Critical parameter study of PWR primary coolant pipe leak failure using probabilistic fracture mechanics

This paper discusses the critical parameters which influence the failure probabilities of a PWR primary coolant loop. Probabilistic fracture mechanics (PFM) is applied for the parametric study, using the Monte Carlo program PRAISE to predict the failure probabilities of a PWR primary coolant loop from various distributions of input parameters. Parameters such as nondetection probability of preservice and inservice inspection, vibratory stress, residual stresses, and their correlations are extensively studied. Critical crack depth which causes immediate failure are calculated in the presence of various vibratory stresses with and without residual stresses. Crack growth schemes are determined with various initial defect depth and depth-length ratio as a function of plant operation time. The results show quantitatively how PWR primary coolant loop reliability can be greatly improved by increasing the sensitivity and decreasing the uncertainty of preservice nondestructive inspection.