Monitoring an underground radioactive waste repository is a key issue in order to ensure the long-term safety of such a project. It relies on very specific structures. A continuous monitoring of structural health implies not only numerous sensors but also a dedicated data processing. This paper presents several studies to introduce an automatic processing of continuous data for a gallery health monitoring.
The French National Radioactive Waste Management Agency (Andra) is in charge of long-term management of radioactive waste produced in France. A Planning Act charged Andra to design and to request the construction licensing for an industrial underground disposal facility called Cigeo. With a permeability of 4.10-14 m/s (for a steady-state groundwater flow), the Callovo-Oxfordian claystone located at 500 m underground in the east of France provides a natural barrier against the radioactive elements migration. However, during the implementation of the package, the first barrier will be the disposal cells.
In this framework, Andra foresees to realize two kinds of repository cells: cells that will receive Intermediate Level - Long Lived radioactive waste packages (ILLW) and High Level radioactive waste packages (HLW). They are very different: the structure of the HLW is a horizontal steel pipe of 80 cm in diameter and 80 m long and the other is a tunnel of 7.5 m in diameter and 500 m long supported by poured concrete. The consequence of drilling and introducing several kinds of radioactive waste package would be the alteration of the host rock properties and hence would ruin the natural barrier necessary for the long-term safety.
Therefore, in addition to a conventional control system that will ensure underground workers safety in case of fires or lack of oxygen by using gas detectors, for instance, Andra decided to design a specific monitoring system.