United States Nuclear Regulatory Commission - Protecting People and the Environment

Bases for Predicting the Earliest Penetrations Due to SCC for Alloy 600 on the Secondary Side of PWR Steam Generators (NUREG/CR-6737)

On this page:

Download complete document

Publication Information

Manuscript Completed: August 2001
Date Published: September 2001

Prepared by:
Dr. Roger W. Staehle, Consultant
22 Red Fox Road
North Oaks, MN 55127

Under Contract to:
Argonne National Laboratory
9700 South Cass Avenue
Argonne, IL 60439

J. Muscara, NRC Project Manager

Prepared for:
Division of Engineering Technology
Office of Nuclear Regulatory Research
U.S. Nuclear Regulatory Commission
Washington, DC 20555-0001

NRC Job Code Y6063

Availability Notice


The purpose of this report is to define the bases for predicting the earliest failures in a set of tubes in steam generators where the mode of failure is stress corrosion cracking. By "earliest failure" is meant the first ones that occur at a probability in the range of 0.0001 to 0.01, depending on the number of tubes at risk for a particular mode-location failure. Such early failures are far removed from the mean time-to-failure where the probability of failure is 0.5. Most testing for the purpose of predicting performance is directed, intentionally or not, toward determining the mean; however, using the mean time-to-failure as a measure of design life implies that 50% of the tubes would have failed during the design life. Failure times at probabilities of 0.0001 to 0.01 may be less than 10-4 to 10-2 of the mean time-to-failure when the shape parameter of the Weibull distribution,ß, is unity. In developing predictions of the earliest failures, knowledge of both the mean, whether measured as µ or as the Weibull scale parameter, Ø, and the shape parameter,ß, are necessary. In general, data obtained that are taken as the mean values are the same as correlations that are usually developed from testing; however, there are few data and no theory for predicting the shape parameter although certain inferences are possible with respect to the critical failure processes being random or subject to accumulation processes. To a first approximation, a conservative shape parameter for predicting the earliest failures is unity. However, larger shape parameters (an increasing shape parameter for the Weibull distribution means a narrower range of data) can be justified depending on the behavior of the most susceptible heats of material in a set. The capacity to predict the earliest failures depends on understanding the inherent variability, the metallurgically-induced variability, the environmentally-induced variability, and the effects of experimental methods on variability. Data for these contributions are analyzed. Information is available that permits developing useful inferences about the mechanistic bases for statistical parameters.

Page Last Reviewed/Updated Wednesday, June 12, 2013