountain. The study area is 145 km northwest of Las Vegas, Nevada, and covers approximately 45 km2. The unsaturated zone is 500 m thick with the potential repository located 300 m below land surface. Precipitation ranges from less than 130 mm/yr at the lower elevations in the south to more than 280 mm/yr at the higher elevations in the north and averages 170 mm/yr across the potential repository. Yucca Mountain is in the Basin and Range physiographic province and consists of a Tertiary volcanic sequence that varies between 1 and 3 km in thickness with an unsaturated zone that varies from 500 m in thickness near the potential repository to more than 750 m thick in the north part of the study area. The sequence consists of alternating layers of welded and nonwelded tuffs ranging in age from approximately 11.6 million years to 13.5 million years. The layers have been downcut in fault-controlled washes and in erosion-controlled washes forming on uplifted fault blocks.
During the last 15 years, several iterations of conceptual models for Yucca Mountain have been developed. Most of the models have the same general nature but differ significantly in detail. The most persistent concepts assumed negligible infiltration rates, extensive lateral flow in the PTn, and virtually no fracture flow
through the TSw (the potential repository horizon), resulting in extremely low vertical percolation rates and long (105 yr) travel times. Most of the conceptual models allowed some lateral flow in the underlying CHn for various reasons. The current conceptual model has relatively high infiltration rates (as much as 80 mm/yr in some locations), fracture-dominated flow in the TCw, vertical matrix-dominated flow in the PTn (little lateral flow), fracture-dominated flow in the TSw, vertical matrix-dominated flow in the vitric rocks of the Calico Hills and Prow Pass, and extensive lateral flow above the zeolitic boundary in those units, all of which lead to much shorter (<104 yr) travel times through the system.
The faults are assumed to be permeable to water and air and to transmit bomb-pulse isotopes (fast-flow paths) through continuous, connected, fractured paths through the PTn, in which flow is otherwise slow because it is matrix-dominated. Once through the PTn, the water can continue in the fault zone or move into the fracture-dominated flow field of the TSw. Perched water bodies form when lateral flow along a zeolitic contact reaches a less permeable rock unit that has been offset by faulting. The fault will act as a conduit to slowly drain the perched water body in equilibrium with the large area inflow.
This paper has presented the most recent conceptual model of the unsaturated zone at Yucca Mountain but is by no means the final word. During the next several years as underground testing continues and more data are collected, analysis of and insights into the mechanisms of unsaturated flow likely will lead to a better understanding of the unsaturated zone at Yucca Mountain.
Altman, S. J., B. W. Arnold, R. W. Barnard, G. E. Barr, C. K. Ho, S. A. McKenna, and R. R. Eaton, 1996. Flow Calculations for Yucca Mountain Groundwater Travel Time (GWTT-95) . Albuquerque, N. Mex.: Sandia National Laboratories. SAND96-0819.
Anderson, L. A., 1981. Rock Property Analysis of Core Samples from the Calico Hills UE25a-3 Borehole, Nevada Test Site, Nevada. Denver, Colo.: U.S. Geological Survey Open-File Report 81-1337. 30 p.
Barnard, R. W., M. L. Wilson, H. A. Dockery, J. H. Ganthier, P. G. Kaplan, R. R. Eaton, F. W. Bingham, and T. H. Robey, 1992. TSPA 1991: An Initial Total-System Performance Assessment for Yucca Mountain. Albuquerque, N. Mex.: Sandia National Laboratories. SAND91-2795.
Bodvarsson, G. S., and T. M. Bandurraga, eds., 1996. Development and Calibration of the Three-Dimensional Site-Scale Unsaturated-Zone Model of Yucca Mountain, Nevada-Yucca Mountain Project Milestone OBO2. Berkeley, Calif.: Lawrence Berkeley National Laboratory Report LBNL-39315.
