By Linda L. Lehman and Tim P. Brown

L. Lehman & Associates, Inc.

Reader's Note: L. Lehman & Associates is an internationally recognized consulting firm specializing in hydrogeology and nuclear and hazardous waste regulation. Linda Lehman, President of L. Lehman & Associates, served as an advisor to the Nevada Agency for Nuclear Projects, Nuclear Waste Project Office on hydrogeologic issues related to Yucca Mountain. The studies discussed below include analyses of water level data and computer modeling exercises of the Yucca Mountain saturated zone hydrology that lead to an alternative to the DOE preferred models of the saturated zone.


The summary report submitted by L. Lehman & Associates, Inc. (LLA) summarized several studies of Yucca Mountain saturated zone hydrology performed by over the past few years that were funded by the Nuclear Waste Project Office. The results of these individual studies, together with observations on chemistry, potentiometric surface, and temperature data led to the development of a different, more complex model of the saturated zone flow field than that presented in DOE performance assessments for Yucca Mountain.

As site characterization progressed at Yucca Mountain, data were produced that challenged the fundamental DOE concepts of flow in the unsaturated and saturated zones, as well as their linkages. For example:

Focused recharge. Evidence exists that recharge can be focused along fault and fracture zones in both the saturated and unsaturated zone in quantities that may allow penetration to repository depth through interconnected fractures;

Fault-bounded hydrologic domains and their hydrologic responses. Data indicate that fault-bounded hydrologic domains are loosely connected due to high transmissivities along domain boundaries. Hydraulic performance of these fault zones can vary depending on several factors, one of which is transmissive characteristics.

Coupled hydraulic and tectonic behavior of faults. Measured responses of water levels to earthquakes indicate significant coupling between the hydrologic system and local stress conditions.

Importance of heat in groundwater flow calculations. Due to significant vertical and lateral gradients that may be generated from a deep heat source in the vicinity of Yucca Mountain, the need for a three-dimensional, non-isothermal model of the pre-waste emplacement groundwater flow field becomes important to the calculation of groundwater travel time for the site.

The current DOE site characterization program is not designed nor is adequate to discriminate between alternative conceptual flow models.


The first major study undertaken by LLA was the examination of the water table configuration and oscillatory behavior. Water chemistry, water levels, and their fluctuation with time were studied to develop an understanding of site hydrology and behavior under the influence of repository-induced changes. A critical piece of information was the groundwater recharge data. The rate at which water would move through the repository and its volume are values that influence the quantities of radionuclides potentially transported to the environment.

Water level data at Yucca Mountain were and continue to be collected primarily by the U.S. Geological Survey. This study used water level records collected intermittently since the early 1980s.

Time series analyses have been applied to hydrologic data by many researchers. Traditional spectral analysis could not be performed for this study because the data were not uniformly spaced in time. It was possible to determine if the data possessed periodic components by applying a numerical method to fit a cosine function to irregularly spaced data (Figure 1). This method was developed and titled FIT.M by Rice (1989).

WT-7 1012.2 177.7 0.09 0.47 0.000107 1.5
WT-10 925.4 182.4 0.70 0.22 0.000074 2
WT-12 1240.0 169.8 0.70 0.35 0.000101 1.25
WT-1 889.2 249.5 0.10 0.44 0.000191 2
WT-11 887.7 253.4 0.12 0.58 0.000100 1.25
WT-16 860.6 266.9 0.11 0.68 0.000240 1.25
WT-6 2975.2 738.1 1.30 0.75 0.003230 0.5
H-5 1936.8 416.6 0.54 0.45 -0.000044 0.5
H-5 1888.4 417.9 0.31 0.28 -0.000330 0.5
WT-1* 1597.8 159.5 0.06 0.32 -0.000085 1
WT-10* 935.5 163.3 0.06 0.22 0.000083 1.75
WT-16* (fit 1) 226.4 279.7 -0.04 0.24 -0.000130 5.25
WT-16* (fit 2) 1229.4 143.2 0.04 0.24 -0.000130 0.75

*Indicates offsets were subtracted from original data (see text).

The water level records were grouped according to the parameters of the cosine function that fits the data. Two main groups of wells were apparent: one group with an average water level period of 970 days and an average phase shift of 180 days; the other group with an average water level period of 879 days and an average phase shift of 257 days. The two groups of wells with similar periodicities demonstrated a spatial distribution parallel to the major north/south trending fault zones. The pattern was similar to that reported by several researchers for the chemistry of groundwater. Results suggested the flow field at Yucca Mountain is structurally controlled and that there are two separate or weakly coupled flow systems. Also the analysis indicated that systematic water level fluctuations occur at Yucca Mountain, but the cause of the fluctuations could not be discerned.


To help understand the dynamics of the saturated zone flow system and the regional hydrologic system, analysis of variations and trends in water levels at Devil's Hole in Amargosa Valley, approximately 20 miles south of Yucca Mountain, were performed. Data on water levels in Devil's Hole were provided by the National Park Service for the period 8/30/89-3/31/95. Trends and cycles in the data were analyzed using the FIT.M program (Rice, 1989)(Figure 2). The analyses indicated that Devil's Hole water levels showed an overall downward trend. Periodicity of roughly one year was found in the overall data. Examination of the data also suggested the possibility of a longer 3.6 year cycle superimposed on the yearly cycle.


