PHYSICAL EVOLUTION OF A BENTONITE BUFFER DURING 18 YEARS OF HEATING AND HYDRATION

Abstract

The FEBEX in situ experiment was a full-scale test reproducing the near-field of an underground nuclear waste repository. It was performed in a gallery excavated in granite, two heaters simulated the thermal effect of the waste canisters and a bentonite barrier composed of highly-compacted blocks surrounded them, acting as buffer between the heaters and the crystalline host rock. The barrier slowly hydrated with the natural incoming groundwater. The bentonite and rock were instrumented and the sensors provided information about the state of the barrier. Half of the experiment was dismantled after five years of operation (partial dismantling), and the other half was left running for subsequent thirteen years before the complete, final dismantling. During both the partial and the final dismantling numerous samples of bentonite were taken for the on-site determination of dry density and water content. This work compares the physical state of the bentonite barrier after two different periods of time, drawing conclusions about the performance of the barrier and the factors affecting its saturation rate and evolution.

The physical state of the barrier was mostly conditioned by the heating and hydration processes, although at some points it was affected by installation particularities. The dry density gradients generated proved to be persistent, and maybe largely irreversible, since they were already observed after five years of operation and remained for another thirteen years, despite the fact that the degree of saturation at the end of the experiment was overall quite high. These gradients did not impair the performance of the barrier and its sealing ability.

To properly compute the bentonite degree of saturation the differences between the microstructural and macrostructural water density have to be taken into account, and this is essential for the proper estimation of the time needed for full saturation of the barrier. In any case, the water content changes evidenced the slowing down of the hydration rate over time.