Faire avancer la sûreté nucléaire

La Recherchev2


3-D core modelling of RIA transient: the TMI-1 benchmark



Email :

Mot de passe :

3-D core modelling of RIA transient: the TMI-1 benchmark P. Ferraresi*, E. Studer**, A. Avvakumov***, V. Malofeev***, D. Diamond****, B. Bromley**** *Institut de protection et de sûreté nucléaire, Département de Recherches en Sécurité, CEA Cadarache - 13108 ST Paul-lez-Durance **Direction des Etudes Nucléaires, Département de Modélisation des Systèmes et des Structures, CEA Saclay - 91191 Gif sur Yvette - France ***Nuclear Safety Institute of Russian Research Center “Kurchatov Institute, Kurchatov square 1 - Moscow - Russia ****Nuclear Energy & Infrastructure Systems Division, Brookhaven National Laboratory - BNL 130 – Upton - NY 11973-5000 - USA

Type de document > *Congrès/colloque

Mots clés > code de calcul, combustible nucléaire, RIA

Unité de recherche > Département de recherche en sécurité (DRS)

Auteurs > [et al.]

Date de publication >  


The increase of fuel burn up in core management poses actually the problem of the evaluation of the deposited energy during Reactivity Insertion Accidents (RIA). In order to precisely evaluate this energy, 3D approaches are used more and more frequently in core calculations. This “best-estimate" approach requires the evaluation of code uncertainties. To contribute to this evaluation, a code benchmark has been launched. A 3D modelling for the TMI-1 central Ejected Rod Accident with zero and intermediate initial powers was carried out with three different methods of calculation for an inserted reactivity respectively fixed at 1.2 $ and 1.26 $. The studies implemented by the neutronics codes PARCS (BNL) and CRONOS (IPSN/CEA) describe an homogeneous assembly, whereas the BARS (KI) code allows a pinby- pin representation (CRONOS has both possibilities). All the calculations are consistent, the variation in figures resulting mainly from the method used to build cross sections and reflectors constants. The maximum rise in enthalpy for the intermediate initial power (33 % PN) calculation is, for this academic calculation, about 30 cal/g. This work will be completed in a next step by an evaluation of the uncertainty induced by the uncertainty on model parameters, and a sensitivity study of the key parameters for a peripheral Rod Ejection Accident.