Abstract: Nuclear waste continuously releases the decay heat after it being stored in the disposal repository. For the safe operation of the disposal repository, it is necessary to accelerate the decay heat to dissipate to surrounding rock. A good solution to solve this problem is to improve the thermal conductivity performance of buffer layers. In this study, the natural graphite powder was mixed into Na-bentonite as backfilling material, which can exert both of their advantages: fast heat conduction for graphite and isolated function of buffer materials. A series of tests, including swelling pressure, free swelling ratio, permeability and thermal conductivity, was conducted on the graphite-bentonite mixture with different graphite contents (Rg = 5%, 10%, 15%, 20%, 30%, 40%) and different granularities (297, 150, 74, 44 μm). The hydro-mechanical-thermal performance of mixtures showed that the added graphite improved the thermal conductivity significantly, and its influence degree depended on the graphite content, initial water content and dry density. Based on parameters of hydro-mechanical-thermal performance, it was found that the optimum graphite content was about 15%?20% (Wt.), and the graphite with a particle size of 150 μm or 74 μm was more beneficial for the optimum requirements. The pore-size distribution curves of the compacted mixture showed that too large or too small graphite particles were conducive to form macropores easily with the same graphite content. The reason is that as the natural graphite particles are flat and a majority of bentonite particles or agglomerates are smaller than graphite, they are formed in the point-edge contacted mode. Especially for mixtures compacted relatively loose, there are a large number of pores at the contact interface between bentonite and graphite. Moreover, graphite is hydrophobic with a low capacity for adsorbing water molecules. Hence, water can easily pass through the surface of the graphite sheet, even after the bentonite is swollen by soaking sufficiently.

Key words: graphite-bentonite, particle size, thermal conductivity, pore size distribution