Bodvarsson, G. S., E. Sonnenthal, and Y-S. Wu., eds., 1998. Unsaturated Zone Flow and Transport Modeling of Yucca Mountain, FY98 . Berkeley, Calif.: Lawrence Berkeley National Laboratory Milestone Report SP3CKJM4.
Botkin, D. B., R. A. Nisbet, S. Bicknell, C. Woodhouse, B. Bentley, and W. Ferren, 1991. Global climate change and California's natural ecosystems. In: Knox, J. B., and A. F. Scheuring, eds. Global Climate Change and California, Potential Impacts and Responses : Berkeley, Calif., University of California Press. p. 123-146.
Buesch, D. C., R. W. Spengler, T. C. Moyer, and J. K. Geslin, 1996. Proposed Stratigraphic Nomenclature and Macroscopic Identification of Lithostratigraphic Units of the Paintbrush Group Exposed at Yucca Mountain. Nevada: U.S. Geological Survey Open-File Report 94-469. 47 p.
Burger, P. A., and K. M. Scofield, 1994. Perched water occurrences at Yucca Mountain and their implications on the exploratory studies facility. Unpublished memorandum to the USGS Technical Project Officer.
Byers, F. M., Jr., W. J. Carr, P. P. Orkild, W. D. Quinlivan, and K. A. Sargent, 1976. Volcanic Suites and Related Cauldrons of the Timber Mountain-Oasis Valley Caldera Complex, Southern Nevada: U.S. Geological Survey Professional Paper 919. 70 p.
Carr, W. J., F. M. Byers, Jr., and P. P. Orkild, 1986. Stratigraphic and Volcano-Tectonic Relations of the Crater Flat Tuff and Some Older Volcanic Units, Nye County, Nevada: U.S. Geological Survey Professional Paper 1323. 28 p.
Fabryka-Martin, J. T., A. L. Flint, D. S. Sweetkind, A. V. Wolfsberg, S. S. Levy, G. J. C. Roemer, J. L. Roach, L. E. Wolfsberg, and M. C. Duff, 1997a. Evaluation of Flow and Transport Models of Yucca Mountain, Based on Chlorine-36 and Chloride Studies for FY97. Los Alamos, N. Mex.: Los Alamos National Laboratory, Yucca Mountain Project Milestone Report SP2224M3.
Fabryka-Martin, J. T., and B. Liu, 1995. Distribution of Chlorine-36 in UZ-14, UZ-16, Perched Water, and the ESF North Ramp, Yucca Mountain, Nevada. Los Alamos, N.Mex.: Los Alamos National Laboratory Milestone Report 3431.
Fabryka-Martin, J. T., S. J. Wightman, W. J. Murphy, M. P. Wickham, M. W. Caffee, G. J. Nimz, J. R. Southon, and P. Sharma, 1993. Distribution of chlorine-36 in the unsaturated zone at Yucca Mountain: an indicator of fast transport paths. In: FOCUS ‘93, Site Characterization and Model Validation, Las Vegas, Nev., Sept. 26-29, 1993. Proceedings. La Grange Park, Ill.: American Nuclear Society, pp. 58-68.
Fabryka-Martin, J. T., S. J. Wightman, B. A. Robinson, and E. W. Vestal, 1994. Infiltration Processes at Yucca Mountain Inferred from Chloride and Chlorine-36 Distributions. Los Alamos, N. Mex.: Los Alamos National Laboratory. Milestone Report 3417.
Fabryka-Martin, J. T., A. V. Wolfsberg, P. R. Dixon, S. Levy, J. Musgrave, and H. J. Turin, 1997b. Summary report of Chlorine-36 Studies: Sampling for Chlorine-36 in the Exploratory Studies Facility. Los Alamos, N. Mex.: Los Alamos National Laboratory. Report LA-13552-MS.
Fabryka-Martin, J. T., A. V. Wolfsberg, S. S. Levy, K. Campbell, P. Tseng, J. L. Roach, and L. E. Wolfsberg, 1998. Evaluation of Flow and Transport Models of Yucca Mountain, Based on Chlorine-36 and Chloride Studies for FY98. Los Alamos, N. Mex.: Los Alamos National Laboratory . Yucca Mountain Project Milestone Report SP32D5M3.
