Assessment of Hydrologic Transport of Radionuclides from the Rulison Underground Nuclear Test Site, Colorado (Uc-703)

Page 1

University and Community College System of Nevada Water Resources Center ASSESSMENT OF HYDROLOGIC TRANSPORT OF RADIONUCLIDES FROM THE RULISON UNDERGROUND NUCLEAR TEST SITE, Sam Earman, Jenny Chapman, and Roko Andricevic Nevada Operations Office U.S. Department of Energy

Page 2

This report was prepared as an account of work sponsored by the United States Government. Neither the United States nor the United States Department of Energy, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, mark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. This report has been reproduced directly from the best available copy. Available to DOE and DOE contractors from the Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831; prices available from (423) 576-8401. Available to the public from the National Technical Information Service, U.S. Department of Commerce, 5285 Port Royal Rd., Springfield, VA 22161.

Page 4

ASSESSMENT OF HYDROLOGIC TRANSPORT OF RADIONUCLIDES FROM THE RULISON UNDERGROUND NUCLEAR TEST SITE, Sam Earman, Jenny Chapman, and Roko Andricevic Water Resources Center Desert Research Institute University and Community College System of Nevada Nevada Operations Office U.S. Department of Energy Las Vegas, Nevada The work upon which this report is based was supported by the U.S. Department of Energy under

Page 5

The U.S. Department of Energy (DOE) is operating an environmental restoration program to characterize, remediate, and close non-Nevada Test Site locations that were used for nuclear testing. Evaluation of radionuclide transport by groundwater from these sites is an important part of the preliminary risk analysis. These evaluations are undertaken to allow prioritization of the test areas in terms of risk, provide a quantitative basis for discussions with regulators and the public about future work at the sites, and provide a framework for assessing data needs to be filled by site characterization. The Rulison site in west-central Colorado was the location of an underground detonation of a 40-kiloton nuclear device in 1969. The test took place 2,568 m below ground surface in the Mesaverde Formation. Though located below the regional water table, none of the bedrock formations at the site yielded water during hydraulic tests, indicating extremely low permeability conditions. The scenario evaluated was the migration of radionuclides from the blast-created cavity through the Mesaverde Formation. Transport calculations were performed using the solute flux method, with input based on the limited data available for the site. Model results suggest that radionuclides from the test are contained entirely within the area currently administered by DOE. The transport calculations are most sensitive to changes in the mean groundwater velocity and the correlation scale of hydraulic conductivity, with transport of strontium and cesium also sensitive to the sorption coefficient. This modeling was performed to investigate how the uncertainty in various physical parameters affects calculations of radionuclide transport at the Rulison site, and to serve as a starting point for discussion regarding further investigation at the site; it was not intended to be a definitive simulation of migration pathways of radionuclide concentration values. Given the sparse data, the modeling results may differ significantly from reality. If needed, confidence in transport predictions can be increased by obtaining more site data, including ascertaining the amount of radionuclides available for transport (i.e., not trapped in melt glass or vented during gas flow testing), and determining accurate values for the mean groundwater velocity, correlation scale for hydraulic conductivity, and sorption coefficients for strontium and cesium.

Page 6

LIST OF FIGURES.................................................................................................................... iii LIST OF TABLES....................................................................................................................... iii Discharge Mixing Area....................................................................................................... 9 Distance to Control Plane ................................................................................................... 10 Mean Groundwater Velocity .............................................................................................. 11 SpatialVariability in Hydraulic Conductivity.................................................................. 12

Page 7

1. Rulison site location map..................................................................................................... 2 2. Generalized geologic cross section at the Rulison emplacement hole........................... 5 3. Locations of wells R-E and R-EX in SW/4, section 25................................................... 6 4. Modeled tritium concentrations at the drilling exclusion boundary resulting from the Rulison test............................................................................................ 13 5. Sensitivity of mean tritium concentration 286 m west of surface ground zero to changes in correlation scale..................................................................................... 16 6. Sensitivity of mean tritium concentration 286 m northeast of surface ground zero to changes in mean groundwater velocity................................................................... 16 7. Sensitivity of mean tritium concentration 286 m northeast of surface ground zero to changes in variance................................................................................................... 17 8. Sensitivity of mean tritium concentration 286 m northeast of surface ground zero to changes in the transverse hydrodynamic dispersion coefficient......................... 17 1. Well shut-in pressures at selected depths in the mesaverde formation ........................ 7 2. Radionuclides produced by the Rulison nuclear test........................................................ 8 3. Results of groundwater transport calculations for radionuclides produced by the Rulison test................................................................................................. 14

