In the framework of a collaboration between Idaho National Laboratory (INL, USA) and the Department of Astronautical, Electrical and Energy Engineeing (DIAEE) of Sapienza, the Nuclear research group of DIAEE is a contributor for the developing of the ¿Parallel and Highly Innovative Simulation for INL Code System¿ (PHISICS), a neutronic (NK) toolkit intended to provide a modern analysis tool for reactor physics investigations. PHISICS is capable to work coupled with RELAP5-3D®, the reference thermal hydraulic code for transient analyses in nuclear reactors, developed by INL. The analyses for fast reactors need the using a large number of energy groups for the neutrons, in comparison with the actual four groups as the default RELAP5-3D® neutron kinetic solver. In addition, thanks to the proposed coupling, detailed 3D thermal hydraulic (TH) transient analysis with a full 3D transport (or diffusion) neutronic will be possible.
The capability to perform TH-NK coupled calculations is one of the main open topics in the GEN IV fast reactors development. This research would realize a complete methodology by using the state-of the art tools, capable to analyze also the unprotected and dissymmetric transient analyses, with a reduction of uncertainty margins.
An optimization of the coupling, object of this proposal, is needed for this kind of calculations, because of the required very large number of nodes and time step, which would make not feasible a simulation, even by exploiting a powerful computing cluster.
A validation of this coupling will be performed through a comparison with the results of an experimental test performed during the ¿Sodium Natural Circulation Test¿ in the framework of ¿PHENIX End-of-Life Experiment¿, selected for a IAEA (International Atomic Energy Agency) benchmark.
Risk Analysis Virtual ENvironment (RAVEN) code will be used as sensitivity and uncertainty evaluation tool for the calculations and for a statistical comparison with experimental data.
The PHISICS code is designed with the mindset to maximize accuracy for a given availability of computational resources. This is obtained by implementing several different algorithms and meshing approaches which the user can choose from in order to optimize his computational resources and accuracy needs. The coupling with RELAP5-3D© has been available starting from 2013, providing new features for advanced core design coupled simulations. The PHISICS and RELAP5-3D© codes are both developed and maintained by the respective developer teams at the Idaho National Laboratories (USA).
These fantastic features are obviously contrasted by a very high computational time. Due to this reason, two main optimizations are required for this challenging application.
The first one, just implemented during a fruitful collaboration between INL and our research group, is the time step decoupling in R5-3D (TH) calculation and coupled PHISICS (NK) calculation.
The theory for the time step control has been developed and implemented in the PHISICS/RELAP5-3D© coupled code resulting in three different methodologies and two different ways to define the tolerance. More than 2,000 lines have been added to the PHISICS code and to the RELAP5-3D© interface. The three methodologies have been successfully tested on a simplified model of the Japanese HTTR, obtaining a good speedup while maintaining the relative error for the total power under 1 %.
Another developing activity is needed for the use of the 3D transient calculation, because the actual computational time (despite a cluster used for the calculation) is very long; the use of a binary format instead the actual xml format for the Cross Sections (Xsec).
This optimization will sensibly reduce the calculation time because the Xsec database is very large (many GB) and the conversion from the "xml" format to binary for each calculation time step is extremely time consuming.
The expected result is a complete methodology for the TH-NK calculation applicable in all interesting transients, which will be analyzed during the reactors design, especially in order to maximize their safety by reducing the uncertainty boundaries related to the methodology approximations.
The innovation proposed is focused on the optimization of a best platform for the TH-NK transient analysis for fast reactors, with the maximization of the accuracy in the neutronic calculation maintaining an acceptable calculation time.
From the tools point of view, PHISICS (with 3D transport solver) allows the non-symmetric transient calculation and thanks to the unlimited number of energy group a very accurate simulation of fast reactor cores is possible.
If the comparison with experimental data would be confirm the accuracy of the methodology with the use of developed tools, an advancement in the TH-NK state of the art simulation will be obtained.