Flint, A. L., and L. E. Flint, 1994. Spatial Distribution of Potential Near Surface Moisture Flux at Yucca Mountain, Nevada. In: Proceedings of the International High-Level Radioactive Waste Conference, Las Vegas, Nev., May 22-26, 1994. La Grange Park, Ill.: American Nuclear Society , pp. 2352-2358.
Flint, A. L., J. A. Hevesi, and L. E. Flint, 1996a. Conceptual and Numerical Model of Infiltration for the Yucca Mountain Area, Nevada. Las Vegas, Nev.: U.S. Department of Energy Milestone Report 3GUI623M, September 1996, 210 p.
Flint, L. E., 1998. Characterization of Hydrogeologic Units Using Matrix Properties. U.S. Geological Survey Water-Resources Investigation Report 96-4342, 61 p.
Flint, L. E., and A. L. Flint, 1990. Preliminary Permeability and Moisture Retention of Nonwelded and Bedded Tuffs at Yucca Mountain, Nevada. U.S. Geological Survey Open-File Report 90-569, 57 p.
Flint, L. E., and A. L. Flint, 1995. Shallow Infiltration Processes at Yucca Mountain, Nevada. Neutron Logging Data 1984-93. U.S. Geological Survey Water-Resources Investigations Report 95-4035, 46 p.
Flint, L. E., A. L. Flint, and J. A. Hevesi, 1993. Shallow infiltration processes in arid watersheds at Yucca Mountain, Nevada,. In: Fifth International High Level Radioactive Waste Management Conference, Las Vegas, Nev., May 22-26. Proceedings. La Grange Park, Ill.: American Nuclear Society , pp. 2315-2322.
Flint, L. E., A. L. Flint, C. A. Rautman, and J. D. Istok, 1996b. Physical and Hydrologic Properties of Rock Outcrop Samples at Yucca Mountain, Nevada. U.S. Geological Survey Open-File Report 95-280, 70 p.
Flint, A. L., L. E. Flint, G. S. Bodvarsson, E. M. Kwicklis, and J. T. Fabryka-Martin, in press. The hydrology of Yucca Mountain. Reviews of Geophysics.
Ganthier, J. H., M. L. Wilson, and F. C. Lauffer, 1992. Estimating the consequences of significant fracture flow at Yucca Mountain. In: Third International Conference on High-Level Radioactive Waste Management, Las Vegas, Nev. Proceedings. La Grange Park, Ill.: American Nuclear Society , pp. 727-731.
Glass, R. J., 1993. Modeling gravity-driven fingering in rough-walled fractures using modified percolation theory. In: High Level Radioactive Waste Management Conference, Las Vegas, Nev. Proceedings. La Grange, Ill.: American Nuclear Society, v. 2, pp. 2042-2049.
Glass, R. J., A. L. Flint, V. C. Tidwell, W. Peplinski, Y. Castro, 1994. Fracture-matrix interaction in Topopah Spring tuff: Experiment and numerical simulation. In: International High-Level Radioactive Waste Conference, Las Vegas, Nev., May 22-26. Proceedings. La Grange Park, Ill.: American Nuclear Society, 9 p.
Grayson, D.K., 1993. The Desert's Past: A Natural Prehistory of the Great Basin. Washington, D.C.: Smithsonian Institution Press, 356 p.
Hammermeister, D. P., D. O. Blout, and J. C. McDaniel, 1985. Drilling and coring methods that minimize the disturbance of cuttings, core and rock formation in the unsaturated zone, Yucca Mountain, Nevada, in National Water Well Association Conference on Characterization and Monitoring of the Vadose (Unsaturated) Zone. Proceedings. Denver, Colo.: National Water Well Association , p. 507-541.
