Résumé
The French concept of drift sealing in the framework of a IL/HL-LL waste underground disposal consists of a swelling clay based core confined between two concrete blocks. The resaturation of the core generates a swelling pressure against the drift wall that induces a compaction of the EDZ and achieves a low permeability of the system (up to 10-11m/s). The concrete blocks function is to ensure a mechanical confinement of the swelling clay core, thus orienting the swelling pressure radially. The concrete block will therefore be subject to high stresses. A specific in-situ experiment (NSC, for Sealing Core) has been designed and is currently carried out in the Meuse Haute Marne URL, to test the performance of such a sealing. The core, 5 m long and near 5 m in diameter, is confined by a concrete block 2.5 m thick. The concrete block has been designed to withstand the swelling stress of the sealing core with its own weight, the induced friction between the block, the gallery and the key. This anchoring consists of a concrete key inside the surexcavated gallery called "tooth". Its diameter is that of the gallery with 4.6 m in its narrowest part and 5.6 m in its surexcavated part. One of the challenges is to check whether the in situ high pressure generated by the sealing core can damage the concrete block and if so to what extent. An ultrasonic monitoring has been designed and installed to monitor the damage in concrete and rock vicinity. A portion of the device has been initially installed in the block volume prior to concreting and a second wall portion after concrete formwork. A 3D velocity model has been designed taking into account the P-wave velocity of the concrete (4293 m/s: average of all raypaths), in the argilites (2500 m/s: estimated from previous EDZ ultrasonic tomography) and bentonite (1000 m/s estimated from Tisato et Marelli, 2013). This model was validated with the location of the emitted pulses from the sensors and simulation of AE in argillite and bentonite (Figure 1). The error in the localization is of the order of 5 centimeters in the concrete. The large number of triggers on changes in the level of ultrasonic noise shows variations of activities that can be linked to the preliminary solicitation of the sealing core. Only a few AE have been recorded by the device. Two of them were located, they are in areas that are expected to be pressurized by the swelling of the sealing core: near the concrete raft of the gallery and key inking of the rock mass. The few AE are used to test the analysis of focal mechanisms. P-wave velocity increases on the majority of rays with a daily average of 0.2 m / s. These changes are related to the concrete hardening process. Given the large number of raypath on which the P-wave velocity is measured (over a hundred), it is intended to a volume tomography of the concrete block to see the most stressed zone and potentially with time the damaged areas. A preliminary analysis of the energy recorded in the sensors according to the length of the raypath shows that wave attenuation is relatively low and therefore that the quality factor Q of the concrete block is high (147) confirming its good mechanical characteristics which continue to improve with time (due to the drying process). Daily monitoring of the averaged quality factor is possibly a good indicator of the overall progress of the block physical properties in the short and medium term. Finally, to make operational monitoring, automatic processing of data coupled with a web-page has been developed. It will detect the AE highest energy and abnormal variations in the velocity of P waves. This system will enable the operator to quickly have information on the health of the concrete block.