Page 8

The U.S. Department of Energy (DOE) and its predecessor agencies are responsible for nuclear weapons research and development as part of the national defense program. These activities include underground nuclear testing, and a small number of such tests have been conducted at sites distant from the Nevada Test Site (NTS). The Rulison site is located in west-central Colorado, approximately 64 km northeast of Grand Junction (Figure 1). Project Rulison was part of the Plowshare Program, and was designed to increase the natural gas production of a low permeability reservoir formation. The device had a yield of 40 kilotons, and was detonated 2,568 m below ground surface on September 10, 1969. DOE has implemented an environmental restoration program with the goal of characterizing, remediating, and closing the offsite nuclear test areas. An early step in this process is performing a preliminary risk analysis of the hazard posed by each site. These analyses will allow prioritization of the sites in terms of risk, provide a quantitative basis for discussions with regulators and the public about future work at the sites, and provide a framework for assessing data needs to be filled by site characterization. Desert Research Institute (DRI) is tasked with performing hydrologic risk evaluations for the groundwater transport pathway. This report details the results of the groundwater-transport evaluation for the Rulison site in terms of radionuclide concentrations that could cross the site boundary. These results will be included with evaluations of risk due to surface sources at Rulison to present a comprehensive site risk analysis in a separate report. The basic scenario evaluated for this assessment is the groundwater transport of radionuclides introduced into the subsurface by the Rulison nuclear test. This assessment strives to be as accurate as possible, but the lack of data requires that significant assumptions be made about some critical parameters. As a consequence of these limitations, the results of this modeling are meant to serve merely as a tool to guide further discussion and investigation, not as a definitive assessment of radionuclide migration at the Rulison site. The analysis relies solely on unclassified data available to the general public. Although this may increase the uncertainty of the source term data, and result in the lack of transport calculations for certain radionuclides introduced by the Rulison detonation, these issues can be investigated more thoroughly, and with much greater accuracy, after the acquisition of further data regarding contaminant transport at the site. Measured values were used wherever possible, but given the lack of data, calculations were performed for ranges of certain parameters. The assessment can be made more realistic with the acquisition of additional site data. A screening tool approach outlined in Daniels etal. (1993), Andricevic etal. (1994), Cvetkovic et al. (1992), and Andricevic and Cvetkovic (1996) was used to model radionuclide transport from the Rulison site. The employed modeling approach incorporates real physical phenomena, such as instantaneous and/or slow release from the source, advection, dispersion, sorption, mass transfer, and possible uncertainty in the model parameters. The output is the expected concentration profile as a function of time (e.g., concentration breakthrough curves) at the compliance point downgradient from the source, as well as the uncertainty around the expected concentration resulting from the natural hydrogeologic heterogeneity, in general, and from the spatially variable groundwater velocity, in particular.

Page 9

Figure 1. Rulison site location map (from DOE, 1984). The solute flux method is described in detail by Andricevic and Cvetkovic (1996), while important elements of the approach can also be found in Daniels etal. (1993) and Andricevic etal. (1994). The following summary is derived from these sources, but the reader is directed to these references for a detailed treatment of the method. The contaminant migration process is described in the solute flux method through the Lagrangian concept of motion following a particle on the Darcy scale. In the absence of direct information on groundwater velocities near Rulison, the mean velocity, t/, is calculated using