Hanks, T. C., I. C. Winograd, R. E. Anderson, T. E. Reilly, and E. P. Weeks, 1999. Yucca Mountain as a radioactive-waste repository at Yucca Mountain . A report to the Director: U.S. Geological Survey Circular 1184, 19 p.
Hevesi, J. A., and A. L. Flint, 1996. Geostatistical Model for Estimating Precipitation and Recharge in the Yucca Mountain Region, Nevada-California. U.S. Geological Survey Water-Resources Investigations Report 96-4123.
Hevesi, J. A., A. L. Flint, and J. D. Istok, 1992. Precipitation estimation in mountainous terrain using multivariate geostatistics. II. Isohyetal maps. Journal of Applied Meteorology 31(7): 677-688.
Ho, C. K., 1995. Assessing alternative conceptual models of fracture flow. In: TOUGH Workshop '95, K. Pruess, ed. Proceedings. Berkeley, Calif.: Lawrence Berkeley National Laboratory, Lawrence Berkeley Laboratory Report LBL-37200.
Ho, C. K., 1997. Models of fracture-matrix interactions during multiphase heat and mass flow in unsaturated fractured porous media. In: Sixth Symposium on Multiphase Transport in Porous Media. Dallas, Tex. 1997 ASME International Mechanical Engineering Congress and Exposition .
Ho, C. K., S. J. Altman, and B. W. Arnold, 1995. Alternative Conceptual Models and Codes for Unsaturated Flow in Fractured Tuff: Preliminary Assessments for GWTT-95, Yucca Mountain Site Characterization Project Report. Albuquerque, N. Mex.: Sandia National Laboratories, SAND95-1456.
Hudson, D. B., and A. L. Flint, 1995. Estimation of Shallow Infiltration and Presence of Fast Pathways for Shallow Infiltration in the Yucca Mountain Area, Nevada. Las Vegas, Nev.: U.S. Department of Energy Milestone Report 3GUI611M.
Istok, J. D., C. A. Rautman, L. E. Flint, and A. L. Flint, 1994. Spatial variability in hydrologic properties of a volcanic tuff. Groundwater 32: 751-760.
Klavetter, E. A., and R. R. Peters, 1986. Estimation of Hydrologic Properties of an Unsaturated Fractured Rock Mass. Albuquerque, N. Mex.: Sandia National Laboratories, SAND84-2642.
Klavetter, E. A., and R. R. Peters, 1987. An Evaluation of the Use of Mercury Porosimetry in Calculating Hydrologic Properties of Tuffs from Yucca Mountain, Nevada. Albuquerque, N. Mex.: Sandia National Laboratories, SAND86-0286, 74 p.
Kwicklis, E. M., and J. P. Rousseau, 1999. Analysis of percolation flux based on heat flow estimated in boreholes . In: Rousseau, J. P., E. M. Kwicklis, and D. C. Gillies, eds. Hydrogeology of the Unsaturated Zone, North Ramp Area of the Exploratory Studies Facility. Yucca Mountain, Nevada: U.S. Geological Survey Water-Resources Investigations Report 98-4050, 244 p.
LeCain, G. D., 1997. Air-Injection Testing in Vertical Boreholes in Welded and Nonwelded Tuff. Yucca Mountain, Nevada: U.S. Geological Survey Water-Resources Investigations Report 96-4262, 33 p.
Maxey, G. B., and T. E. Eakin, 1950. Ground Water in White River Valley, White Pine, Nye, and Lincoln Counties. Nevada. Nevada State Engineer, Water Resources Bulletin (8):59 p.
Montazer, P., and W. E. Wilson, 1984. Conceptual Hydrologic Model of Flow in the Unsaturated Zone, Yucca Mountain, Nevada. U.S. Geological Survey Water-Resources Investigation Report 84-4345, 55 p.
Moyer, T. C., and J. K. Geslin, 1995. Lithostratigraphy of the Calico Hills Formation and Prow Pass Tuff (Crater Flat Group) at Yucca Mountain, Nevada. U.S. Geological Survey Open-File Report 94-460, 59 p.
