Faire avancer la sûreté nucléaire

La Recherchev2


Modeling of iodine radiochemistry in the ASTEC severe accident code: Description and application to FPT-2 PHEBUS test



Email :

Mot de passe :

Nuclear Technology / Volume 171 / Number 1 / Pages 88-107

July 2010

Type de document > *Article de revue

Mots clés > ASTEC Code, FPT2, iodine chemistry, severe accident

Unité de recherche > IRSN/DPAM/SEMIC/LETR

Auteurs > BOSLAND Loïc, CANTREL Laurent, GIRAULT Nathalie, CLEMENT Bernard

Date de publication > 01/06/2010


In the case of a hypothetical severe accident in a nuclear power plant, iodine is one of the fission products of major importance. It may be present in various gaseous forms that could be released to the environment, impacting population health. In such a case, the amount released (the so-called “source term”) has to be estimated in order to help the safety authorities protect the population from radiological consequences. This estimation is one of the main objectives of the Accident Source Term Evaluation Code (ASTEC) that is developed jointly by the French Institut de Radioprotection et de Sûreté Nucléaire (IRSN) and the German institute Gesellschaft für Anlagen- und Reaktorsicherheit. ASTEC is composed of various modules able to model the nuclear reactor behavior during an accident. One of these modules, named IODE, predicts iodine behavior in the reactor containment.
It is able to model the kinetics of about 35 chemical reactions and mass transfer processes. IODE is validated against separate effect tests, semi-integral experiments, and integral experiments. This paper presents the experimental phenomena that would take place in reactor containment in the case of a severe accident. Then, IODE is used to model the experimental gaseous concentration of organic and inorganic iodine in the PHEBUS FPT-2 test carried out by IRSN. The comparison of experimental data and the modeling show a general good agreement for inorganic iodine even if some differences are evidenced. For organic iodides the modeling is not satisfying. These differences might be explained by the deficiencies of some models and by some assumptions that still have to be validated by dedicated experiments.