Publisher's Synopsis
Various forms of degradation have been observed in the containment vessels of a number of operating nuclear power plants in the United States. Examples of degradation include corrosion of the steel shell or liner, corrosion of reinforcing bars and prestressing tendons, loss of prestressing, and corrosion of bellows. The containment serves as the ultimate barrier against the release of radioactive material into the environment. Because of this role, compromising the containment could increase the risk of a large release in the unlikely event of an accident. Previous work in this area has assessed the effects that degradation has on the pressure retaining capacity of the containment vessel through structural analysis that account for degradation. These analyses have provided useful information about the effects of the degradation on the structural capacity of the containment in both deterministic and probabilistic fashions. However, the appropriate metric to use in assessing containment degradation effects during a severe accident was determined to require additional study. The previous work with probabilistic descriptions of the containment capacity were obtained from the results of the structural analysis models, and used as input for the risk models (Probabilistic Risk Assessment analyses). The risk was formulated in terms of the large early release frequency (LERF). The relative LERF values were computed for various postulated cases of degradation. In that study, an instance of degradation was treated as a change in the plant's licensing basis and assessed with U.S. NRC Regulatory Guide 1.174, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis." The Regulatory Guide provides the limits of the acceptable increases in LERF due to changes in the plant. Many of the cases of postulated degradation were those consisting of local corrosion in the liner or shell that produce leaks that do not contribute significantly to LERF, and in some cases, cause no change in LERF. Early releases due to small leaks were found to contribute to the small early release frequency (SERF). Since Regulatory Guide 1.174 does not provide guidance on the limits of SERF, additional deterministic analyses were performed in this study to assess the effects of degradation on the consequences using metrics other than LERF. This study was performed using the Sandia codes MELCOR and MACCS. These codes were used to simulate two different accident scenarios (long- and short-term station black-outs) and compute the resulting consequences for a PWR plant with a reinforced concrete containment. The structural analyses used in the previous probabilistic study were used to develop the containment behavior models for the accident simulations. Several different postulated cases of liner corrosion were considered to enable a comparison of the consequences.
