This paper describes two quarter-scale experiments (1.4 m diameter) and associated numerical analyses on granular backfill engineered barrier systems in support of the Yucca Mountain Project for the potential repository. The two configurations include a sloped capillary barter and a plain backfill. The tests involve application of dyed water as a constant line infiltration source along the top of the test set-up, monitoring water movement through the test, and measuring water exiting the experiments. A complete water balance estimate is made for each test, and observed water movement is compared with (1) detailed numerical analyses conducted using the TOLIGH2 code for unsaturated flow in porous media and (2) posttest observations. The results of the testing and analyses show that for the injection rates and configuration applied, the capillary barter design diverts a significant amount of all injected water and the TOUGH2 pretest predictions show qualitative and quantitative agreement with the experimental data.
INTRODUCTION
The Yucca Mountain Project (YMP) is currently developing the Site Recommendation (SR) design for the underground facilities including the emplacement drifts and the Engineered Barter Systems (EBS) components within them. The SR design includes alternative configurations for key subsystems that will demonstrate the ability of the design to accommodate unexpected conditions. Beginning in 1999, possible design configurations were explored using scaled models. The Pilot Scale (?fi scale) experiments described here are intended to develop an engineering assessment of the viability of key engineered barrier components for use at YMP. The first two Pilot Scale experiments are (1) a capillary barrier consisting of a fine granular material overlying a coarse granular material, and (2) a standard coarse granular backfill. Neither the capillary barrier nor backfill are part of the current repository design but may be added at a later date. The current repository design uses a drip shield to protect the waste packages from rock falls and water intrusion.
Capillary barter test data and observations are presented for the first 70 days of testing as the barrier is initially stressed by the infiltrating water. The plain backfill test was operated for only approximately 34 days as downward infiltrating water quickly contacted the simulated waste package.
A limited database of information exists regarding backfill behavior. Backfill has long been used in the mining industries to stabilize mine workings and as a convenient method for disposing of waste rock. Little is known about the thermal or coupled themalhydrological-mechanical-chemical processes that are likely to occur within backfill systems for the YMP. The only large-scale data currently available for crushed tuff backfill materials at elevated temperatures and scales appropriate for the YMP were obtained by Ryder et al. (1996), who evaluated the thermal behavior of coarse crushed tuff backfill. Fernandez and Richardson (1994) evaluated available technologies for sealing the potential repository at Yucca Mountain, including technologies for emplacing backfill materials. A significant literature base exists for the thermal behavior of two-phase (rock and air) systems or porous beds, as discussed in Kaviany (1991). Capillary barriers have been considered as alternative engineered barrier systems for backfilling of emplacement drifts .by Conca et al. (1998). Capillary barriers have also been extensively studied as alternative approaches to landfill cover systems, particularly in arid environments (e.g. Webb 1997a, 1997b; Ho and Webb 1998a, 1998b; Stormont 1995a, 1995b; Ross 1990; Oldenberg and Pruess 1993), although lit
