Hydrogeologic Investigations of Flow in Fractured Tuffs, Rainier Mesa, Nevada Test Site (American Geophysical Union, 1987)

Page 1

Hydrogeologic Investigations of Flow in Fractured Tuffs, Rainier Mesa, Nevada Test Site Charles E. Russell, John W. Hess, and Scott W. Tyler

Page 2

HYDROGEOLOGIC INVESTIGATIONS OF FLOW IN FRACTURED TUFFS RAINIER MESA, NEVADA TEST SITE Charles E. Russell, John W. Hess, and Scott W. Tyler Water Resources Center, Desert Research Institute, Las Vegas, Nevada, 89120 Abstract. Rainier Mesa, a primary site for nuclear testing, is located in the north central area of the Nevada Test Site, and is composed of highly fractured and altered Tertiary tuffs. A hydrogeologic study was conducted within the mesa concentrating on several parameters: the source of ground water found in Rainier Mesa, period of principal recharge, ground-water travel time between the mesa surface and the tunnel level, period of hydraulic response, and total amount of recharge per year to Ul2n Tunnel recharge basin. The data base consists of: the precipitation record, discharge record of seeps within Rainier Mesa, the gross chemistry and stable isotopic composition of these seeps, and two tracer stud ies conducted from the mesa surface. Results indicate that ground water is of recent meteoric origin, winter is the period of principal re charge, the period of hydrologic response is at least four months, the total recharge is approxi mately 8 percent of the precipitation which falls on the U12n recharge basin, and travel time is estimated as greater than one year and less than six. These observations describe an envirorment which has been subjected to nuclear test ing since 1957. It is certain that this testing has altered the hydrogeochemical environment and it is suspected that some alteration of the hy draulic system has also occurred. Since 1957, nuclear testing has been conducted within the adit systems mined into Rainier Mesa, Nevada Test Site (NTS). These adits are com monly referred to as tunnels. There is concern that migration of radionuclides may occur from the testing areas in the adits to the regional ground-water system approximately 500 m below. The primary purpose of this paper is to quantita tively delineate the hydrogeologic processes which occur within Rainier Mesa so that later studies may predict the rate and extent of radio nuclide migration. This paper represents one of the few field data sets describing vadose zone Copyright 1987 by the American Geophysical Union. fracture flow in tuffs and should be of interest to others working in this field. The objective of this study is to quantita tively investigate the hydrogeologic regime of Rainier Mesa. The hydrogeologic parameters of concern are: origin of perched water within the vadose zone, period of principal recharge, total area, period of hydrologic response and travel time of ground-water from the mesa surface to the tunnel level approximately 350 m below the sur To delineate the flow system of Rainier Mesa, the following parameters were monitored: perched ground-water discharge from two seeps within U12n Tunnel, as well as the portal discharge of U12n Tunnel, the stable isotopic and geochemical data from the previously mentioned seeps, two tracer studies incorporating four different tracers, precipitation data from the surface of the mesa and tritium samples from within the tunnels. Rainier Mesa is located in the north central portion of the NTS, approximately 140 km north west of Las Vegas, Nevada. The top of the mesa covers approximately 11.4 km^ and has an aver age elevation of 2200 m. The mesa is the highest point of the north-south trending Belted Range, and is a typical fault block mountain of the basin and range province [Hansen et al., 1963]. Rainier Mesa is the remnant of a volcanic plateau which was uplifted during an episode of tectonic extension in the middle to late Cenozoic. It is composed of a series of nearly parallel, roughly planar Miocene Tuffs which dip 10 to 25° to the west. These tuffs originated from a series of calderas to the west and south of Rainier Mesa. The stratigraphy of Rainier Mesa is listed in Table 1 and is a summary of Maldonado et al. [1979]. The welded tuffs within

