Paleocorrosion studies in deep sea sediments and the geological disposal of nuclear wastes

Abstract Uncertainties still surround assessment of the safety of disposal of nuclear wastes incorporated into “radwaste” matrices. This is mostly due to the long time required for radioactive decay of 237 Np. The present work explores the usefulness of an experimental approach in “paleocorrosion”, which should help in minimizing such uncertainties. In this approach polished sections of sediments containing high concentrations of natural analogues of radwaste matrices are subjected to element micromapping. Thus it is possible to characterize the long-term interactions of such analogues in their geological repositories, and to identify which generate reaction aureoles and protective and/or unprotective coatings. These analogues include grains incorporated in deep sea sediments (uraninite and quartz from the Oklo uranium ore deposit; volcanic ash particles; magnetic cosmic spherules). The present results indicate that uraninite should be a much more durable rawaste matrix than any type of glass in deep sea sediments. They also suggest that the biological “weathering” of solidified nuclear wastes should be promptly investigated. They finally provide guidelines for developing standard corrosion probes for relating the weathering of natural analogues to an evaluation of the long-term durability of radwaste matrices.