Analysis of main steam isolation valve leakage in design basis accidents using MELCOR 1.8.6 and RADTRAD
Analyses were performed using MELCOR and RADTRAD to investigate main steam isolation valve (MSIV) leakage behavior under design basis accident (DBA) loss-of-coolant (LOCA) conditions that are presumed to have led to a significant core melt accident. Dose to the control room, site boundary and LPZ are examined using both approaches described in current regulatory guidelines as well as analyses based on best estimate source term and system response. At issue is the current practice of using containment airborne aerosol concentrations as a surrogate for the in-vessel aerosol concentration that exists in the near vicinity of the MSIVs. This study finds current practice using the AST-based containment aerosol concentrations for assessing MSIV leakage is non-conservative and conceptually in error. A methodology is proposed that scales the containment aerosol concentration to the expected vessel concentration in order to preserve the simplified use of the AST in assessing containment performance under assumed DBA conditions. This correction is required during the first two hours of the accident while the gap and early in-vessel source terms are present. It is general practice to assume that at {approx}2hrs, recovery actions to reflood the core will have been successful and that further core damage can be avoided. The analyses performed in this study determine that, after two hours, assuming vessel reflooding has taken place, the containment aerosol concentration can then conservatively be used as the effective source to the leaking MSIV's. Recommendations are provided concerning typical aerosol removal coefficients that can be used in the RADTRAD code to predict source attenuation in the steam lines, and on robust methods of predicting MSIV leakage flows based on measured MSIV leakage performance.