Moyer, T. C., J. K. Geslin, and L. E. Flint, 1996. Stratigraphic Relations and Hydrologic Properties of the Paintbrush Tuff Nonwelded (PTn) Hydrologic Unit, Yucca Mountain, Nevada. U.S. Geological Survey Open-File Report 95-397, 151 p.
Paces, J. B., L. A. Neymark, B. D. Marshall, J. F. Whelan, and Z. E. Peterman, 1996. Ages and Origins of Subsurface Secondary Minerals in the Exploratory Studies Facility (ESF). U.S. Geological Survey Milestone Report 3GQH450M.
Patterson, G. L., 1999. Occurrences of perched water in the vicinity of the Exploratory Studies Facility North Ramp. In: Rousseau, J. P., E. M. Kwicklis, and D. C. Gillies, eds. Hydrogeology of the unsaturated zone, North Ramp area of the Exploratory Studies Facility, Yucca Mountain, Nevada. U.S. Geological Survey Water-Resources Investigations Report 98-4050, Denver, Colo., 244 p.
Peters, R. R., and E. A. Klavetter, 1988. A continuum model for water movement in an unsaturated fractured rock mass. Water Resources Research 24(3):416-430.
Peters, R. R., E. A. Klavetter, I. J. Hall, S. C. Blair, P. R. Heller, and G. W. Gee, 1984. Fracture and Matrix Hydrogeologic Characteristics of Tuffaceous Materials from Yucca Mountain, Nye County, Nevada. Albuquerque, N. Mex.: Sandia National Laboratories, SAND84-1471, 108 p.
Pruess, K., 1998. On water seepage and fast preferential flow in heterogeneous, unsaturated rock fractures. Journal of Contaminant Hydrology 30(3-4):333-362.
Pruess, K., and Y. W. Tsang, 1990. On two-phase relative permeability and capillary pressure of rough-walled rock fractures. Water Resources Research 26(9):1915-1926.
Robinson, B. A., A. V. Wolfsberg, G. A. Zyvoloski, and C. W. Gable, 1995. An Unsaturated Zone Flow and Transport Model of Yucca Mountain. Los Alamos National Laboratory Yucca Mountain Project Milestone 3468.
Rockhold, M. L., B. Sagar, and M. P. Connelly, 1990. Multi-dimensional modeling of unsaturated flow in the vicinity of exploratory shafts and fault zones at Yucca Mountain, Nevada. In: First International High Level Radioactive Waste Management Conference, Las Vegas, Nev. Proceedings. La Grange Park, Ill.: American Nuclear Society, p. 153-162.
Roseboom, E. H., Jr., 1983. Disposal of high-level nuclear waste above the water table in arid regions. Alexandria, Va.: Geological Survey Circular 903, 21 p.
Ross, B., 1990. The diversion capacity of capillary barriers. Water Resources Research 26(10):2625-2629.
Rousseau, J. P., E. M. Kwicklis, and D. C. Gillies, eds., 1998. Hydrogeology of the Unsaturated Zone, North Ramp Area of the Exploratory Studies Facility, Yucca Mountain, Nevada. U.S. Geological Water-Resources Investigations Report 98-4050, 244 p.
Rulon, J. J., G. S. Bodvarsson, and P. Montazer, 1986. Preliminary Numerical Simulations of Groundwater Flow in the Unsaturated Zone, Yucca Mountain, Nevada. Berkeley, Calif.: Lawrence Berkeley Laboratory , LBL 20553, 91 p.
Rush, F. E., W. Thordarson, and L. Bruckheimer, 1983. Geohydrologic and Drill-Hole Data for Test Well USW H-1, Adjacent to the Nevada Test Site, Nye County, Nevada. U.S. Geological Survey Open-File Report 83-141, 68 p.
