Methodology and numerical strategy for forecasting the leakage rate evolution of nuclear reactor buildings inner containments
Mehdi Asali  1, 2, *@  , Bruno Capra  1@  , Jean-Baptiste Colliat  2@  , Jacky Mazars  3@  
1 : OXAND France  -  Site web
OXAND Group
49 avenue Franklin Roosevelt, 77210 Avon -  France
2 : Laboratoire de mécanique de Lille  (LML)  -  Site web
Ecole Centrale de Lille, Université Lille I - Sciences et technologies, CNRS : UMR8107, Arts et Métiers ParisTech
Bâtiment M6 Bvd Paul Langevin 59655 VILLENEUVE D ASCQ CEDEX -  France
3 : Sols, Solides, Structures-Risques  (3SR)  -  Site web
Université Joseph Fourier - Grenoble I, CNRS : UMR5521, Université de Franche-Comté
BP 53, 38041 Grenoble Cedex 9 -  France
* : Auteur correspondant

The containment building represents the third and last protection barrier of nuclear reactors buildings (NRB). Yet ageing mechanisms of prestressed concrete could strongly affect the tightness capacity of the inner containment of a double-wall reactor building over time. That is a major issue considering the long-term operation and the potential life extension of NRBs while ensuring safety and regulatory requirements.

Considering the size of those structures and the complexity of all interacting phenomena (such as drying, creep, shrinkage and cracking), it is very difficult from a computational perspective to build an industrial and operational tool modeling efficiently all the strong couplings occurring at different scales. In that context, a simple yet physically representative chained weakly-coupled strategy based on a macro-element discretization is implemented and applied to the VeRCoRs mock-up (scale 1:3).

The proposed methodology adapts to feedback and data collected with time. It enables operators (1) to take into account variabilities and uncertainties of main parameters in order to quantify their impact on the total leakage rate, (2) to manually introduce defects coming from visual inspections and (3) to preempt and optimize leak mitigation actions in order to avoid outage extensions and associated losses of income.


Personnes connectées : 1