Performance Prediction for Large-Scale Nuclear Waste Repositories: Final Report

Glassley, W E ; Nitao, J J ; Grant, W ; Boulos, T N ; Gokoffski, M O ; Johnson, J W ; Kercher, J R ; Levatin, J A ; Steefel, C I

Lawrence Livermore National Laboratory

关键词: Radioactive Wastes;    Solutes;    Chemical Reactions;    Forecasting;    Mineralogy;   

DOI  :  10.2172/15013584
RP-ID  :  UCRL-ID-142866
RP-ID  :  W-7405-ENG-48
RP-ID  :  15013584

美国|英语

来源: UNT Digital Library

【 摘 要 】

The goal of this project was development of a software package capable of utilizing terascale computational platforms for solving subsurface flow and transport problems important for disposal of high level nuclear waste materials, as well as for DOE-complex clean-up and stewardship efforts. We sought to develop a tool that would diminish reliance on abstracted models, and realistically represent the coupling between subsurface fluid flow, thermal effects and chemical reactions that both modify the physical framework of the rock materials and which change the rock mineralogy and chemistry of the migrating fluid. Providing such a capability would enhance realism in models and increase confidence in long-term predictions of performance. Achieving this goal also allows more cost-effective design and execution of monitoring programs needed to evaluate model results. This goal was successfully accomplished through the development of a new simulation tool (NUFT-C). This capability allows high resolution modeling of complex coupled thermal-hydrological-geochemical processes in the saturated and unsaturated zones of the Earth's crust. The code allows consideration of virtually an unlimited number of chemical species and minerals in a multi-phase, non-isothermal environment. Because the code is constructed to utilize the computational power of the tera-scale IBM ASCI computers, simulations that encompass large rock volumes and complex chemical systems can now be done without sacrificing spatial or temporal resolution. The code is capable of doing one-, two-, and three-dimensional simulations, allowing unprecedented evaluation of the evolution of rock properties and mineralogical and chemical change as a function of time. The code has been validated by comparing results of simulations to laboratory-scale experiments, other benchmark codes, field scale experiments, and observations in natural systems. The results of these exercises demonstrate that the physics and chemistry embodied in the code accurately represents the state-of-the-art in modeling these processes, and that the conceptualization of the models used in the simulations honors the primary processes that are controlling these systems. Application of the code to a wide range of important and strategic problems has been undertaken. Particularly significant are results obtained concerning the evolution of a potential high level nuclear waste repository at Yucca Mountain, Nevada. In these simulations, the results suggest that fluid movement and chemical changes will be such as to encourage solute transport around the sides of the waste emplacement tunnels, thus minimizing the potential for seepage of water and dissolved salts into the tunnels. The results also indicate that the short term response of the geological system to waste emplacement will be complex and rapid, and will be most readily detected below waste emplacement tunnels. A successful monitoring program of repository performance during the early stages of the operational period would thus benefit by coordinating design and execution of sampling strategies with a simulation tool such as NUFT-C. Such an approach would allow efficient and cost-effective sampling strategies, and would facilitate interpretation of what will surely be complex and massive data sets.