Page 3

44 FLOW IN FRACTURED TUFFS, RAINIER MESA TABLE 1. Geologic Units of Rainier Mesa Thickness (m) Brief Geologic Unit Mean Range Descript ion Timber Mountain Tuff Rainier Mesa Member 76.9 13-182 welded to densely welded welded to nonwelded, some members friable Stockade Wash Tuff 8.7 0-14 Ash-flow, nonwelded, Bedded and Ash Flow Tuffs of Area 20 40.7 0-63 Lava and Tuff of Dead Horse Flat 14.6 0-39 Indian Trail Formation Belted Range Tuff Tub Spring Member 6.5 0-21 partially welded in some Indian Trail Formation Crater Flat Tuff 10.6 0-68 Ash-flow, nonwelded to densely welded, some Indian Trail Formation 18.1 0-82 Ash-fall, nonwelded, Rainier Mesa are highly fractured, whereas the zeolitic units are fractured to a lesser degree. The only unit not appreciably fractured is the friable tuff of the Paintbrush Formation. Nuclear testing within Rainier Mesa has al tered the hydrogeology, predominately by increas ing the number of fractures [Clebsch, 1961]. As a result of increased fracturing, fluid flow through the mesa is not expected to depict pre testing recharge conditions. The presence of continued discharge at the tunnel seeps indicates that recharge from the raesa surface is occurring to date and offers a unique opportunity to study vadose zone fracture flow. The hydrogeology of Rainier Mesa is controlled primarily by a heterogeneous, poorly connected fracture system. Flow within the upper for mations of the mesa occurs primarily within the fractures due to low matrix permeability. The friable tuff of the Paintbrush Formation is characterized by unsaturated interstitial flow. Below this, flow in the Indian Trail Formation is predominately saturated fracture flow. This formation acts as an aquitard within the Mesa. Perched water lenses exist within the fracture systems of the Indian Trail Formation at an ele vation of 1820 100 m and extend throughout the unit. It is within this formation where most of the nuclear testing is done. Below the Indian Trail Tuff, unsaturated fracture flow continues until the regional ground-water system is reached 900 m below the surface of the mesa (Thordarson, 1965). An idealized cross section of Rainier Mesa, showing the major units and the occurrence of saturated fractures is on Figure 1.

Page 4

RUSSELL ET AL. 47 Fig. 5. The estimated U12n recharge basin. There are several problems with the vapor phase transport estimate. First, the percent of vapor contributed to the total relative humidity by water used for mining activities and water evaporated from interstitial pores is unknown, thus the percent of the vapor transport contrib uted by the evaporation of fracture waters is unknown. If the combined contribution from all three sources is used in our calculations, then an estimated maximum discharge will be the re sult. Second, the humidity measurements were taken during summer when the evaporation compo nent is larger. Therefore the estimate is a conservative estimate with respect to vapor From the following equation the total discharge from Ul2n tunnel is A source of error in this estimate is the pos sible contribution to the total flow from dril ling fluid. Over the last 30 years an estimated 377,000 nr* of drilling fluid were lost to the entire mesa. Although large, this fluid was dis tributed between the five major tunnel systems over a thirty year period. Averaged over this space and time, this contribution is approximate ly 7% of the total yearly flow from U12n Tunnel. Due to the large uncertainty of this estimate, the effect of drilling fluid on total discharge will be ignored. A second problem is the existence of nuclear test generated fractures, and the drill holes themselves. These features tend to increase recharge by increasing permea bility, thus the recharge estimate from this study may be larger than that of the pre-nuclear period for Rainier Mesa. Figure 5 shows the estimated recharge basin for U12n Tunnel. The boundaries of this recharge basin were plotted using the orientation and extent of known faults and the topography of the mesa surface. It is estimated that 50 to 60% of faults carry ground water in Rainier Mesa, and these faults are orientated approximately north south and are steeply dipping, [Thordarson, Secondary basins, which are +_ 30% the size of the first basin, are also included to encompass reasonably expected deviations in the parameters which affect fracture recharge. Assuming that Rainier Mesa is homogeneous with respect to frac ture transport of ground water and assuming the recharge basin falls into the above range, then the recharge per unit area for U12n recharge basin may be calculated where: Ry equals the recharge per unit area for discharge from U12n Tunnel, and U is the area of the U12n tunnel recharge basin. The area of the estimated recharge basin is 1.47X10° m^. The basins containing +_ 30% of this area are Thus for the best estimate

