Resolution of Generic Safety Issues: Issue 163: Multiple Steam Generator Tube Leakage ( NUREG-0933, Main Report with Supplements 1–34 )
The NRC identified1031 this issue in June 1992 to address an NRC staff member’s concern, given in a DPO dated December 3, 1991,1936 and modified March 27, 1992,1937 about the potential for a main steamline break (MSLB) accident to cause significant primary-to-secondary leakage that could damage the reactor core. The DPO was prompted by widespread outer-diameter stress-corrosion cracking (ODSCC) at the steam generator tube support plates (TSPs) at the Trojan Nuclear Power Plant, which the DPO author claimed could not be reliably detected, and by the staff’s approval of alternate repair criteria (ARC) that would allow many tubes known to contain such cracks to remain in service.
In accordance with NRC Management Directive 6.4, “Generic Issues Program,” dated November 17, 2009,1858 the staff screened the issue and classified it as GSI-163 on June 16, 1992.1031 The principal assertion addressed by GSI‑163 was the potential for multiple steam generator (SG) tube leaks during an MSLB that cannot be isolated outside containment to lead to core damage that could result from the loss of all primary system coolant and safety injection fluid from the refueling water storage tank.
The intent of GSI‑163 was to address the adequacy of regulatory requirements relating to the management of SG tube integrity to ensure that all tubes will continue to exhibit acceptable structural margins against burst or rupture under normal operating conditions, as well as during postulated design-basis accidents (DBAs) (including MSLB), and that leakage from one or multiple tubes during postulated DBAs will be limited to very small amounts, consistent with the applicable regulations for offsite and control room doses. In contrast, any actions needed to address containment bypass scenarios due to tube failure during severe accidents would likely involve changes to accident management procedures and, perhaps, hardware modifications not involving the steam generators and, therefore, were outside the scope of GSI‑163. Similarly, iodine spiking and radiological assessment issues were outside the scope of GSI‑163. DPO issues outside the scope of GSI‑163 were managed under the SG Action Plan umbrella.
Importance to Safety
The SG tubes function as an integral part of the reactor coolant pressure boundary (RCPB) and, in addition, isolate radioactive fission products in the primary reactor coolant from the secondary coolant and the environment. Thus, the SG tubing serves a containment function as well as an RCPB function. SG tube leakage (i.e., primary-to-secondary leakage) or ruptures have a number of potential safety implications, including those associated with allowing fission products in the primary coolant to escape into the environment through the secondary system. In the event of an MSLB accident or stuck open SG safety valve, leakage of primary coolant through the tubes could contaminate the flow out of the ruptured steamline or safety valve, respectively. In addition, leakage of primary coolant through the SG tubing could deplete the inventory of water available for long-term cooling of the core in the event of an accident.
Regulatory Framework for Ensuring Steam Generator Tube Integrity
Title 10 of the Code of Federal Regulations (10 CFR), “Energy,” establishes the fundamental regulatory requirements for the integrity of the SG tubes. Specifically, the general design criteria (GDC) in Appendix A, “General Design Criteria for Nuclear Power Plants,” to 10 CFR Part 50, “Domestic Licensing of Production and Utilization Facilities,” state that the RCPB—
shall have “an extremely low probability of abnormal leakage…and gross rupture” (GDC 14, “Reactor Coolant Pressure Boundary”)
“shall be designed with sufficient margin" (GDCs 15, “Reactor Coolant System Design,” and 31, “Fracture Prevention of Reactor Coolant Pressure Boundary”)
shall be of “the highest quality standards practical” (GDC 30, “Quality of Reactor Coolant Pressure Boundary”)
shall be designed to permit “periodic inspection and testing...to assess...structural and leak-tight integrity” (GDC 32, “Inspection of Reactor Coolant Pressure Boundary”)
To this end, 10 CFR 50.55a, “Codes and Standards,” specifies that components that are part of the RCPB must meet the requirements for Class 1 components in Section III of the American Society of Mechanical Engineers Boiler and Pressure Vessel Code (ASME Code).1938 In 10 CFR 50.55a, the NRC further requires that, throughout the service life of a PWR facility, ASME Code Class 1 components meet the requirements (except for the design and access provisions and preservice examination requirements) in Section XI, “Rules for Inservice Inspection [ISI] of Nuclear Power Plant Components,” of the ASME Code, to the extent practical. This requirement includes the inspection and repair criteria of Section XI of the ASME Code. Section XI requirements pertaining to ISI of SG tubing are augmented by additional requirements in the plant technical specifications (TS).
As part of the plant licensing basis, applicants for PWR licenses are required to analyze the consequences of postulated DBAs, such as an SG tube rupture (SGTR) and MSLB. These analyses consider primary-to-secondary leakage that may occur during these events and must show that the offsite radiological consequences do not exceed the applicable limits of 10 CFR 50.67, “Accident Source Term,” or 10 CFR Part 100, “Reactor Site Criteria,” for offsite doses, GDC 19, “Control Room,” criteria for control room operator doses (or some fraction thereof as appropriate to the accident), or the NRC-approved licensing basis (e.g., a small fraction of these limits).
Operating experience has proven that SG tubing is subject to a variety of mechanically and corrosion-induced degradation mechanisms that may impair the structural and leakage integrity of the SG tubing. The licensee’s plant TS require the implementation of SG tube surveillance programs to ensure that tubes are repaired, or removed from service by plugging the tube ends, before the structural or leakage integrity of the tubes is impaired. The TS include a generally applicable depth-based tube repair limit, typically 40 percent of the nominal tube wall thickness, beyond which the tubes must be repaired or plugged. This depth-based tube repair limit is intended to ensure that tubes accepted for continued service will not leak and will retain safety factors against burst consistent with the design basis (i.e., the stress limits in the ASME Code, Section III1938) with allowance for flaw depth measurement uncertainty and for incremental flaw growth before the next scheduled inspection. The plant TS also include a limit on operational primary-to-secondary leakage, typically 150 gallons per day, beyond which the plant must be promptly shutdown.
