An evaluation of non-volatile fission product core release models in original and upgrade ISAAC 2.0 codes

To enhance the understanding of the non-volatile fission product core release models in ISAAC (Integrated Severe Accident Analysis Code for CANDU plants) code, which is an integrated severe accident computer code for PHWR plants, and to evaluate fission product core release behavior under severe accident conditions, this report analyzes release fractions according to core release models and/or options for major non-volatile fission product species. The existing core release models in original ISAAC 1.0 version (1995) and the upgrade models in ISAAC 2.0 beta version (2003) computer codes are used as simulation tools and the reference plant is the Wolsong 1 and 2/3/4 units (CANDU-6). For the analyzed sequence, a hypothetical conservative large LOCA is selected initiated by a guillotine break in the reactor outlet header with total loss of feed water assuming that most safety systems are not available. The upgrade core release models added in ISAAC 2.0 is based on the MAAP code models and the upgrade process is performed in cooperation with FAI (Fauske and Associates Inc.). These model evaluation and upgrade will help users both to predict the difference and uncertainty among models and to have more model choices which results in easier comparison with other competitive codes. As analysis results, an ISAAC 2.0 error is found in calculating releases of Mo and Ce species for a Wolsong 1 unit, which will be modified in ISAAC 2.0 formal version. In the calculations for Wolsong 2/3/4 units, the release fractions of upgrade models were higher in the order of ORNL-B, CORSOR-M and CORSOR-O (or CORSOR-M/O) models and the release fractions of existing models were similar with the CORSOR-O (or CORSOR-M/O) model. Therefore, though more study is needed, considering both these results and recent ISP 46 viewpoints that the ORNL-Booth model was an optimum model for volatile and some non-volatile (Mo, Ru) species, the ORNL-Booth model rather than the CORSOR-M model is recommended as a default selection. The results for Wolsong 2/3/4 units appear to be almost the same with a little faster accident progression compared with Wolsong 1 unit cases because of a little higher thermal power. In conclusion, most non-volatile fission products except Sb species whose initial inventory is very small are transported together with corium under severe accident conditions while only small amount (less than maximum several percents) are released and distributed into other regions due to their non-volatile characteristics. The ISAAC 2.0 upgrade models showed good simulation for these phenomena