Over-view of in-situ thermomechanical experiments in clay: Concept, results and interpretation

The HADES project (High Activity Disposal Experimental Site) aims at demonstrating the technical feasibility and the long-term safety of geological disposal of reprocessed HLW (High-Level Wastes) radioactive wastes. This disposal could be realised in the Tertiary Boom clay formation below the Mol/Dessel nuclear site. Previous studies in the 80's on the geomechanical behaviour of Boom clay, at host rock temperature (15 degrees C), have demonstrated the mining capabilities of this clay. European partners have collaborated to increase the number of in-situ tests to be developed and operated from the Underground Research Facility( URF). Integrated large-scale experiments have been developed during the last four years, within the framework of the Commission of the European Communities (CEC) research contracts, in order to gain more insight into the thermal influence of heat-emitting wastes on the clay behaviour in the near field. This paper summarizes the knowledge gained from the in-situ hydro-thermo-mechanical experiments realised from the URF. The objectives and main issues of the tests are described. Information regarding selection of parameters and representativeness of their measurements is discussed. Despite their different aims and designs, three hydro-thermo-mechanical experiments are compared in order to emphasize similarities in the general behaviour of the clay massif submitted to a thermal load. It is observed that its behaviour is qualitatively similar to the reaction upon a constant total stress increase. With the information now available, it is not obvious to draw conclusions with respect to the irreversibility of some phenomena. Simple in-house analytical codes are able to simulate with a good accuracy the hydro-thermo-mechanical behaviour of clay. Nevertheless, these codes are not sufficient to describe the complexity of the phenomena that are involved in hydro-thermo-mechanical processes as pointed out by field data. In order to validate more complex models, the possibilities to improve the representativeness of the measurements are investigated.