Application of MCNP for predicting power excursion during LOCA in Atucha-2 PHWR

Atucha-2 is a Siemens-designed pressurized heavy water reactor in the Republic of Argentina. The correct prediction of the negative reactivity introduced in the moderator by an Emergency Boron Shutdown System (EBSS) is of great relevance for the correct safety evaluation of a double-ended guillotine large break LOCA scenario. During such event the EBSS is in charge to compensate the insertion of positive reactivity, caused by the void generated in the coolant channels by a sharp system pressure drop, in order to avoid severe core damage. The correct simulation of such event implies the minimization of the so called ‘‘numeric boron self-shielding effect’’ or the over-estimation of the inserted negative reactivity caused by the adoption of relatively large numerical meshes. In fact, because during the first phases of the injection, a very high concentrated boron solution is introduced in a small volume of the moderator tank, non-conservative reactivity estimation can be calculated if a ‘‘numeric boron dilution’’ is resulting by the adoption of large thermal-hydraulic and neutronic meshes. A methodology based on Monte Carlo transport code MCNP5 has been developed in order to predict power and reactivity excursions, representing a boron distribution in the moderator with different spatial resolutions. In such a way, it was possible to investigate the negative reactivity over-estimation due to the ‘‘boron self-shielding effect’’. This investigation is generally not possible by system codes used for safety analysis. The methodology was extended also to benchmark the EBSS negative reactivity insertion evaluated by the reference system code RELAP-3D/NESTLE. Final results were used for the development of the Chapter 15 of the Final Safety Analysis Report of Atucha-2 NPP, demonstrating the conservative predictions of the reference system code employed for the safety evaluation.