I don't think I'd want to bank on stable permafrost for a century. And one would probably want to put some care into site selection and construction as seismic

I don't think I'd want to bank on stable permafrost for a century. And one would probably want to put some care into site selection and construction as seismic hazards got us here to begin with. The residents of Hokkaido might reasonably not be warmly welcoming to this idea. I don't know enough about tritium impact to sublimation rates, but that would presumably be a problem. But point is taken about enclosed and chilled storage for 84-96 years.