A comparative study of the thermo-hydraulics of the SNUF test facility core using FLOWNEX and RELAP5

Abstract

In the current energy mix, South Africa sources its energy from coal power, nuclear power, natural gas, hydro power, and various renewable energy sources. From the Integrated Resource Plan released by the Department of Energy in 2019 (DoE, 2019), it has been envisaged that South Africa aims to pursue an energy mix which is diversified with a reduced reliance on a single or a few primary energy sources. Also, since the current coal fleet is close to the end of its design life together with South Africa's commitment to reduced carbon emissions post 2030, an opportunity was identified for a restructured energy mix. One of the technologies chosen for continued inclusion in this new energy plan was nuclear energy. With the impending possibility of further nuclear power stations being added to the energy grid in South Africa it is necessary to focus some attention on the extent to which the localisation effort of the country would be carried out in this nuclear build program. Thermo-hydraulic analysis was identified as the area of expertise at the NWU which was to be expanded upon with a specific focus on the modelling of the core region of the Seoul National University Facility (SNUF). SNUF is a small-scale nuclear testing facility located at the Seoul National University and is based on the Advanced Power Reactor 1400 MWe (APR1400) (Lee, 2008). Simulation tools are used for thermo-hydraulics studies, with the RELAP5 being a popular benchmark simulation package in the nuclear field. Flownex®, which is a locally developed simulation tool in South Africa, is a thermo-hydraulics code which offers a very broad range of options that provide the most relevant solutions to aid in solving various system level and sub-system level simulations (M-Tech, 2013). By further developing this code in the nuclear systems arena, a good foundation could be laid towards the nuclear new build localisation effort. Therefore, the primary objective of this study was to develop a RELAP5 model of the core region and downcomer section of the SNUF facility and then to replicate this model with a corresponding Flownex® model. The ability of both models was then evaluated in predicting the thermo-hydraulic behaviour of the core region of the SNUF facility. The simulation fidelity of both the RELAP5 and Flownex® models were investigated and compared with each other, specifically the calculation of the heat transfer coefficient and the distribution of the power by the heater elements. The Dittus-Boelter correlation used in

Flownex®, produced unrealistic wall temperatures, while the Chen correlation produced good agreement with the RELAP5 results for the set of conditions used. The fluid bulk temperatures were found to differ by a maximum of 0.03% between RELAP5 and Flownex® but the wall temperatures were found to differ by 359%. It was therefore observed that when Flownex® applied the Dittus-Boelter correlation in its calculations, it produced unrealistic wall temperatures for the set of conditions used. In contrast, when a C# script was embedded within Flownex® using the Chen correlation for its calculations, good results were produced which were in agreement with the RELAP5 results. The conduction analysis study showed that discretisation of the component for both RELAP5 and Flownex® should be properly applied to get consistent results.