Journal article

Towards modeling and validation enhancements of the PSI MCNPX fast neutron fluence computational scheme based on recent PWR experimental data

At the Paul Scherrer Institute (PSI), a computational scheme aimed at high fidelity fast neutron fluence estimations for Light-Water-Reactors (LWRs) was in previous years developed. In this scheme, the neutron transport calculations are performed with the stochastic Monte Carlo N-Particle Transport Code MCNPX using as basis a three-dimensional pin-level volumetric source obtained from validated deterministic CASMO/SIMULATE models. While first validation studies confirmed a satisfactory performance, the strategy is to continually add new validation cases in order to achieve an enlarged and comprehensive qualification basis that also integrates latest advances in methods and/or nuclear data. Thereby, new sets of experimental data from a Swiss operating pressurized water reactor plant that became available recently were adopted for a further validation of the scheme. The first set consists of Mn-54 and Nb-93m activity measurements from so-called gradient probes located at an elevation corresponding to the top end of active fuel and increasing thereby the computational challenges because of very strong axial flux gradients. The second set consists of fluence estimates derived from Mn-54 and Nb-93m activity measurements of scraping samples extracted from the reactor pressure vessel. All dosimeters have been analyzed after the 27th cycle of operation of the reactor, providing thereby an opportunity to assess the computational methodology for modern fuel management schemes. This paper presents the validation study of the PSI fast neutron fluence scheme against these two new experimental sets including scheme adaptations and optimizations that were implemented to that aim. In particular, a new variance reduction approach, easily adaptable to any location but established here in order to tackle the challenging features of the gradient probe locations, is outlined. (C) 2015 Elsevier Ltd. All rights reserved.


Related material