# Assessing Groundwater Model Uncertainty for the Central Nevada Test Area

## Abstract

The purpose of this study is to quantify the flow and transport model uncertainty for the Central Nevada Test Area (CNTA). Six parameters were identified as uncertain, including the specified head boundary conditions used in the flow model, the spatial distribution of the underlying welded tuff unit, effective porosity, sorption coefficients, matrix diffusion coefficient, and the geochemical release function which describes nuclear glass dissolution. The parameter uncertainty was described by assigning prior statistical distributions for each of these parameters. Standard Monte Carlo techniques were used to sample from the parameter distributions to determine the full prediction uncertainty. Additional analysis is performed to determine the most cost-beneficial characterization activities. The maximum radius of the tritium and strontium-90 contaminant boundary was used as the output metric for evaluation of prediction uncertainty. The results indicate that combining all of the uncertainty in the parameters listed above propagates to a prediction uncertainty in the maximum radius of the contaminant boundary of 234 to 308 m and 234 to 302 m, for tritium and strontium-90, respectively. Although the uncertainty in the input parameters is large, the prediction uncertainty in the contaminant boundary is relatively small. The relatively small prediction uncertainty is primarily due to themore »

- Authors:

- Publication Date:

- Research Org.:
- Desert Research Institute, University and Community College System of Nevada (US)

- Sponsoring Org.:
- USDOE Office of Environmental Management (EM) (US)

- OSTI Identifier:
- 795454

- Report Number(s):
- A-2002-01

TRN: US0201326

- DOE Contract Number:
- AC08-00NV13609

- Resource Type:
- Conference

- Resource Relation:
- Conference: SPECTRUM 2002, Reno, NV (US), 08/04/2002--08/08/2002; Other Information: PBD: 14 Jun 2002

- Country of Publication:
- United States

- Language:
- English

- Subject:
- 54 ENVIRONMENTAL SCIENCES; BOUNDARY CONDITIONS; DIFFUSION; FLOW MODELS; POROSITY; SPATIAL DISTRIBUTION; STRONTIUM 90; TRITIUM; TUFF; RADIONUCLIDE MIGRATION; DATA COVARIANCES; SORPTIVE PROPERTIES; WASTE FORMS; DISSOLUTION; MONTE CARLO METHOD; GROUND WATER; GROUNDWATER; MODELING; UNCERTAINTY; CNTA; CONTAMINANTS; STRONTIUM-90

### Citation Formats

```
Greg Pohll, Karl Pohlmann, Ahmed Hassan, Jenny Chapman, and Todd Mihevc.
```*Assessing Groundwater Model Uncertainty for the Central Nevada Test Area*. United States: N. p., 2002.
Web.

```
Greg Pohll, Karl Pohlmann, Ahmed Hassan, Jenny Chapman, & Todd Mihevc.
```*Assessing Groundwater Model Uncertainty for the Central Nevada Test Area*. United States.

```
Greg Pohll, Karl Pohlmann, Ahmed Hassan, Jenny Chapman, and Todd Mihevc. Fri .
"Assessing Groundwater Model Uncertainty for the Central Nevada Test Area". United States. https://www.osti.gov/servlets/purl/795454.
```

```
@article{osti_795454,
```

title = {Assessing Groundwater Model Uncertainty for the Central Nevada Test Area},

author = {Greg Pohll and Karl Pohlmann and Ahmed Hassan and Jenny Chapman and Todd Mihevc},

abstractNote = {The purpose of this study is to quantify the flow and transport model uncertainty for the Central Nevada Test Area (CNTA). Six parameters were identified as uncertain, including the specified head boundary conditions used in the flow model, the spatial distribution of the underlying welded tuff unit, effective porosity, sorption coefficients, matrix diffusion coefficient, and the geochemical release function which describes nuclear glass dissolution. The parameter uncertainty was described by assigning prior statistical distributions for each of these parameters. Standard Monte Carlo techniques were used to sample from the parameter distributions to determine the full prediction uncertainty. Additional analysis is performed to determine the most cost-beneficial characterization activities. The maximum radius of the tritium and strontium-90 contaminant boundary was used as the output metric for evaluation of prediction uncertainty. The results indicate that combining all of the uncertainty in the parameters listed above propagates to a prediction uncertainty in the maximum radius of the contaminant boundary of 234 to 308 m and 234 to 302 m, for tritium and strontium-90, respectively. Although the uncertainty in the input parameters is large, the prediction uncertainty in the contaminant boundary is relatively small. The relatively small prediction uncertainty is primarily due to the small transport velocities such that large changes in the uncertain input parameters causes small changes in the contaminant boundary. This suggests that the model is suitable in terms of predictive capability for the contaminant boundary delineation.},

doi = {},

journal = {},

number = ,

volume = ,

place = {United States},

year = {2002},

month = {6}

}