Expected Thermal and Hydrothermal Environments for Waste Emplacement Holes Based on G-Tunnel Heater Experiments

INTRODUCTION Volcanic tuffs on and adjacent to the Nevada Test Site (NTS) are being considered by the Department of Energy (DOE) for the possible geo- logic disposal of high-level radioactive wastes. The Nevada Nuclear Waste Storage Investigations (NNWSI) project was established in 1977 to evaluate such disposal. Sandia National Laboratories (SNL), one of the participants in the NNWSI project, has as one of its responsibilities the development of a rock mechanics program for the Nevada site. SNL conducted three small-diameter heater experiments in welded and nonwelded tuffs in G-Tunnel on the NTS, to investigate the thermal (heat transfer) and hydrothermal (thermal- induced water migration) behavior as part of the rock mechanics program. These two aspects were investigated to increase knowledge about the behavior of tuff so that predictive capabilities are available for use by repository designers, waste-package designers, and performance analysts. Also, knowledge of the behavior of the tuff is invaluable for planning and designing future in situ experiments at Yucca Mountain, the candidate site for the repository. The evaluations of these three experiments is the subject of this paper. Designers, analysts, and experimenters are seeking a better understanding of the behavior of tuffs under the influence of thermal loads expected from radioactive-waste storage, because of two physical features peculiar to tuffs. First, tuffs are inherently porous because of the manner in which they are deposited (Smith, 1960). Tuffs utilized in G-Tunnel testing have porosities ranging from 15 to 46%, which cover the ranges of expected porosities in the target horizon at Yucca Mountain. It has been established that the G-Tunnel tuffs have similar bulk, thermal, and mechanical material properties to those at Yucca Mountain (Zimmerman et al., 1984), thus, G-Tunnel testing can be used to evaluate the effects of porosity on the thermal and mechanical behavior. The second feature is that tuffs above the water table contain pore water that influences the thermal properties of the rock; displacement of the water must be accounted for in repository designs and analyses. For example, G-Tunnel tuffs, located some 600 m above the static water level, have saturations in excess of 60%. Similar percentages are expected at the target horizon in Yucca Mountain (Zimmerman et al., 1984). The presence of and the removal of the pore water affects the thermal behavior of the tuffs, and these three experiments were designed first to determine the phenomenological behavior and then assess influences on predictive capabilities. The focus in this paper is to present the results and evaluations of the experiments as they apply to improving predictive capabilities. The thermal aspects are emphasized by comparing the measured temperatures (in and around heater emplacement holes) with calculations of predicted temperatures using numerical models, so that the effects of significant parameters can be evaluated and integrated into future predictive models. Observation of the hydrothermal aspects is needed to assess the effects of loss of pore water and subsequent migrations on the temperature fields during the heating, and also to define the expected environment in the vicinity of the outer metallic surface around the waste. The metallic boundary can be either the outer shell of the waste package system, which is the main engineered barrier for waste containment, or the borehole liner. A final decision has not been made as yet about the necessity for lining boreholes; thus, the emphasis is on predicting the characteristics of the environment around the outer metallic surface. Two aspects that limit the scope of this paper are (1) the experiments were performed at a smaller than canister scale for the purpose of