Linear fit to daily average water level for Devil's Hole, where m=the slope and b=the y intercept

Statistical comparison of the data from various other physical features was also performed and indicated a weak correlation between water supply pumping near Mercury on the Nevada Test Site and Devil's Hole. Significant correlation was found between local rainfall, Ash Meadow's spring discharge west of Devil's Hole, and upgradient monitor well water levels with Devil's Hole water levels.

The study concluded that water levels in Devil's Hole temporarily rebounded from a low in July 1992, peaked in the summer of 1993, and then resumed the previously observed long-term downward trend. The data also displayed periodicity in water levels superimposed over the long-term trend. The study provided little or no evidence of impacts on Devil's Hole from recent historical levels of water well pumping near Mercury. The results suggested complexity within the regional hydrologic system and the possibility of compartmentalized zones of flow in the area.


In order to evaluate an alternative conceptual model of saturated zone flow for the area around and including Yucca Mountain based on previous work done by LLA and other State of Nevada researchers, a numerical model was assembled and tested. The conceptual model postulates that faults and fractures control the flow of groundwater through the volcanic tuffs underlying Yucca Mountain.

The implications of this conceptual model are significant. The potential for fault/fracture zones to transmit groundwater rapidly and, by implication, transport contaminants with minimal dispersion and adsorption is high. Further, potential sensitivity to tectonic forces (earthquakes) means that the hydrologic system can change unpredictably over time. Earthquakes and/or fault movement may adjust flow properties of hydrologic conduits or create new fracture zones that could result in a major realignment of the hydrologic system.

Different approaches were used in developing the conceptual model. Examination of the distribution of temperature at the water table suggested that saturated zone flow was more complex than the simple 2-dimensional uniform flow proposed in DOE's conceptual model. Temperature distributions suggested localized recharge or an area of relatively high permeability along the strike of the Ghost Dance Fault, and a thermal hot spot along the western flank of Yucca Mountain. This information may indicate that faults could serve as pathways for both cool water from the surface as well as warm water upwelling from depth. Some faults may act as conduits while others may act as barriers.

Another approach examined water level responses in wells and springs to earthquakes. Responses were analyzed for the following earthquakes: June 28, 1992: a 7.5 magnitude earth-quake at Landers, California; June 28, 1992: a 6.6 magnitude earthquake near Big Bear, California; and June 29, 1992: a 5.6 magnitude earthquake at Little Skull Mountain, Nevada, 12 miles southeast of Yucca Mountain. Water level data showed four distinct types of response to the earthquakes:

an upward temporary spike

a rapid upward change with an apparent long-term stabilization at a higher level

a downward temporary spike

a rapid downward change with an apparent long-term stabilization at a lower level.

After the well locations were plotted and compared with known faults in the area (Figure 3) , it was found that most of the wells that experienced increased water levels were closely associated with northwest-trending shear zones, while wells that experienced downward-trending water levels aligned more closely with normal faults.


These water level responses to earthquakes have several implications relative to understanding the regional flow system and the processes which control flow. First, that the level of the water table is structurally controlled and is closely linked to ongoing regional tectonics. The responses may also suggest that the hydraulic conductivity field in these fault zones is transient, and thus not predictable over time. Lastly, localized changes in hydraulic gradient and direction or diversion of the flow field may be induced by earthquakes.

Taken together, the above information suggests a different conceptualization of the Yucca Mountain flow field than that currently being considered by DOE. The alternative conceptual model is structurally controlled by fault or fracture zones. Fault/fracture zone intersections play a key role in the distribution of recharge, velocity fields, and pathways. Also, the model is dynamic rather than static and has the potential to change rapidly due to tectonic movements.

The conceptual model(Figure 4) postulates that some water movement occurs across the Yucca Mountain block from west to east, primarily via discrete northwest trending fault zones. A steep hydraulic gradient exists at the location of the Solitario Canyon Fault which suggests a resistance to easterly flow but not a barrier to flow. Colder flow also enters the Yucca Mountain block from the northwest along the Ghost Dance Fault, and then flows southeasterly along the Drill Hole Wash structure, the Sundance Fault, and likely another fault zone at the southern end of the proposed repository that to date has not been identified or observed.


Saturated zone flow conceptual model compared with potentiometric surface (Ervin et al, 1994) and fault locations

The conceptual model emphasizes focused flow paths in the saturated zone along structural zones and weakness. It is hypothesized that fault and fracture zones are interconnected and dominate saturated zone flow.


Based upon the previous analyses, several findings regarding the saturated zone flow system at Yucca Mountain have been made:

1) A complex conceptual model of the saturated zone flow system is required to adequately define the system. The potential for fault/fracture zones to transmit groundwater rapidly and transport contaminants with minimal dispersion and absorption is high.

2) Saturated zone flow is likely compartmentalized, fault controlled, and dynamic. Predictions of future flow rates, volumes, and velocities within an active tectonic environment that meet the required regulatory assurance of confidence are not likely.