Modeling to Support Groundwater Contaminant Boundaries for the Shoal Underground Nuclear Test

Abstract

EXECUTIVE SUMMARYGroundwater flow and radionuclide transport at the Shoal underground nuclear test are characterized using three-dimensional numerical models, based on site-specific hydrologic data. The objective of this modeling is to provide the flow and transport models needed to develop a contaminant boundary defining the extent of radionuclide-contaminated groundwater at the site throughout 1,000 years at a prescribed level of confidence. This boundary will then be used to manage the Project Shoal Area (PSA) for the protection of the public and the environment.The PSA is located in north-central Nevada along the crest of the Sand Springs Range. The Shoal nuclear test was detonated on October 26, 1963. It had a yield of 12 kilotons and was located at a depth of 367 m below land surface. The Sand Springs Range is comprised of fractured granite and is bounded on the east and west by alluvium-filled valleys. Faulting is present in the range, with a prominent shear zone and other major faults striking southwest to northeast across the site. Groundwater recharge occurs by infiltration of precipitation on the mountain range, with regional discharge occurring in the valleys. A groundwater divide occurs below the upland area of the range, separating flow to the east and west. The nuclear test is located on the eastern side of the divide such that groundwater from the nuclear test area moves toward Fairview Valley.Geologic and hydrologic data to support the modeling were gathered during three different periods. First, regional hydrogeologic investigations and detailed site geology studies were conducted in support of the nuclear test in the 1960s. Second, the U.S. Department of Energy (DOE) drilled and tested four hydrologic characterization wells in 1996 that provide site-specific hydraulic parameters. After the 1996 data collection, an interim groundwater model was developed by . It was the basis for a Data Decision Analysis (Pohll et al., 1999b) that evaluated model uncertainties and identified optimum methods of reducing uncertainty. As a result of the recommendations of the DDA, the third data collection period occurred in 1999 to 2000, when DOE drilled and tested another four wells at the site, and conducted a tracer test between two of the wells.Data from these studies support a fundamental conceptual model for groundwater flow at the PSA of groundwater flow through the fractured granite, toward the adjoining valleys. Recharge is uniformly distributed across the surface of the Sand Springs Range, originating as precipitation. A hydrologic divide coincides with the upland areas of the range, forming a no-flow boundary west of the nuclear test. A region of higher conductivity and higher porosity is located around the cavity and chimney as a result of the nuclear test. A significant refinement to the conceptual model as a result of the 1999 field work is the recognition of a low-conductivity shear zone creating an impermeable no-flow boundary east of the nuclear test, dipping to the west.Alternate conceptual mode...

Modeling of Groundwater Flow and Radionuclide Transport at the Climax Mine sub-CAU, Nevada Test Site

“…The lack of a tracer test at the Climax stock leads us to use data from a tracer test at a similar subsurface flow system. Based on a tracer test in a fractured granite rock mass at Shoal, equivalent porosity was found to range between 0.027 and 0.054 (Pohlmann et al, 2004). Using this range, Pohlmann et al (2004) fit the data to a lognormal distribution with a mean of 0.025 and a standard deviation of 0.023.…”

“…A random walk particle transfer (RWPT) approach, originally formulated by Liu et al (2000) and improved upon by Pohlmann et al (2004), is used to describe the diffusion of particles from rock fractures into matrix blocks of finite size and, once in the matrix, the diffusion of particles back to the rock fractures. This algorithm is based on transfer probabilities that control the movement of particles between rock fractures and matrix blocks.…”

“…This algorithm is based on transfer probabilities that control the movement of particles between rock fractures and matrix blocks. The transfer probabilities are dependent on parameters such as a constant diffusion coefficient, average fracture spacing and aperture, matrix porosity, retardation coefficients for the matrix and fractures (spatially variable), and advection in the matrix and fractures (refer to Appendix D in Pohlmann et al [2004] for more detail). A constant diffusion coefficient of 1.0E-06 m 2 /d, representative of a generic radionuclide, is used for all transport simulations.…”

Long-term Monitoring Plan for the Shoal Underground Nuclear Test

“…Figure 5 shows the domain for the Shoal flow and transport model, which is oriented parallel to the dominant northeast-trending structural grain of the Sand Springs Range and to the shear zone that is located near Shoal surface ground zero (SGZ) and extends northeast to the land withdrawal boundary. One thousand Monte Carlo realizations of the flow field were produced as described in Pohlmann et al (2004). For the monitoring design, these flow realizations are used to perform transport simulations and obtain the geometric characteristics of the plume as it crosses a number of control planes (CPs) or siting horizons.…”

“…The RWPT code used in the Pohll et al (1999) model has been substantially updated to improve the handling of the spatial variability of dispersion and porosity, improve the velocity interpolation scheme, and incorporate a new approach for simulating the matrix diffusion process. This improved code was applied to the 2004 model presented in Pohlmann et al (2004). A brief overview of this approach is presented here.…”