Prioritization and Regulatory Assessment
The NRC gave the issue a HIGH priority ranking in 1997.1091 The NRC originally planned to develop a rule involving a more flexible and more effective regulatory framework for SG tube surveillance and maintenance activities (compared with TS requirements existing at that time) that would allow a degradation-specific management approach. The staff discontinued this effort in 1997 after a regulatory analysis indicated that rulemaking was unnecessary. With Commission approval, the staff began to develop a generic letter requesting that all PWR licensees submit proposed changes to their plant TS that would ensure SG tube integrity is maintained. This generic letter initiative included a draft regulatory guide and sample TS incorporating a programmatic, performance based strategy for ensuring SG tube integrity.
On December 1, 1997, the industry informed the NRC staff of an industry initiative, Nuclear Energy Institute (NEI) 97‑06, “Steam Generator Tube Integrity Guidelines,”1939 which paralleled the draft regulatory guide and which all PWR licensees had committed (among themselves) to implement. NEI 97‑061939 provided a programmatic, performance-based approach to ensuring SG tube integrity. With Commission approval, the staff put the generic letter initiative on hold and worked with the industry to identify revised TS that would be aligned with the NEI 97‑061939 initiative and that would ensure all PWR licensees implement programs to ensure that SG tube integrity will be maintained. This effort was completed in May 2005 with the NRC staff’s approval of Technical Specification Task Force (TSTF)‑449, Revision 4, “Steam Generator Tube Integrity,” dated April 14, 2005,1897 which included a new standard TS template governing SG tube integrity. In response to NRC Generic Letter 2006‑01, “Steam Generator Tube Integrity and Associated Technical Specifications,” dated January 20, 2006,1901 all PWR licensees submitted license amendment applications to change their TS in accordance with TSTF‑449.1897
The nature of the DPO evolved considerably in the years after 1991, adding additional concerns related to alternate tube repair criteria, iodine spiking assumptions for radiological analysis, severe accidents, and many other concerns. The staff prepared a DPO consideration document and provided it to the NRC’s Executive Director for Operations (EDO) on September 1, 1999. At the EDO’s request, the ACRS served as an equivalent ad hoc panel to review the DPO issues. The ACRS met with the DPO author and other members of the NRC staff and reviewed the documentation related to the DPO issues. The ACRS issued NUREG-1740, “Voltage-Based Alternative Repair Criteria,”1898 on February 1, 2001, documenting its conclusions and recommendations. By memorandum1899 dated May 11, 2001, the Office of Nuclear Reactor Regulation and the Office of Nuclear Regulatory Research developed a joint action plan to address the conclusions and recommendations in the ACRS report. This action plan and resolution of GSI‑163 was later incorporated into the NRC’s Steam Generator Action Plan (SGAP).1899, 1940 The status of the SGAP was presented to the Commission in SECY‑03‑0080, “Steam Generator Tube Integrity (SGTI)—Plans for Revising the Associated Regulatory Framework,” dated May 16, 2003,1900 and discussed at a Commission meeting on May 19, 2003. In a memorandum1941 to the DPO author dated March 5, 2001, the EDO stated that the NRC concluded that the concerns raised in the DPO were dispositioned and the DPO closed on the basis of the following three points:
|(1)||the ACRS ad hoc subcommittee’s finding that the ARC and condition monitoring program can adequately protect public health and safety|
|(2)||the ACRS ad hoc subcommittee’s conclusion that no immediate regulatory actions were necessary|
|(3)||the NRC staff’s development of an SGAP1899, 1940 to address the conclusions and recommendations in the ACRS ad hoc subcommittee’s report|
By memorandum from B. Sheron to L. Reyes dated July 5, 2007,1942 GSI‑163 was closed in the Generic Issues Program and was transferred to the Office of Nuclear Reactor Regulation for regulatory office implementation.
As of September 30, 2007, new performance-based TS requirements were in place at all U.S. PWRs. These requirements were the culmination of years of work between the NRC staff and the industry to develop a generic template for new TS requirements incorporating a programmatic, performance-based approach for ensuring SG tube integrity (70 FR 24126; May 6, 2005).1943 Each PWR licensee adopted the new TS requirements voluntarily, consistent with the generic template, and not as the result of an NRC backfit. These requirements are intended to ensure that all tubes exhibit adequate structural margins against burst or rupture for the spectrum of normal operating and DBA conditions, consistent with the original design basis. These requirements are also intended to ensure that total leakage from tubes at a plant will not exceed values assumed in licensing-basis accident analyses even if no tubes actually rupture under these conditions. In addition, licensees are required to periodically demonstrate that these structural margin and accident leakage criteria are satisfied for all tubes or, if not satisfied, to report the occurrence in accordance with 10 CFR 50.72, “Immediate Notification Requirements for Operating Nuclear Power Reactors,” and 10 CFR 50.73, “Licensee Event Report System.”
New Technical Specifications Requirements for Ensuring Steam Generator Tube Integrity
As discussed above, NRC requirements for the ISI and repair of SG tubes are contained in the plant TS. Until recently, these TS requirements were entirely prescriptive in nature, consisting of specified sampling plans for tube inspection, specified inspection intervals, and flaw acceptance limits (termed “tube repair limits”) beyond which the tube must be removed from service by plugging or must be repaired. The TS defined the SGs to be operable when the facility met these requirements.
