Advanced Multi-Physics Methods for Safety Investigations of Research Reactors

The main objective of this thesis is to develop an analysis tool for plate-type research reactors capable of performing steady-state and transient calculations using a multiphysics approach between neutronics and thermal-hydraulics. Most research reactors have been licensed using legacy codes that have been developed specifically for those particular reactors. These codes are de¬signed with low-level solutions, which means that certain limitations may exist in terms of accuracy and modeling capabilities. However, in recent years there has been a push to promote the use of more advanced codes, with more mature and highly validated architectures, which are used in power reactor analysis. While the application of these codes to research reactors allows the simulation of steady-state and transient problems, based on heuristic methods and simplifications, the results do not represent all the complexities and details of a real model. In addition, another drawback is that, in many cases, when extended or modified codes are used in research reactors, the verification and validation process is approached in a superficial manner. This implies that a thorough evaluation of the accuracy and reliability of the results obtained is not carried out. ... mehrIn this context, to address these new challenges, this dissertation extends and modifies the Serpent2/Subchanflow code. This state-of-the-art code, normally used for power reactor analysis, undergoes a rigorous verification and validation process using relevant experimental data obtained from reference research reactors. In this way, the quality and reliability of the results obtained in the analysis of research reactors is guaranteed.
The first step is to extend the Subchanflow thermal-hydraulic code to include a heat conduction solver for plate-type fuel elements and different correlations for predicting the heat transfer and pressure drop in narrow rectangular channels, typical for plate-type research reactors, as well as modifications to account for downward flow in addition to upward flows. This extension has been validated with experimental data relevant to the research reactors. The experimental data measured under steady-state conditions have been obtained from various technical reports of the RA-6 facility and IEA-R1 reactor.
Subsequently, the static simulation capabilities of the Monte Carlo Serpent 2 neutron code have been validated by comparison with experimental data of the thermal neutron flux measured in the SPERT IV D-12/25 reactor. In this analysis, static core parameters such as the criticality state and the control rod worth curve have been evaluated. This validation allows verifying the accuracy and reliability of the Serpent 2 code in reproducing the static nuclear characteristics of the reactor, thus demonstrating its ability to accurately model and predict the behavior of the neutron flux under steady-state conditions.
The internal coupling of the Serpent2 and Subchanflow codes based on the master-slave approach is then established and validated using unique data. For the validation of the transient capability of the coupled Serpent2/Subchanflow code, experimental data from the SPERT IV D-12/25 reactor measured during transient control rod ejection tests are used. This work represents the first validation of the transient capability of Serpent2/Subchanflow using very detailed models of the SPERT IV core at the plate/subchannel level considering the local feedback between neutronics and thermal-hydraulics during static and transient simulations.
Then, the validated and coupled Serpent2/Subchanflow code is applied to analyze the generic IAEA 10 MW Benchmark reactor defined by the International Atomic Energy Agency, where a reactivity insertion accident scenario is assumed. The hypothetical transient scenarios encompass slow and fast reactivity insertions to the system.
The results predicted by Serpent2/Subchanflow show excellent agreement with the experimental values, considering a statistical uncertainty of ± 2 sigma. This is particularly apparent in the estimation of core power and critical plate temperature distribution. This high agreement enhances the reliability and accuracy of the code in predicting parameters that allow direct identification of the highest and lowest power plate, its location within the fuel assembly and core, as well as other safety parameters, such as the Departure from Nucleate Boiling Ratio (DNBR) and the temperature distribution at the hottest critical plate, without the need to apply correction fac¬tors for hot channel conditions. Consequently, the development of the Serpent2/Subchanflow code is positioned as a highly reliable and accurate numerical tool, enabling high-fidelity local analysis of MTR research reactors.