Sass, J. H., and A. H. Lachenbruch, 1982. Preliminary Interpretation of Thermal Data from the Nevada Test Site . U.S. Geological Survey Open-File Report 82-973, 30 p.
Sawyer, D. A., R. J. Fleck, M. A. Lanphere, R. G. Warren, and D. E. Broxton, 1994. Episodic volcanism in the Miocene southwest Nevada volcanic field: Stratigraphic revisions, 40Ar/39Ar geochronologic framework, and implications for magmatic evolution . Geological Society of America Bulletin 106(10): 1304-1318.
Scott, R. B., and J. Bonk, 1984. Preliminary Geologic Map of Yucca Mountain with Geologic Sections, Nye County, Nevada. U.S. Geological Survey Open-File Report 84-494.
Scott, R. B., R. W. Spengler, S. Diehl, A. R. Lappin, and M. P. Chornack, 1983. Geologic character of tuffs in the unsaturated zone at Yucca Mountain, southern Nevada. In: Mercer, J. W., P. S. C. Rao, and I. W. Marine, eds. Role of the Unsaturated Zone in Radioactive and Hazardous Waste Disposal . Ann Arbor, Mich.: Ann Arbor Science, p. 289-335.
Sinnock, S., Y. T. Lin, and J. P. Brannen, 1984. Preliminary Bounds on the Expected Post-Closure Performance of the Yucca Mountain Repository Site, Southern Nevada. Albuquerque, N. Mex.: Sandia National Laboratories, SAND84-3918, 83 p.
Sinnock, S., Y. T. Lin, and J. P. Brannen, 1987. Preliminary bounds on the expected post-closure performance of the Yucca Mountain repository site, southern Nevada. Journal of Geophysical Research 92(B8): 7820-7842.
Snyder, D. B., and W. J. Carr, 1982. Preliminary Results of Gravity Investigations at Yucca Mountain and Vicinity, Southern Nye County, Nevada. U.S. Geological Survey Open-File Report 82-701, 36 p.
Striffler, P., G. O'Brien, T. Oliver, and P. A. Burger, 1996. Perched water characteristics and occurrences, Yucca Mountain, Nevada . Unpublished memorandum to the U.S. Geological Survey Yucca Mountain Project Technical Project Officer.
Sweetkind, D. S., J. T. Fabryka-Martin, A. L. Flint, C. J. Potter, and S. S. Levy, 1997. Evaluation of the Structural Significance of Bomb Pulse 36Cl at Sample Locations in the Exploratory Studies Facility, Yucca Mountain, Nevada. Las Vegas, Nev.: U.S. Department of Energy Milestone Report SPG33M4.
Tucci, P., and D. J. Burkhardt, 1995. Potentiometric-Surface Map, 1993, Yucca Mountain and Vicinity, Nevada . U.S. Geological Survey Water-Resources Investigations Report 95-4149.
U.S. Department of Energy, 1984. Draft Environmental Assessment: Yucca Mountain Site, Nevada Research and Development Area, Nevada. Washington, D.C.: U.S. Department of Energy.
U.S. Department of Energy, 1992. Report of Early Site Suitability Evaluation of the Potential Repository Site at Yucca Mountain, Nevada. Washington, D.C.: U.S. Department of Energy.
Waddell, R. K., J. H. Robison, and R. K. Blankennagel, 1984. Hydrology of Yucca Mountain and Vicinity, Nevada-California: Investigative Results Through Mid-1983. U.S. Geological Survey Water-Resources Investigations Report 84-4267, 72 p.
Wang, J. S. Y., and T. N. Narasimhan, 1985. Hydrologic mechanisms governing fluid flow in a partially saturated, fractured, porous medium. Water Resources Research 21(12): 1861-1874.
Weeks, E. P., and W. E. Wilson, 1984. Preliminary Evaluation of Hydrologic Properties of Cores of Unsaturated Tuff, Test Well USW H-1, Yucca Mountain, Nevada. U.S. Geological Survey Water-Resources Investigations Report 84-4193, 30 p.