Although these requirements were intended to ensure SG tube integrity in accordance with the plant design and licensing bases (including the applicable regulations in 10 CFR Part 50), operating experience has shown that these earlier requirements did not necessarily ensure that facilities would meet this objective. For example, the required minimum tube inspection sample sizes and eddy current test (ECT) flaw detection performance were sometimes insufficient to ensure the timely detection of flaws before the desired margins against burst and the desired degree of leak tightness were compromised. In addition, ECT measurement uncertainties and flaw growth rates sometimes exceeded those allowed for by the tube repair criteria. Also, when flaws were detected by ISI and were determined to exceed the tube repair criteria (dictating plugging or repair of the affected tubes), there was no requirement to demonstrate that the affected tubes retained the desired margins against burst and leakage integrity at the time these flaws were detected and plugged or repaired. Thus, implementation of the surveillance requirements alone did not necessarily ensure that the scope, frequency, and methods of inspection would be sufficient to ensure SG tube integrity. These earlier requirements did not directly ensure that the objective of GSI‑163 was being met.
As such, licensees experiencing significant degradation problems frequently found it necessary to implement measures beyond the minimum TS requirements in order to ensure the maintenance of adequate tube integrity consistent with the plant design and licensing bases. Until the 1990s, these measures tended to be ad hoc and licensee-specific. In the meantime, the industry and the NRC staff began initiatives to improve the effectiveness and consistency of the utility programs to ensure SG tube integrity. NEI 97‑061939 provided general, high-level guidelines for a programmatic, performance-based approach for ensuring SG tube integrity. NEI 97‑061939 references a number of detailed guideline documents from the Electric Power Research Institute for programmatic details concerning SG tube inspections, SG tube integrity assessment, in situ pressure testing, and monitoring of operational primary-to-secondary leakage. The NEI 97‑061939 approach was inspired by, and is similar to, an approach developed by the NRC staff in a draft regulatory guide, “Steam Generator Tube Integrity,” published as DG‑10741944 in December 1998.
The new TS requirements1943 address the previous lack of a direct relationship between the TS surveillance requirements and SG tube integrity. The new TS requirements require implementation of an SG program that focuses directly on maintaining tube integrity and periodically verifying that the program continues to be successful in meeting this goal. This required SG program addresses the central objective of GSI‑163 in that it is intended to ensure that all SG tubes will exhibit acceptable structural margins against burst or rupture under normal operating conditions, as well as during postulated DBAs (including MSLB), and that leakage from one or multiple tubes during postulated DBAs (including MSLB), will be limited to very small amounts, consistent with the applicable regulations for offsite and control room dose.
New performance-based TS requirements 1943 include a new limiting condition for operation (LCO) that tube integrity shall be maintained with an associated surveillance requirement and that tube integrity shall be verified in accordance with the SG program. The key elements of the SG program are defined in the TS administrative controls, which specify that an SG program shall be established and implemented to ensure that SG tube integrity is maintained. The TS do not provide specific details on how this objective is to be met; it is the licensee’s responsibility to ensure that the program will meet the stated objective. Industry guidelines in NEI 97‑061939 and other guidance referenced therein provide a resource to utilities for meeting this objective. However, the TS do define a general programmatic framework for the SG program, which must include the following elements:
- performance criteria for SG tube integrity
- provisions for condition monitoring
- provisions for tube repair criteria
- provisions for SG tube inspections
- provisions for monitoring primary-to-secondary leakage
The TS define three different types of performance criteria for evaluating SG tube integrity:
|(1)||structural integrity criteria|
|(2)||accident-induced leakage (primary-to-secondary) criteria|
|(3)||operational primary-to-secondary leakage criterion|
The condition of the tubes relative to the structural integrity criteria and the accident-induced leakage criteria is evaluated periodically, based on inservice inspection results, in situ pressure tests, or other means before the plugging of tubes to confirm that these criteria are met for all tubes. This periodic evaluation is termed a condition monitoring assessment and is performed during each plant outage during which the SG tubes are inspected, plugged, or repaired. The operational leakage criterion corresponds to the TS LCO limit for primary-to-secondary leakage. Primary-to-secondary leakage is monitored while the plant is operating. Should this leakage exceed the TS LCO limit, the plant must be shutdown in accordance with the TS. The structural integrity criteria define the minimum factors of safety against burst or plastic collapse that must be maintained for all tubes under normal operating and DBA loading conditions. These safety factor criteria were developed to be consistent with the safety factors that are ensured by the stress limits in ASME Code, Section III1938 (i.e., the design basis). These safety factor criteria include, for example, a safety factor of 3 against burst under normal steady state full-power operation primary-to-secondary pressure differential and a safety factor of 1.4 against burst applied to design-basis accident primary-to-secondary pressure differentials.
Even if all tubes exhibit safety factors in accordance with the structural integrity performance criteria, tubes with localized flaws can leak under normal operating and accident conditions, without burst or collapse. The central DPO concern1936, 1940 was that such leakage from multiple tubes may lead to significant radiological releases or core melt. The accident-induced leakage criteria address this concern by limiting the allowable total accident-induced leakage in each SG (as determined during condition monitoring assessments) to values assumed in the licensing basis accident analyses to demonstrate that offsite and control-room doses meet applicable regulatory requirements. The accident-induced leakage criteria values are a small fraction of the values associated with a ruptured tube or values that affect peak clad temperature and the likelihood of core melt.
Given the TS LCO operational leakage limit, a separate performance criterion for operational leakage is unnecessary for ensuring prompt shutdown if the limit is exceeded. However, operational leakage is an indicator of tube integrity performance, although it is not a direct indicator. It is the only indicator that can be monitored while the plant is operating. Maintaining leakage within the limit provides added assurance that the plant is meeting structural and accident leakage performance criteria. Thus, inclusion of the TS leakage limit among the set of tube integrity performance criteria is appropriate from the standpoint of completeness of the performance criteria.