Page 5

48 FLOW IN FRACTURED TUFFS, RAINIER MESA TABLE 2. Discharge from Various Tunnels in Rainier Mesa Tunnel Fluid Discharge U12g 0.036 Fernandez and Freshley (1984) U12n 53.0 Russell et al. (1987) U12e 19.0 S. Tyler, Field Inspec. 2/13/87 for the secondary basins The above estimates are an estimated range of recharge per unit area for the Ul2n Tunnel re charge basin. This recharge estimate does not account for ground-water that does not discharge through the Ul2n Tunnel Portal, thus it is a con The area of Rainier Mesa directly above Ul2n Tunnel may be more conducive to recharge than other areas. The aqueduct, the principal surface drainage of Rainier Mesa, is directly over por tions of thetunnel. This could promote in creased infiltration with respect to other areas of the mesa. Thetunnel also cross-cuts the Aqueduct syncline, a major structural feature of Rainier Mesa. Down dip flow in the vadose zone is possible and may be contributing a greater discharge value than would normally be expected. This study needs to be conducted on the other tunnel systems within the mesa in order to arrive at an estimated total recharge. Preliminary discharge estimates have been made for several other tunnels and are reported by Tyler and Mihevic (in preparation), these estimates are on The values on Table 2 indicate that the re charge processes acting on Rainier Mesa are high ly variable. To give an example, Thordarson [1965] estimated a total recharge value of 172,600 m^/yr for Rai nier Mesa. Through the use of this estimate, Ul2n Tunnel accounts for 20% of the total recharge in only 6% of the total area. A detailed study of the discharge from the other tunnel systems could be used to calculate the total recharge through Rainier Mesa. The percent of precipitation which recharges the Ul2n recharge basin can be calculated from where pr is the percent of precipitation which recharges, P is the yearly average precipitation for Rainier Mesa from the last four years, and T is the total discharge from Ul2n Tunnel . From Approximately 8 % of all precipitation which falls on Rainier Mesa acts as recharge. This is very close to Eakins [1962] estimate of 7% as reported by Thordarson [1965]. Period of Hydraulic Response The period of hydraulic response is the period of time it takes for a given recharge event to cause a corresponding increase in discharge at the seeps. Two pieces of information are re quired to determine this parameter. The first is a complete precipitation record for the mesa, and the second is discharge records from the seeps within the mesa covering the same time period. Only large winter precipitation events were used in determining the period of hydraulic response. This was done in order to eliminate those precip itation events which may not have recharged the mesa. The precipitation record was examined from September 1, 1983 to August 31, 1986. Discharge was measured at three points: the 03 drift, 05 drift, and the portal weirs. These records ex tend back to September, 1985. The two types of data were subject to cross- correlation analysis in order to determine if there was any signifi cant correlation between recharge and discharge events. Unfortunately, this technique was un successful. The next step was to apply an average response technique. This methodology averages the discharge records following sus pected recharge events. It is used to isolate an average response time following a given stim ulus [B., Kinnison, Personal Communication, Nov. 1986]. An example is given on Figure 6. This example uses two major storms of the 1985 and 1986 season as stimulus, and three discharge rec ords for each storm. Each record begins on the date of the appropriate storm. ^The storms were on November 11, 1985 and January 29, 1986. The six discharge records were averaged together to create Figure 6. From this figure, a large in crease in discharge is noted at a time lag of approximately 120 days with a secondary increase at a time lag of 30 days. The secondary increase is interpreted as the discharge response of the November 11 storm, as recorded on the record which began on January 29, 1986. The averaged response of the six discharge

Page 6

RUSSELL ET AL. 45 Fig. 1. Cross-section of Rainier Mesa, showing hydrogeologic units and idealized occurrence of Results and Discussion Stable Isotopes from Composition for Rainier Source of Water Thordarson [1965], through the use of the Eakin [1962] reconnaissance method, estimated recharge was occurring into Rainier Mesa. It is one purpose of this study to validate this esti mate with field data. To investigate if recharge is occurring, two ground-water parameters were monitored over time: the stable isotopic composition and discharge of actively flowing seeps within the tunnel system. A graph of ground-water discharge from a seep in U12n.05 drift from March 1986 is presented in Figure 2. The gradual increase in discharge throughout March indicates the hydrologic system responded to an input signal. Other seeps that were monitored have also exhibited such increases in flow during this time. The stable isotopic composition of both the precipitation and tunnel seeps are plotted in Figure 3. An examination of this figure reveals that the isotopic composition of the seep discharge falls on the meteoric water line in the same area as present day mesa precip itation. Figure 3 indicates that the fracture water found in the U12n.03 and U12n.05 seeps are Fig. 2. Discharge from U12n.05 drift for March Fig. 3- The isotopic composition of Rainier Mesa precipitation and ground water relative to the meteoric water line. isotopically similiar to present day precipita tion. The above evidence seems to indicate that the fracture water is derived from recent precip itation. However, other sources could be respon sible for the increased flow. These sources are: a nuclear test near U12n.05 drift, drilling fluid lost from holes near the seeps, or an increase in The increase in discharge due to a ruclear test is characterized by a sharp increase in flow over a one to two day period [C. Russell, J. Hess, and S. Tyler, in preparation]. The dis charge on Figure 2 does not show this. The U12n.05 drift seep is located in the …