Document Type

Technical Report

Publication Date



Geochemical modeling techniques were applied to groundwater flowpaths from Yucca Flat on the Nevada Test Site (NTS) to the Amargosa Desert, south of the NTS to estimate groundwater flow velocities for independent comparison to velocities calculated by other methods. The groundwater flowpaths examined considered flow in the carbonate aquifer beneath Yucca and Frenchman flats mixing with flow from carbonate aquifers east and southeast of the NTS and discharging at wells south of the NTS border. The approach used the computer codes WATEQ4F and NETPATH to calculate chemical speciation, determine the saturation state of mineral phases, and simulate mixing and the possible chemical reactions along the flowpath. The reactions were used to predict the δ13C and 14C values of the final downgradient water. These calculated carbon isotopic values were then compared to data to test the validity of the modeled geochemical reactions. Travel times produced by the modeling were used to calculate groundwater velocities for the various flowpaths.

Fifty-six different groundwater mixing and geochemical evolution scenarios were considered. All simulations used Ca, Mg, HCO3, and SO4 to constrain each simulation; reactive mineral phases were calcite, dolomite, gypsum, Ca-montmorillonite, and Mg-montmorillonite, along with cation exchange between Ca and Na. Mixing proportions for each simulation were calculated two different ways using Cl concentrations or δ18O values. Running the 56 different mixing and groundwater geochemical evolution scenarios with two different mixing proportions for each scenario resulted in a total of 99 simulations evaluated. Resultant models from each simulation were considered valid if the precipitation and dissolution constraints applied to the mineral phases were not ignored by NETPATH and if the modeled δ13C value was within ± 1.0% of the measured δ13C value of the final groundwater.

The results of the geochemical and isotopic modeling suggest that only a few mixing scenarios from Yucca Flat to Army #1 Water Well and Amargosa Tracer Well #2 are possible. These scenarios required mixing of carbonate aquifer water represented by Water Well C-1 with Spring Mountain water represented by Cold Creek Spring and/or a component of overlying volcanic aquifer groundwater represented by UE-5e PW-3. None of the mixing scenarios using either UE-1q or ER-6-1 groundwater to represent Yucca Flat groundwater were successful. Also, none of the mixing scenarios using USGS HTH #3 representing groundwater flow from east of the NTS were successful. This suggests that either the flowpath is not valid or that USGS HTH #3 is not representative of eastward flowing groundwater.

For the successful mixing scenarios, additional NETPATH modeling simulations with Si, Na, and K added as constraints and several major volcanic minerals added as phases were conducted to account for the possible interaction of groundwater with volcanic-rock aquifers in the geologically complex NTS flow system. The additional mineral phases included anorthite, Na-montmorillonite, K-montmorillonite, albite, potassium feldspar, and Mg/Na exchange. The increased complexity of the modeling simulations provided better constraint on the total chemical mass balance and tested the variability of travel times originally calculated with the simpler simulations.

The successful geochemical and isotopic mixing models suggest that the proportion of groundwater flowing from Yucca Flat to Frenchman Flat is larger (20 to 57 percent), the groundwater flow from the east (USGS HTH #3 representing flow from the Sheep Range) and southeast (Cold Creek Spring representing from the Spring Mountains) to Frenchman Flat is smaller (0 to 54 percent), and the flow from the overlying volcanic aquifer into the carbonate aquifer is larger (8 to 60 percent) than the present conceptual model. Only one geochemical evolution model evolving Army #1 Water Well to Amargosa Tracer Well #2 had non-changing conservative tracers (Cl and δ18O) from initial to final groundwater.

A mixing zone for the three valid mixing simulations was delineated by evaluating the potentiometric map of the NTS and vicinity. The mixing zone delineated is roughly southwest of Frenchman Lake in Frenchman Flat. Distances measured from the mixing zone were divided by the range of travel times developed during geochemical and isotopic modeling to produce groundwater velocities. Groundwater velocities from Yucca Flat to Army #1 Water Well and Amargosa Tracer Well #2 ranged from 0.9 m/yr to 10.0 m/yr. Calculated velocities from Yucca Flat are slower (0.9 to 1.3 m/yr) when overlying volcanic waters are not included in the simulations (Water Well C-1 plus Cold Creek Spring to Army #1 Water Well); calculated velocities from Yucca Flat including overlying volcanic waters are somewhat faster (1.0 to 5.8 m/yr for Water Well C-1, Cold Creek Spring, and UE-5e PW5 to Army #1 Water Well and 1.6 to 2.6 m/yr to Amargosa Tracer Well #2). The fastest velocities are from the geochemical evolution simulation from Army #1 Water Well and Amargosa Tracer Well #2 (5.6 to 10.0 m/yr). These groundwater velocities fall at the lower end of groundwater velocities for the area reported by others (1.8 to 18,288 m/yr).


Groundwater; Groundwater flow; Groundwater flow--Measurement; Nevada--Nevada Test Site; United States--Amargosa Desert


Chemistry | Environmental Chemistry | Environmental Monitoring | Environmental Sciences | Fresh Water Studies | Geochemistry | Hydrology | Water Resource Management




Link to related water research in the Historic Landscape of Nevada digital collection:,995