The new TS require that the SG program include periodic tube inspections. This includes a new performance-based requirement that the scope, methods, and intervals of the inspections ensure the maintenance of SG tube integrity until the next inspection. This performance-based requirement complements the requirement for condition monitoring in ensuring that tube integrity is maintained. The requirement for condition monitoring is backward looking in that it is intended to confirm that tube integrity has been maintained before the time the assessment is performed. The inspection requirement, by contrast, is forward looking, as it is intended to ensure that tube inspections, in conjunction with plugging of tubes, are performed so as to ensure that the plant will continue to meet the performance criteria until the next SG inspection. Tube inspections would be followed again by condition monitoring at the next SG inspection to confirm that the performance criteria were in fact met, and so on.
The new TS performance-based requirements are supplemented by a number of prescriptive requirements relating to minimum sample sizes for tube inspections, maximum allowable inspection intervals, and tube repair criteria. Even though the new TS compel implementation of a performance-based program (including inspections and plugging) that ensures tube integrity, the prescriptive requirements pertaining to inspection sample sizes and inspection intervals provide added assurance of tube integrity should new or unexpected degradation mechanisms or changes in previously observed flaw growth rates occur. The tube repair criteria provide added assurance that degraded tubes will be plugged or repaired before the integrity of these tubes is impaired.
For the tube repair criteria, the new TS retain the standard depth-based limit of 40 percent of the nominal tube wall thickness. In addition, any plant-specific requirements pertaining to the use of alternate repair criteria in the old TS have been carried over to the new TS.
The NRC regional offices conduct periodic inspections (typically during each outage inspection) to assess the effectiveness of licensee programs for ensuring tube integrity in accordance with the technical specifications. These regional inspections are performed in accordance with the NRC Inspection Manual, Inspection Procedure 71111.08, “Inservice Inspection Activities,” dated November 9, 2009.1945
Failure to meet any of the TS tube integrity performance criteria is reportable pursuant to 10 CFR 50.72 and 50.73 in accordance with guidelines in NUREG‑1022, Revision 2, “Event Reporting Guidelines: 10 CFR 50.72 and 50.73,” issued September 2004.1946 In addition, the NRC regional office would follow up on such an occurrence, as appropriate, consistent with the NRC Reactor Oversight Program1947 and the risk significance of the occurrence.
Finally, the new TS requirements include a requirement that the following information be submitted to the NRC within 180 days of each SG inspection:
a description of the inspections performed
the results of these inspections
the active degradation mechanisms found
the number of tubes plugged or repaired
the results of the condition monitoring assessments (vis-à-vis the tube integrity performance criteria)
The NRC staff reviews these reports for the purposes of monitoring SG tube degradation trends and assessing the effectiveness of licensee programs. These reviews, like the regional office inspection reports, are documented and publicly available.
Effectiveness—Steam Generator Program
Although the new TS requirements have only been in place since 2005–2007 (depending on the plant), all PWR licensees have been implementing the basic performance-based elements of these requirements since 1999–2000 following their commitment to the industry’s NEI 97‑061939 initiative. The NEI 97‑06 initiative was an evolutionary change in licensee programs for ensuring tube integrity, because the effectiveness of these programs has been constantly evolving and improving since the 1970s. Industry guidelines relating to secondary water chemistry control and inservice inspection have been available since this period and have been frequently updated to reflect research findings, technology developments, and operating experience. In the late 1980s, licensees became sensitized to the need to monitor operational primary-to-secondary leakage on as close to a real-time basis as possible to provide added assurance of plant shutdown before rupture of a leaking tube. Industry guidelines for monitoring and responding to operational primary-to-secondary leakage have been available since the mid-1990s. Another trend dating from the 1970s was an ever-increasing awareness among licensees of the need for their SG programs to address tube integrity in addition to satisfying TS surveillance requirements. Industry guidelines for tube integrity assessment became available in the mid-1990s and led to improved consistency, rigor, and completeness of licensee tube integrity assessments.
In parallel with these SG programmatic improvements, tube integrity reliability appears to have improved significantly since the 1970s. This is evidenced by the sharply declining trends in frequency of SGTR and of forced shutdowns because of SG leakage.1948 The use of tubing that is more resistant to stress-corrosion cracking (i.e., thermally treated alloy 600 and 690 tubing in lieu of the mill annealed (MA) alloy 600 tubing used in SGs manufactured through the late 1970s) in new (post-1970s) and replacement SGs has been responsible for some of this improvement. However, even plants with alloy 600 MA tubing have experienced sharply improved performance trends in forced outage and SGTR frequencies. The improving trends for the plants with alloy 600 MA tubing are due to a variety of factors relating to tube integrity management programs. These include more effective secondary water chemistry programs and steps taken to control copper and impurity ingress from the feed system. These improvements in water chemistry programs, however, are not the only reason for the improving tube integrity trends, because even with the improved water chemistry programs, plants with alloy 600 MA tubing have continued to experience extensive degradation, including stress-corrosion cracking. As a result, it is clear that improved, more effective inspection programs and tube integrity management have played very important roles in reducing the frequency of forced outages because of SG leakage and SGTRs.
Even with the improved SG programs, operating experience provided examples of tube flaws that were not detected by inservice inspection. These flaws were later discovered to not satisfy the required structural and accident leakage integrity margins. There have been three such occurrences from 2000 to 2009:
Indian Point 2—SGTR event in February 2000.1949 This represented a failure to meet structural and leakage integrity performance criteria.