Page 10

where K is the mean hydraulic conductivity, 7 is the mean hydraulic gradient, and ne is the mean effective porosity. Hydrogeologic parameters such as K and ne can be highly variable as a result of geologic heterogeneity. Numerous studies of the variability of hydraulic conductivity have concluded that conductivity is generally log-normally distributed (Freeze and Cherry, 1979; Hoeksema and Kitanidis, 1985). Thus, the natural logarithms of hydraulic conductivity data can be described by a normal distribution with a mean pin/^ and variance o2^. The variance represents the variability of K about its mean, and may range from near zero for homogeneous deposits to five, or higher, for extremely variable porous media (Hoeksema and Kitanidis, 1985). Because it is distributed in space, K usually has some degree of spatial correlation. The negative exponential function is often used to describe the K correlation structure because it is found to correspond to log K data and is easy to use. The correlation length of K, X, represents the distance beyond which data points show weak correlation. The higher the value of X, the greater the spatial continuity of K. When the log-normal distribution and the negative exponential covariance function are assumed, the heterogeneous, isotropic hydraulic conductivity field can be statistically characterized by three If the parameters on the right-hand side of the Darcy equation are log-normally distributed, then so is U and the estimate of the mean velocity is [%£/ = + [Xin/ - pime. The variance of the estimated mean U, a2in[/, can be calculated as the sum of the variances of the other parameters, if sufficient data are available. The estimation error in U, o2\nu, represents the magnitude of uncertainty in the estimate of U contributed by the estimation errors of K, J, and ne. The magnitude of the uncertainty in the mean velocity, o2inu, depends on the number of measurements used to estimate the parameters in the Darcy equation. In the case of independent measurements, o2\nu = where o2\nu is the variance in the velocity field and N is the number of measurements. For spatially correlated measurements, o2inu is scaledbyA^'n+oOV-l)], where @ is an averaged spatial correlation between data points. The solute flux method evaluates movement of a solute from the source to a plane perpendicular to the direction of flow. Aquifer heterogeneity' is included and represented by the variance of log-hydraulic conductivity, o2^, and the hydraulic conductivity integral scale, X. The combination of the spatial variability of aquifer properties and the uncertainty in the estimates of these properties causes the solute flux to be a random function described by a probability density function (PDF). The mean and variance of the solute flux are converted to the flux-averaged concentration needed for risk calculations by dividing by the groundwater flux, Q. The first two moments of the flux-averaged concentration are important in determining the total risk level. The larger the magnitude of variance in the flux-averaged concentrations, the larger the maximum The geology and hydrogeology of the Rulison site are described by many authors, including Hurr et al. (1969), Nork (1969), Voegeli (1969), Larson and Beetem (1970), Nork and Fenske

Page 11

(1970), Reynolds etal. (1970), Voegeli and West (1970), Voegeli andClaassen (1971a), Voegeli and Claassen (1971b), and DOE (1984). The Rulison event was conducted at a depth of 2,568 m below ground surface, within the Mesaverde Formation, a Cretaceous body of shale and sandstone (Figure 2). The Mesaverde was deposited in several near-shore depositional environments, including coastal swamp, floodplain, and marine conditions. As a result, sandstone occurs in lenticular form, and these lenses are not continuous. Aquifers in the Rulison area are generally limited to alluvium and terrace deposits, as the underlying formations are typically impermeable, and produce little water, if any (Reynolds et al., 1970). Nork and Fenske (1970) suggest that groundwater in the Mesaverde may be immobile. An inventory of wells and springs in the surrounding area indicated that only one well produced any significant amount of water from bedrock; this well is reported to be 233 m deep (completed in the Green River Formation), over 2,300 m above the shot point (Voegeli and West, 1970). Testing in well R-EX (Figure 3) indicated that no water was produced from any formation, although the investigation focused primarily below 1,828 m (Voegeli, 1969; Reynolds etal., 1970). Due to the fact that no water was obtained during testing, no site-specific information is available regarding hydrogeologic parameters such as transmissivity or groundwater velocity for any formation at the Rulison site. There are two potential sources for groundwater contamination by radionuclides at the Rulison site: surface contamination from site activities (e.g., drillback and flaring operations), and subsurface contamination resulting from the radionuclides produced by the shot. Any tritium on the land surface could possibly migrate through the vadose zone to reach shallow groundwater contained in surficial alluvium. A total of 2.824 x 1015 pCi of tritium (Colorado Dept, of Health, 1980) were released to the atmosphere during four gas flaring operations in 1970 and 1971. Monitoring conducted during the flaring suggests that radionuclides were dispersed by winds over an area covering many square kilometers. For example, tritium concentrations in air were well above background at the old control point pad, 3.8 km from the flare stack (Boysen, 1976). The site cleanup guideline for tritium in soil moisture was 3.0 x 107 pCi/L, and the few samples out of the 559 collected that exceeded that guideline (the highest being 4.7 x 107 pCi/L) were associated with known locations of spills (Eberline Instrument Corporation, 1977). During a site cleanup operation in 1976, roughly 227 L of water with a tritium concentration of approximately 2.3 x 108 pCi/L were spilled in the vicinity of surface ground zero (SGZ), but decontamination operations (consisting mainly of removing contaminated soil from the site for disposal in Beatty, Nevada) appear to have been successful. Two other minor spills at the site occurred in 1971 and 1972, and sampling of soil water from these areas during the final cleanup operation in 1976 found that decay and dilution had reduced contamination at the spill sites to negligible levels (Eberline Instrument Corporation, 1977). The …