MAAP-MELCOR Crosswalk Phase 1 Study

The Modular Accident Analysis Program (MAAP), Version 5 (MAAP5) and Methods of Estimation of Leakages and Consequences of Releases (MELCOR) are widely used integral plant response analysis computer codes. Both programs have been developed over the past 30 years for the purpose of simulating a range of beyond-design-basis accidents. The codes are benchmarked against numerous separate-effects experiments that reflect, to varying degrees, conditions expected to arise in light water reactor accidents. Such separate-effects tests, however, do not completely represent the novel physics that can arise through the interaction of multiple phenomena and physical processes at a reactor scale. Furthermore, aside from the Three Mile Island Unit 2 (TMI-2) core damage event, there is limited information available to evaluate reactor-scale behavior. Both MAAP5 and MELCOR have developed models to capture reactor-scale accident progression that, to a certain extent, extrapolate from separate-effects experiments, with assessment against the TMI-2 event only. Because of the limited information available to assess these extrapolated reactor-scale models, differences in MAAP5 and MELCOR code predictions do exist, most notably in the simulation of in-vessel core-melt progression. While these differences are not necessarily influential for the key metrics evaluated in probabilistic risk assessments, they can have a more pronounced impact on studies assessing the efficacy of accident management measures. This paper reports the first phase of a MAAP-MELCOR crosswalk designed to identify the key core-melt progression modeling differences. The results of this study highlight the impact that assumptions about reactor-scale, in-vessel core debris morphology have on (a) the potential for high temperatures to develop above the reactor core and in the main steam lines and (b) the magnitude and extent of the period for in-vessel hydrogen generation. These examples play critical roles in the evolution of challenges to the reactor pressure vessel pressure boundary and containment and are ultimately central to the evaluation of accident management effectiveness.