Comanche Peak 1—Failure to meet structural and leakage integrity performance criteria in Fall 2002, as determined by in situ pressure testing during condition monitoring.1950
Oconee 2—Failure to meet structural integrity performance criteria in fall 2002, as determined by in situ pressure testing during condition monitoring.1951
Another occurrence, at Crystal River 3 in 2003, involved an apparent failure to satisfy the accident leakage criterion.1952 The initial finding that the accident leakage rate exceeded the performance criteria was based on use of a leakage calculation model that was overly conservative. In 2005, the NRC staff approved a more realistic, but still conservative, leakage model than that used in the 2003 calculation.1953
Of these three occurrences, only the tube that ultimately ruptured under normal operating conditions at Indian Point would likely have ruptured had an MSLB event occurred during a several-month period preceding the SGTR event. This experience indicates that the frequency at which tubes may be vulnerable to rupture (or leakage from multiple tubes comparable to a ruptured tube) under MSLB is well within the conditional probability value of 0.05 assumed in NRC risk studies.681, 1954
On the basis of the above, the staff concludes that SG program improvements in the areas of inservice inspection and tube integrity management and assessment have contributed significantly to improved SG tube integrity performance. Improved water chemistry practices and the increasing number of PWRs with SGs of improved design and more stress corrosion cracking resistant tubing have also contributed to this trend.
Disposition of ACRS (DPO Review Panel) Recommendations
An ACRS ad hoc subcommittee served as the NRC DPO review panel for the DPO, documenting its conclusions and recommendations1800, 1898 in February 2001. This section addresses the subcommittee’s conclusions and recommendations as they relate to the adequacy of NRC requirements to ensure that tube structural and leakage integrity will be maintained such that there is reasonable assurance that public health and safety will continue to be maintained.
Voltage-Based Alternate Repair Criteria Issues
Background on Voltage-Based Tube Repair Limits
The DPO concerns were first prompted by the finding of intergranular attack (IGA) and ODSCC at the tube-to-TSP intersections at the Trojan nuclear power plant in 1991, the challenges that were encountered in reliably detecting such flaws, and consideration being given at the time to allowing some tubes with greater than 40 percent through-wall flaws to remain in service. At Trojan, and subsequently at many other PWRs with Westinghouse-designed SGs, ECT inspections identified hundreds of indications at the tube-to-TSP intersections. Examination of tube specimens removed from the field (i.e., pulled tube specimens) identified the degradation mechanism as stress-corrosion cracking initiating from the ODSCC, with varying degrees of general IGA. These examinations showed the ODSCC IGA to be confined to within the 0.75‑inch thickness of the TSPs. Burst testing of these specimens revealed the failure mode to be axial.
ECT techniques were not capable of accurately sizing the depth of the ODSCC IGA flaws relative to the applicable TS tube repair limit of 40 percent of the nominal tube wall thickness. For this reason, it was necessary to assume that all detectable ODSCC/IGA indications exceeded the 40-percent tube repair limit, thus necessitating the plugging or repair of all affected tubes. However, the number of affected tubes at each plant ranged from hundreds to, sometimes, thousands of tubes. This had significant economic implications for the industry. Plugging such a large number of tubes would potentially significantly shorten the useful life of the SGs, after which SG replacement would be necessary. Depending on the plant, the useful SG lifetime could potentially expire before replacement SGs were available. Sleeve repairs at each TSP intersection did not appear to offer a practical, cost-effective alternative. For this reason, around 1990 the industry began to investigate alternative approaches to ensuring the integrity of tubing affected by ODSCC IGA at the TSPs.
The 40‑percent, depth-based tube repair limit is intended to ensure that tubes accepted for continued service will not leak and will retain safety factors against burst consistent with the design basis (i.e., the stress limits in ASME Code, Section III1938) with allowance for flaw depth measurement uncertainty and for incremental flaw growth before the next scheduled inspection. These safety factors include a factor of 3 relative to normal operating pressure differential (between primary system and secondary system pressures) and 1.4 relative to postulated accident pressure differentials. The 40‑percent limit was developed with the conservative assumption that degradation results in uniform thinning of the tube wall thickness in both the axial and circumferential directions. Burst testing of pulled tube samples with ODSCC IGA flaws showed the degrading effect of these flaws on tube burst pressure to be significantly less than is the case for tubes that are uniformly thinned to the same depth. This result is explained by the limited axial extent of the flaws (i.e., less than the thickness of the TSP (0.75 inches)), the nonuniformity of the depth profile, and the often segmented rather than continuous nature of the cracks. In some cases, crack segments could penetrate up to 100 percent through the tube wall while maintaining structural safety margins consistent with the design basis.
It was also observed from burst and leak tests performed on the pulled tube samples that those ODSCC IGA indications that had exhibited low-voltage ECT signals in the field tended to exhibit high burst strengths and low potential for leakage compared to indications exhibiting higher voltage responses. This observation led the industry to develop a database from pulled tube specimens and lab specimens correlating voltage response of the ODSCC IGA indications with burst strength, probability of leakage (POL) under MSLB differential pressure, and leak rate (given that leakage occurs) under MSLB differential pressure. This database was used as the basis for developing voltage-based ARC. Statistical/mathematical models were developed for each of these correlations. The burst and leak rate correlations were represented by a mean regression curve and an associated variability distribution to capture the scatter or variability of the data. The POL correlation was modeled as a log-logistic function with an associated uncertainty distribution. Separate sets of correlations were developed for SGs with 7/8-inch diameter tubing and 3/4-inch diameter tubing respectively.
