| Performance Assessment of the Monticello Mill Tailings Repository |
 |
 |
| Background | Objectives | Scenarios | Conceptual Models | Total-System Framework Model | Parameter Distribution for Process Models | Results | Alternative Designs | Sensitivity Analysis | Summary and Conclusions | ||||||||||||||||||||||||
The Monticello Mill Tailings Site is located in southeastern Utah, south of the town of Monticello (see image to the right). The present climate at Monticello is “sub-humid,” with an average annual precipitation of ~38 cm (15 inches) and an average annual temperature of 7.8 ºC (46 ºF). In 1941, the Monticello mill was constructed and used to process nearly a billion kilograms of ore. By 1960, when operations were terminated, approximately 2 million cubic meters of radioactive uranium mill tailings had been left behind from the operations. To contain the mill tailings, DOE began construction of a repository south of the original mill site in 1995, and in 1996 the construction of a composite double- liner system at the base of the repository was completed. The cover was designed to mitigate the release of radon gas to the surface and to minimize water infiltration to the mill tailings. It consists of a thick topsoil layer with vegetation that can store precipitation and allow evaporation and transpiration via vegetation. This top layer overlies a coarse sand layer that acts as a capillary barrier and is intended to drain any infiltrating water laterally above a high-density polyethylene geomembrane. Beneath the geomembrane is a compacted clay layer that serves as a barrier to radon gas transport and water infiltration. The clay layer rests directly on top of the mill tailings. At the base of the repository beneath the mill tailings is a double composite-liner system composed of sand, two geomembrane liners, two geosynthetic clay-liners, and a transmissive leachate collection system. The entire repository is surrounded by Quaternary deposits consisting of sandy loam, clay, and pediment gravels. Beneath the repository, two aquifers exist--a perched alluvial aquifer, as close as several meters below the bottom of the repository and the regional Burro Canyon aquifer beneath the alluvial aquifer. The perched aquifer was contaminated by mill tailings prior to construction of the repository. The contaminants of concern include uranium, as well as its radioactive decay products (thorium-230, radium-226, radon-222), and heavy metals such as vanadium, lead-210, and arsenic. The Burro Canyon aquifer and has not been contaminated. Between the alluvial aquifer and the Burro Canyon aquifer are unsaturated layers of shale and sandstone. The water from the upper alluvial aquifer is used for irrigation purposes, but all drinking-water wells are located in the lower Burro Canyon aquifer.
- Top of Page - Back - Objectives The metrics and objectives used for the process assessment of the Monticello Mill Tailings Repository case study are as follows:
Scenarios The first step in the performance assessment is to develop relevant scenarios based on performance objectives and applicable features, events, and processes at the site. For a list of features, events, and processes at the site refer to section 4.1 of SAND Report 2002-3131: Development of a Risk-Based Probabilistic Performance-Assessment Method for Long-Term Cover Systems--2nd Edition (Full reference). A total of eight scenarios were chosen for this study:
- Top of Page - Back - Conceptual Models Once the scenarios were developed, conceptual models of the features, events, and processes for each scenario could be formulated. The conceptual models include contaminant source and release information, a description of transport mechanisms and pathways, and a definition of modeling endpoints.
Total-System Framework Models The system used for this study was the Framework for Risk Analysis in Multimedia Environmental Systems (FRAMES). FRAMES was developed by Pacific Northwest National Laboratory (PNNL) with funding from DOE and EPA. The FRAMES system provides a user-friendly platform for integrating medium specific computer models, an extensive and editable contaminant database, a powerful and flexible sensitivity/uncertainty module, and textual and graphical viewers for presenting modeling outputs.scenarios 1 and 2 were simulated separately using HELP v. 3.07 as a stand-alone code (Schroeder et al., 1994). As part of this project, HELP was integrated into FRAMES so that stochastic HELP simulations could be performed from within the FRAMES architecture. - Top of Page - Back - Parameter Distribution for Process Models Six process models, each with necessary input parameter distributions, were included in the total-system model: (1) water percolation through the cover; (2) radon gas trasport through the cover; (3) source-term release; (4) vadose-zone transport; (5) saturated-zone transport; and (6) human exposure. For details on parameter distributions for each of the process models, refer to section 4.3 of SAND Report 2002-3131: Development of a Risk-Based Probabilistic Performance-Assessment Method for Long-Term Cover Systems--2nd Edition (Full reference). - Top of Page - Back - Results Overall the results of this performance assessment illustrated how probabilistic analyses could be used to evaluate long-term performance of covers against regulatory metrics. The performance metrics can be risk-based, such as groundwater concentration or dose, or they can be prescriptive metrics such as conductivity for a particular layer of the cover. For detailed results, refer to section 4.4 of SAND Report 2002-3131: Development of a Risk-Based Probabilistic Performance-Assessment Method for Long-Term Cover Systems--2nd Edition (Full reference). An example of the results follow.
- Top of Page - Back - Alternative Designs An alternative cover design was also evaluated for the Monticello disposal site using probabilistic calculations. The performance of an evapotranspiration (ET) cover was evaluated using the same probabilistic methods described for the existing design. The ET cover consists of a simplified design consisting of only the top layer. The ET cover concept makes use of the natural moisture storage capacity of this upper layer of soil and the relatively high efficiency of natural vegetation for the removal of moisture via evapotranspiration. An assessment of landfill-cover cost versus performance can be made based on these results.
For detailed results, refer to section 4.5 of SAND Report 2002-3131: Development of a Risk-Based Probabilistic Performance-Assessment Method for Long-Term Cover Systems--2nd Edition (Full reference). - Top of Page - Back - Sensitivity Analysis A sensitivity analysis of the probabilistic assessment results can provide valuable information regarding which components of the landfill cover system and which parameters are most important to the simulated performance metric(s). This information provides understanding about the relationship between uncertainty in individual input parameters and the uncertainty in the performance of the system. In addition, knowledge of the parameters having the greatest influence on future performance can be used to help prioritize site characterization activities, to help optimize landfill cover design, and to assist in the design of monitoring systems. Using a sensitivity analysis provides the quantitative information necessary to ensure that resources are directed to those aspects of the cover system that “drive” performance and not on those aspects of cover design that have little significance. The following graph illustrates relative importance of stochastic input parameters on the simulated peak cummulative dose.
Summary and Conclusions Overall the results of this performance assessment illustrated how probabilistic analyses could be used to evaluate long-term performance of covers against regulatory metrics. The performance metrics can be risk-based, such as groundwater concentration or dose, or they can be prescriptive metrics such as conductivity for a particular layer of the cover. In both cases, probabilistic performance assessments can provide uncertainty and sensitivity analyses that identify the parameters that are most important to long-term performance. These parameters may be important for engineering design, environmental studies, and long-term monitoring efforts to assist in prioritizing their efforts. In addition, alternative designs for long-term covers can be evaluated using risk-based performance metrics that are intended to protect human health and the environment. These comparisons provide a more quantitative means to compare the performance of cover designs while factoring in additional issues such as cost and schedule. - Back - |
||||||||||||||||||||||||
|
Back to top of page Background | Objectives | Scenarios | Conceptual Models | Total-System Framework Model | Parameter Distribution for Process Models | Results | Alternative Designs | Sensitivity Analysis | Summary and Conclusions  © Sandia Corporation | Site Contact | Site Map | Privacy and Security |
