Impact of PWSCC and Current Leak Detection on Leak-Before-Break Acceptance

This paper assesses the effect of using primary water stress corrosion cracking (PWSCC) crack morphology parameters (roughness, number of turns, and actual flow path/pipe thickness) in determining the difference in the leakage crack length, and how the difference in the leaking crack lengths changes typical margins from past LBB submittals and published reports. Several past LBB submittal cases were selected; in addition, cases from generic LBB reports published by EPRI were also selected. The results of the analyses showed that the past submittals by nuclear steam system supply (NSSS) companies frequently used the surface roughness comparable to an air-fatigue crack with no turns and the actual flow path equal to the thickness of the pipe. This condition would give the shortest possible leakage flaw length. The roughness, number of turns, and actual flow path to thickness ratio for PWSCC cracks were determined from photomicrographs of service-removed cracks. When using the PWSCC crack morphology parameters that corresponded to the crack growing parallel to the long direction of the dendritic grains (V.C. Summer and Ringhals cases), then the leakage flaw length increased 69 percent over the air-fatigue crack length at the same leak rate. Using the same critical crack length as was used in the initial LBB submittals and the published documents, the margins on the crack length changed from 1.77 to 6.0 for the initial submittals (which we also reproduced) to 0.88 to 2.74 from our calculations for a PWSCC crack. If the crack grew in the buttered region of the bimetallic weld, then based on metallographic sections from service-removed flaws, there would be a more tortuous flow path. For this crack condition, in all but one case, the margins on the normal operating versus N+SSE crack lengths were below the safety factor of two required for LBB approval. The average margin decreased from 3.39 for the air-fatigue crack to 1.55 for the PWSCC crack growing transverse to the long direction of the dendritic grains. This was about an additional 20 percent decrease in the margin from the case of having the PWSCC grow parallel to the long direction of the dendritic grains. These results show that LBB is difficult to satisfy for PWSCC susceptible pipe using the current SRP 3.6.3 LBB approach. This LBB assessment did not consider the possible development of a long circumferential surface crack, which would be more detrimental to LBB behavior. Such cracking behavior would violate the LBB screening criterion.