The NRC approved the voltage-based ARC on an interim basis for Trojan in 1992, and subsequently for other plants. In 1995, the staff issued Generic Letter 95‑05, “Voltage-Based Repair Criteria for Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking,” dated August 3, 1995, with guidance for submitting applications for permanent voltage-based ARC amendments.1804 Over the next several years, the NRC approved voltage-based ARC TS amendments for 27 units. However, because of subsequent SG replacements at many of these plants, only three units continue to have TS that allow implementation of the voltage-based ARC.
The supporting databases for the burst and leakage correlations are periodically updated as additional data become available. The conditional leak rate correlation for 7/8-inch diameter tubing has been, and continues to be, weak. For this reason, use of the linear regression fit of the conditional leak rate correlation is subject to demonstrating that the fit is valid at the 5‑percent level with the “p-value” test. If this condition is not satisfied, the linear regression fit is assumed to be constant with voltage.
When implementing the voltage-based ARC, an upper limit on voltage is established such as to provide a factor of 1.4 against burst under MSLB conditions. (The TSP constrains radial expansion of the tube under normal operating conditions, ensuring that the factor-of-3 criterion for normal operating conditions is met.) For MSLB, the TSP is conservatively assumed to be displaced axially by hydraulic blowdown loads. Thus, the TSP is assumed not to constrain radial expansion and burst under MSLB conditions. This voltage limit is deterministically based, corresponding to the voltage in the burst pressure versus voltage correlation where the lower 95‑percent prediction interval burst pressure equals 1.4 times the MSLB differential pressure. This voltage is adjusted downward on a plant-specific basis to allow for voltage growth between inspections and for voltage measurement variability. The voltage growth value is a generic or plant specific value, whichever is larger. Plant-specific values are based on the average value observed during the most recent one or two inspection intervals. The voltage measurement variability is an upper 95‑percent cumulative probability estimate based on industry data.
Given the scatter and variability of the burst and leakage correlations, the growth rate distribution, and the voltage measurement variability distribution, there remains a probability that an indication at the upper voltage limit may burst at a pressure less than 1.4 times MSLB pressure. For this reason, it must also be demonstrated that the conditional probability of one or more tubes bursting under MSLB conditions, from among the entire population of indications projected to exist at the next scheduled inspection, is less than 0.01 (operational assessment). This forward projection (using a Monte Carlo sampling method) is performed assuming that the probability of detection (POD) for ODSCC IGA flaws during the current, or most recent, inspection is 0.6, independent of voltage amplitude. This conditional probability criterion was developed 681 to ensure that implementation of the voltage-based ARC would not significantly increase risk. In addition, a similar analysis is done during each inspection based on the as-found indications (without consideration of voltage growth) to confirm that the conditional probability criterion was met during the prior period of operation (condition monitoring assessment).
All tubes with bobbin coil indications exceeding the upper voltage criterion must be plugged. In addition, lower limits on voltages of 1 volt for 3/4‑inch diameter tubing and 2 volts for 7/8‑inch diameter tubing have been established for conservatism. Tubes with bobbin indications higher than the lower voltage limit, but less than or equal to the upper voltage limit, may be left in service if rotating probe inspections do not confirm the bobbin coil indication.
When implementing the voltage-based ARC, the licensee must also demonstrate that leakage under MSLB conditions will not exceed values assumed in the licensing-basis accident analyses. The MSLB leakage assessment is performed in a similar manner as the conditional probability of burst analyses, except that the Monte Carlo sampling is performed on the POL and leak rate correlations instead of the burst correlation. This analysis yields a probability distribution of leak rates. The MSLB leak rate is the upper 95‑percent percentile value from the distribution, evaluated at an upper 95‑percent confidence bound.
ACRS Ad Hoc Subcommittee’s Conclusions
The ACRS ad hoc subcommittee’s conclusions supported the technical adequacy of the voltage-based ARC1804 subject to two recommendations described later. Specific conclusions included the following:
“There is a need for ARCs.” The subcommittee did not focus on the economic benefits of ARCs (from avoided tube plugging and repairs and extended SG life), but rather on the need for different plugging criteria to address the different types of degradation being encountered in the field. The subcommittee noted that ODSCC in the tube at the TSP intersections is difficult to detect and characterize relative to the standard 40‑percent, depth-based repair criterion. The subcommittee noted the conservatism of the standard 40‑percent, depth-based criterion for this type of degradation and the attractiveness of voltage-based ARCs for this type of degradation, especially if supplemented by characterizations that ensure flaws producing the signal meet explicit and implicit assumptions about the possible growth and behavior of flaws.
The staff notes that the voltage-based ARC includes specific requirements for verifying that these assumptions continue to be valid. For example, the assumptions that the ODSCC has a predominant axial orientation and that it is confined to within the thickness of the TSP is verified by laboratory examinations of representative tube samples, which are periodically removed from the SGs, and by rotating coil inspection of all tube-to-TSP intersections with bobbin coil responses exceeding 1 volt for ¾‑inch diameter tubing and 2 volts for 7/8‑inch diameter tubing.1804 As discussed later in response to the subcommittee’s recommendation that the staff should develop a program to monitor the predictions of flaw growth for systematic deviations from expectations, the staff believes that any systematic deviations from expectations in flaw growth will be identified and addressed in the staff review of the reports submitted after each outage during which the voltage-based ARC is implemented.
“Plants will be operated with flaws in the SG tubes and this need not be risk significant.” The subcommittee noted that, provided risk is managed properly, it is acceptable to operate plants with known, small flaws as well as undetected flaws in the SG tubes. As discussed in the previous section, the staff notes that the new technical specifications ensure low risk by requiring implementation of an SG program that ensures that all tubes satisfy the performance criteria for structural and leakage integrity. The staff also notes that the performance criteria associated with implementation of voltage-based ARCs differ somewhat from those in the new generic TS (which are applicable when not implementing the voltage-based ARC). The ARC-specific performance criteria include a conditional probability criterion for induced tube ruptures to ensure that the conditional probability for induced ruptures is within values assumed in past risk assessments.
