![]() |
Search Options | |||
Index | Site Map | FAQ | Facility Info | Reading Rm | New | Help | Glossary | Contact Us | ![]() |
July 23, 2001 The Honorable Richard A. Meserve
Dear Chairman Meserve: During the 484th meeting of the Advisory Committee on Reactor Safeguards, July 11-13, 2001, we heard presentations by and held discussions with representatives of the NRC staff and the Electric Power Research Institute (EPRI) Materials Reliability Program regarding industry and staff actions relative to cracking and leaking observed in pressurized water reactor (PWR) Alloy 600 reactor vessel head penetrations, including control rod drive mechanism (CRDM) nozzles. This matter was also discussed during a July 10, 2001, meeting of the Materials and Metallurgy and the Plant Operations Subcommittees. During our reviews, we had the benefit of the documents referenced. Conclusions and Recommendations
Discussion Cracks were recently detected during inspections of CRDM nozzles at Oconee Units 1, 2, and 3 and Arkansas Nuclear One (ANO) Unit 1. Preliminary risk assessment indicates that the issuance of a bulletin is appropriate to request operational information from the licensees as soon as possible. The staff's in-depth analysis has raised a number of technical concerns. Although plans are in place to resolve them, the following concerns are of particular importance:
The risk assessment activities should be expanded to include rod ejection with coincident small-break loss of coolant accident and potential damage to adjacent control rods.
Inspection schedule prioritization during the upcoming refueling outages will be based on an analysis of the susceptibility of cracking of CRDM nozzles in different plants. This approach relies on the assumption that susceptibility is determined by time of service and vessel head temperature. This has led to the grouping of each PWR into one of four "bins." The 14 reactors in the two highest susceptibility bins should receive highest priority in inspections of all CRDM nozzles in 2001. Although this approach is reasonable from a technical standpoint at present, its accuracy will become apparent as inspections proceed. It is prudent to consider potential modifications to this methodology including the following:
The staff should be prepared to modify any proposed inspection program and timing depending on the results of inspections of the first group of plants (i.e., Fall 2001). These early inspection results may show that it is imperative to inspect the vessel heads of the remaining pressurized water reactors promptly. On the other hand, they may show that it is appropriate to delay the inspections of the remaining plants to allow improvements in diagnostic capabilities.
The current visual inspection process, which relies on detecting boron crystals at the top of the annulus, indicates the possible presence of circumferential cracks at the base of the annulus, but gives no information on the size and/or orientation of these cracks in the Alloy 600 material. In addition, the absence of visible boron crystals does not give complete assurance that a concentrated chemical environment at the annulus does not exist, resulting in the rapid growth of a circumferential crack. This concern could be addressed during the fall outage by a full volumetric inspection of all CRDM nozzles (i.e., including those with no boron crystals) at Oconee Units 1, 2, and 3, and ANO Unit 1. Volumetric inspections by a qualified process in such cases makes abundant sense. Assessment of the inspection methods used to detect and size cracks in CRDM nozzles and nozzle welds is necessary, especially for the circumferential cracks initiating at the base of the annulus between the CRDM nozzles and the pressure vessel head.
The inspection intervals once cracks are detected depend on knowledge of crack propagation rates as a function of the local material, environmental, and stress conditions. There are data for Alloy 600 cracking as a function of stress intensity and the temperature of the PWR primary coolant. Also, there are limited data relevant to the axial cracking in the Inconel 182 J-weld connecting the CRDM nozzle to the vessel head. The quality of these data is being evaluated by separate expert committees convened by industry and the staff. There is no similar data set relevant to the circumferential cracks that initiate in and adjacent to the J-weld and that present the greatest potential structural integrity concern. The reason for this lack of cracking data is that the local environment in the annulus between the pressure vessel and the CRDM nozzle is not known with sufficient certainty. This problem is also being addressed by the staff. Consideration of the above issues in conjunction with the issuance of the bulletin should ensure that this matter is satisfactorily addressed for the short term. The Committee wishes to be updated once the licensee responses to the bulletin are evaluated. A crucial issue confronted in the proposed bulletin is the urgency of inspections of vessel head penetrations, especially for plants thought to be less susceptible to CRDM stress corrosion cracking. Risk would be the metric best suited for determining the urgency. Unfortunately, neither the NRC's phenomenological capabilities, such as the ability to predict time-dependent stress corrosion cracking, nor the NRC's risk assessment capabilities are sufficiently developed at this time to provide defensible bases for decisions on the urgency of vessel head inspections. Sustained research to better the agency's integrated capabilities in probabilistic fracture mechanics and risk assessment will be needed to assist NRC in confronting future issues of reactor coolant system degradation. Dr. William J. Shack did not participate in the Committee's deliberations regarding this matter.
References:
|