Whitfield, M. S., Jr., C. M. Cope, and C. L. Loskot, 1984. Borehole and Geohydrologic Data for Test Hole USW UZ-6, Yucca Mountain Area, Nye County, Nevada. U.S. Geological Survey Open-File Report 92-28, 41 p.
Wilson, M. L., 1996. Lateral diversion in the PTn unit: Capillary-barrier analysis. In: High Level Radioactive Waste Management Seventh Annual International Conference, Las Vegas, Nev., April 29-May 3. Proceedings. La Grange Park, Ill.: American Nuclear Society, SAND95-2186C, p. 111-113.
Winograd, I. J., 1971. Hydrogeology of ash-flow tuff: A preliminary statement. Water Resources Research 7(4): 994-1006.
Winograd, I. J., 1972. Near-surface storage of solidified high-level radioactive waste in thick (400-2,000 foot) unsaturated zones in the southwest. Geological Society of America, Abstracts with Programs 4: 708.
Winograd, I. J., 1974. Radioactive waste storage in the arid zone. EOS, Transactions of the American Geophysical Union 55: 884-894.
Winograd, I. J., 1981. Radioactive waste disposal in thick unsaturated zones. Science 212(4502): 1457-1464.
Winograd, I. J., and W. Thordarson, 1975. Hydrogeologic and Hydrochemical Framework, South-Central Great Basin, Nevada-California, With Special Reference to the Nevada Test Site . U.S. Geological Survey Professional Paper 712-C, 126 p.
Wittwer, C. S., G. S. Bodvarsson, M. P. Chornack, A. L. Flint, L. E. Flint, B. D. Lewis, R. W. Spengler, and C. A. Rautman, 1992. Design of a three-dimensional site-scale model for the unsaturated zone at Yucca Mountain, Nevada. In: International High-Level Radioactive Waste Conference, Las Vegas, Nev., April 12-16. Proceedings. La Grange Park, Ill.: American Nuclear Society , p. 263-271.
Wittwer, C. S., G. Chen, G. S. Bodvarsson, M. P. Chornack, A. L. Flint, L. E. Flint, E. M. Kwicklis, and R. W. Spengler, 1995. Preliminary Development of the LBL/USGS Three-Dimensional Site-Scale Model of Yucca Mountain, Nevada. Berkeley, Calif.: Lawrence Berkeley National Laboratory , LBL-37356, UC0814.
Wolfsberg, A. V., J. T. Fabryka-Martin, K. S. Campbell, S. S. Levy, and P. H. Tseng, 1999. Use of chlorine-36 and chloride data to evaluate fracture flow and transport models at Yucca Mountain. In: International Symposium: Dynamics of Fluids in Fractured Rocks: Concepts and Recent Advances, in Honor of Paul A. Witherspoon, February 10-12, 1999. Proceedings. Berkeley, Calif.: Lawrence Berkeley National Laboratory.
Wu, Y. S., S. Finsterle, and K. Pruess, 1996. Computer models and their development for the unsaturated zone model at Yucca Mountain, Chapter 4 in Development and Calibration of the Three-Dimensional Site-Scale Unsaturated-Zone Model of Yucca Mountain, Nevada. G. S. Bodvarsson and T. M. Bandurraga, eds. Berkeley, Calif.: Lawrence Berkeley National Laboratory Milestone OBO2, MOY-970317-04.
Yang, I. C., G. W. Rattray, and P. Yu, 1996. Interpretations of Chemical and Isotopic Data from Boreholes in the Unsaturated-Zone at Yucca Mountain, Nevada. U.S. Geological Survey Water-Resources Investigations Report 96-4058.
Yang, I. C., P. Yu, G. W. Rattray, and D. C. Thorstenson, 1998. Hydrogeochemical Investigations and Geochemical Modeling in Characterizing the Unsaturated Zone at Yucca Mountain, Nevada. U.S. Geological Survey Water-Resources Investigation Report 98-4132.