The subcommittee also noted that additional, defense-in-depth management of risk can be achieved by restricting known flaws in the tubes to those unlikely to grow significantly during an operating cycle. The staff agrees, noting there have been cases in which preventive plugging of tubes not in violation of the voltage-based repair criteria was performed to prevent high-voltage growth from occurring during the next operating cycle.
“The general features of the procedures that the staff has established to limit the number and size of flaws left in operating SG tubes are adequate.” The subcommittee found no fault with the concept of voltage-based ARC and found the voltage repair criterion of 1 volt for ¾‑inch diameter tubing and 2 volts for 7/8‑inch diameter tubing to be conservative. The subcommittee did not attempt to reach conclusions about occasions when the staff granted exceptions to these criteria, except to note that these exemptions should have been accompanied by more complete risk analyses. The staff notes that the 1- and 2‑volt criteria are lower threshold limits and that all indications below these limits are acceptable.1804 However, the voltage-based ARC includes higher upper bound voltage threshold limits, which are determined in accordance with the voltage-based ARC methodology in Generic Letter 95‑05.1804 This methodology is based on satisfying the voltage-based ARC performance criteria, including the criterion on conditional probability of induced ruptures during MSLB, with allowance for voltage measurement variability and voltage growth rate distribution. As noted by the subcommittee, the staff approved an increase in the lower voltage threshold limit to 3 volts for three plants (with ¾‑inch diameter tubing) where a number of tubes were expanded against the tube support plates for purposes of limiting axial support plate deflection under MSLB conditions.
These changes are no longer in effect, because these plants have undergone SG replacement and the provisions for implementing voltage-based ARCs have been eliminated from the TS for these plants. Under these changes, the licensees were required to demonstrate that the conditional probability of burst criterion continued to be met. Thus, the staff believes there were no risk implications associated with the 3‑volt criterion.
“The general features of the condition monitoring program are adequate.” The subcommittee found the general approach used to assess the probabilities of leakage and tube burst to be conservative. The subcommittee felt that the development of empirical correlations of burst pressure and leakage with voltage amplitude are technically defensible. The subcommittee found no evidence that the supporting databases were flawed in any nonconservative, systematic way. The subcommittee felt that the constant POD assumption in the voltage-based ARC methodology approved by the staff could potentially deter technical improvements, but acknowledged that the staff would consider approving alternative POD assumptions that recognize that POD can depend on flaw size (with a sufficient technical justification). In fact, the NRC staff has approved an alternative1955 to the constant POD model that replaces the POD parameter with a parameter known as the “probability of prior cycle detection.” This empirical, plant-specific parameter is voltage dependant and relates the total number of indications found during a given inspection in a given voltage bin to the subset of these indications that were also detected during the previous inspection.
The subcommittee concluded that the condition monitoring program that licensees adopt in conjunction with the ARC, although not perfect, can produce a better understanding of the conditions and vulnerabilities of steam generators and afford additional protection to the public than has been possible in the past. The staff agrees with this conclusion and notes that the voltage-based ARC was an important step that contributed to the ultimate development of the performance-based strategies in DG‑1074,1944 NEI 97‑06,1939 and the new TS for ensuring SG tube integrity.
ACRS Ad Hoc Subcommittee’s Recommendations
The following two recommendations accompanied the above conclusions by the ACRS ad hoc subcommittee:
|(1)||ACRS ad hoc subcommittee recommendation:1898 “The databases for 7/8‑inch diameter tubes need to be greatly improved to be useful.”|
The subcommittee observed that the correlation of leakage with voltage for the 7/8‑inch diameter tubes does not correspond well with that for 3/4‑inch diameter tubes. The subcommittee could identify no mechanistic reasons why this should be the case. The subcommittee felt that the poor correspondence may reflect stochastic scatter and the limited size of the database. Therefore, the subcommittee felt that the staff should consider requiring a near-term expansion of the database.
Evaluation of the leakage data has not led to a conclusive explanation for the poor correlation of the 7/8‑inch diameter tube leakage data compared with ¾‑inch diameter tube leakage data.
The poor correlation notwithstanding, the methodology for assessing leak rate is conservative for the following reasons:
Pre-pull voltage responses are used for the correlations. If the crack tears as a result of the tube pull operation, the measured voltage is expected to be higher than if the tube were not damaged.
The leak rate analysis yields a probability density function of total leak rate (using Monte Carlo sampling of the input parameter distributions and leak rate distributions as a function of voltage) for a given population of voltage responses. This probability density function is evaluated at the upper 95th percentile value at an upper 95‑percent confidence bound vis‑à‑vis the applicable performance criterion for accident leakage.
If a statistical correlation between leak rate and voltage cannot be demonstrated to within criteria specified in Generic Letter 95‑05,1804 Generic Letter 95‑051804 specifies that leakage shall be treated as independent of voltage, which is conservative (because most indications left in service are relatively low-voltage indications, which tend to leak less than the mean).
