NPP Krško 3 inch Cold Leg Break LOCA Calculation using RELAP5/MOD 3.3 and MELCOR 1.8.6 Codes

Authors

  • Vesna Benčik
  • Davor Grgić
  • Siniša Šadek
  • Štefica Vlahović

DOI:

https://doi.org/10.37798/2019682-3197

Keywords:

MELCOR, RELAP5, code comparison, input data verification, small break Loss of Coolant Accident

Abstract

NPP Krško input deck developed at Faculty of Electrical Engineering and Computing (FER) Zagreb, for severe accident code MELCOR 1.8.6 is currently being tested. MELCOR is primarily used for the analyses of severe accidents including in-vessel and ex-vessel core melt progression as well as containment response under severe accident conditions. Accurate modelling of the plant thermal-hydraulic behaviour as well as engineering safety features, e.g., Emergency Core Cooling System, Auxiliary feedwater system and various containment systems (e.g., Passive Autocatalytic Recombiners, Fan Coolers and Containment spray) is necessary to correctly predict the plant response and operator actions. For MELCOR input data verification, the comparison of the results for small break (3 inch) cold leg Loss of Coolant Accident (LOCA) for NPP Krško using MELCOR 1.8.6 and RELAP5/MOD 3.3 was performed. A detailed RELAP5/MOD 3.3 model for NPP Krško has been developed at FER and it has been extensively used for accident and transient analyses. The RELAP5 model has been upgraded and improved along with the plant modernization in the year 2000. and after more recent plant modifications. The results of the steady state calculation (first 1000 seconds) for both MELCOR and RELAP5 were assessed against the referent plant data. In order to test all thermal-hydraulic aspects of developed MELCOR 1.8.6 model the accident was analysed, and comparison to the existing RELAP5 model was performed, with all engineering safety features available. After initial fast pressure drop and accumulator injection for both codes stable conditions were established with heat removal through the break and core inventory maintained by safety injection. Transient was simulated for 10000 seconds and overall good agreement between results obtained with both codes was found.

Downloads

Download data is not yet available.

Published

2022-07-08