A PROBABILISTIC MECHANISTIC APPROACH FOR ASSESSING THE RUPTURE FREQUENCY OF SMALL MODULAR REACTOR STEAM GENERATOR TUBES USING UNCERTAIN INPUTS FROM IN-SERVICE INSPECTIONS
MetadataShow full item record
One of the significant safety issues in nuclear power plants is the rupture of steam generator tubes leading to the loss of radioactive primary coolant inventory and establishment of a path that would bypass the plant's containment structure. Frequency of steam generator tube ruptures is required in probabilistic safety assessments of pressurized water reactors to determine the risks of radionuclide release. The estimation of this frequency has traditionally been based on non-homogeneous historical data that are not applicable to small modular reactors consisting of new steam generator designs. In this research a probabilistic mechanistic-based approach has been developed for assessing the frequency of steam generator tube ruptures. Physics-of-failure concept has been used to formulate mechanistic degradation models considering the underlying degradation conditions prevailing in steam generators. Uncertainties associated with unknown or partially known factors such as material properties, manufacturing methods, and model uncertainties have been characterized, and considered in the assessment of rupture frequency. An application of the tube rupture frequency assessment approach has been demonstrated for tubes of a typical helically-coiled steam generator proposed in most of the new small modular reactors. The tube rupture frequency estimated through the proposed approach is plant-specific and more representative for use in risk-informed safety assessment of small modular reactors. Information regarding the health condition of steam generator tubes from in-service inspections may be used to update the pre-service estimates of tube rupture frequency. In-service inspection data are uncertain in nature due to detection uncertainties and measurement errors associated with nondestructive evaluation methods, which if not properly accounted for, can result in over- or under-estimation of tube rupture frequency. A Bayesian probabilistic approach has been developed in this research that combines prior knowledge on defects with uncertain in-service inspection data, considering all the associated uncertainties to give a probabilistic description of the real defect size and density in the tubes. An application of the proposed Bayesian approach has been provided. Defect size and density estimated through the proposed Bayesian approach can be used to update the pre-service estimates of tube rupture frequency, in order to support risk-informed maintenance and regulatory decision-making.