On the basis of the above, the staff concluded1956 that item number 3.7 (the leakage correlation issue) is adequately addressed and is, therefore, closed. In addition, the staff stated that it would continue to assess the leakage correlations as more data are added to the database. The ACRS reviewed these findings.1862 The ACRS continues to believe that the leakage correlation for 7/8‑inch diameter tubing should not be used, which is contrary to the staff’s position, as stated above. As previously noted, the voltage-based ARC, including the leakage correlation, continues to be used at one plant with 7/8‑inch diameter tubes (as of February 2009) and is approved for use at two additional plants (but not currently implemented). However, the ACRS stated that it agrees with the staff that the choice of a 2‑volt limit for 7/8‑inch diameter tubes is conservative with respect to the risk posed and that item number 3.7 should be closed.
|(2)||ACRS ad hoc subcommittee recommendation:1898 “The staff should establish a program to monitor the predictions of flaw growth for systematic deviations from expectations.”|
One step in the voltage-based ARC methodology is the prediction of the change in the voltage distribution over an operating cycle. The subcommittee noted that this is done assuming a linear change in the distribution with time. The subcommittee noted that this is inconsistent with behavior of stress corrosion cracks observed in NRC research. These studies show that cracks grow slowly until they interlink, after which it is possible for flaws to grow very quickly. Flaw growth, then, is inherently nonlinear and can be treated as linear with time only in a bounding manner. The subcommittee stated that, even then, stochastic variability means that occasionally individual cracks can violate even very conservative linear bounds. Thus, the subcommittee found that it will be important for the staff to be vigilant in monitoring the implementation of the ARC to watch for such systematic errors in the crack growth predictions.
In accordance with GL 95‑05,1804 licensees submit information related to the structural and leakage integrity of the tubes within 90 days (the 90‑day report) of completion of the steam generator tube inspections. The information submitted includes the actual voltage distribution and the projected voltage distribution for the next operating cycle. It also includes the tube burst probability and calculated leakage under main steamline break differential pressure conditions. The projected voltage distribution with the resultant tube burst probability and leakage estimates account for flaw growth.
The staff routinely reviews these 90‑day reports and compares the tube burst probability and leakage to the criteria specified in GL 95‑05.1804 In addition, the staff compares the predicted values to actual values. If the predicted values are conservative, the flaw growth distribution used in the prediction is typically considered to be within expectations. If the predicted values are not conservative when compared to the actual values, the staff evaluates the root cause and ensures appropriate corrective actions are taken by the licensee.
In summary, the staff concluded1957 that any systematic deviations from expectations in flaw growth will be detected and addressed in the staff review of the 90‑day reports. The staff also concluded that crack growth rates will continue to be adequately monitored as part of the implementation of the voltage-based ARC and considers SGAP item number 3.8 to be closed.1958
Damage Progression Issues
The ACRS ad hoc subcommittee recommended: “Risk analyses that the staff considers need to account for progression of damage in a more rigorous way.”1898 This recommendation stemmed from a DPO concern that dynamic loads induced in steam generator tubes by an MSLB or other secondary-side breaches would lead to growth of cracks and increased steam generator tube leakage or ruptures outside the range of analyses and experiments performed by the NRC staff. In addition, an MSLB may impose dynamic loads on the TSPs beyond simply those associated with differential pressure loads, and these loads could be transferred to the tubes. The subcommittee noted that this concern affects any consideration of SG tube integrity and is not unique to use of voltage-based ARCs. The staff opened a new generic issue, GSI‑188, “Steam Generator Tube Leaks or Ruptures Concurrent with Containment Bypass from Main Steam Line or Feedwater Line Breaches,” in part to address this concern. This work was performed under item number 3.1 of the SGAP1899, 1940 and was completed. Key conclusions of the staff in resolution of GSI‑188 included:1870
Dynamic loads and resonance vibrations following an MSLB are low and have little impact on growth of existing cracks beyond the effects of differential pressure stress alone.
Dynamic loads from an MSLB or feedwater line break do not affect the structural integrity of tubes in service and do not lead to additional leakage or ruptures beyond what would be determined using differential pressure loads alone.
Therefore, the principal assertion of GSI‑188 is closed, and no changes to existing regulations and guidance are recommended.
The dynamic load effects from an MSLB or feedwater line break need not be taken into account in evaluating the potential for multiple tube ruptures under GSI‑163.
The ACRS reviewed the technical basis for these findings1862 and concluded that item number 3.1 of the SGAP is appropriately closed out. Confirmatory information requested by the ACRS1862 was subsequently provided to the ACRS.1870
Jet Impingement Issue
The ACRS ad hoc subcommittee considered a DPO concern that particulate-laden fluids flowing from a cracked SG tube can pierce adjacent tubes. The staff evaluated this concern as item number 3.2 of the SGAP.1899, 1940 This item addressed both MSLB and severe accident conditions. In its review of the DPO concerns,1898 the ACRS ad hoc subcommittee concluded that the staff had undertaken adequate research (under item number 3.2 of the SGAP) to address this issue. The subcommittee stated that, although it is necessary to carry this research to an appropriate conclusion, early results suggest that damage progression by the jet cutting mechanism is not likely.
Item number 3.2 has been completed,1959 and the detailed results of this study for MSLB conditions are documented.1960 The study was based on tests that provided a conservative simulation of an MSLB to determine the susceptibility of SG tubes to erosive damage from impacting jets of superheated steam leaking from adjacent tubes. This study showed that the likelihood of failure propagation by jet erosion is low under these conditions.
The detailed results for severe accident conditions are documented.1961 Erosion tests were conducted in a high-temperature, high-velocity erosion rig using micron-sized nickel and aluminum oxide particles mixed in a high-temperature gas. The erosion results, together with analytical models for crack opening area and jet velocities, were used to estimate the erosive effects of superheated steam with entrained aerosols from the core during severe accidents. It was determined that failure of an adjacent tube by jet impingement would take more than 10 hours after the subject crack had undergone significant crack opening displacement by creep at high temperature. However, once the system has reached these high temperatures, failure of some primary system component, including unflawed SG tubes, would be expected to occur in less than 1 hour. Thus, jet impingement is very unlikely to contribute in any significant way to severe accident risk.
The ACRS agreed with the staff’s conclusion that the probability of damage progression via jet cutting of adjacent SG tubes is low and need not be considered in accident analyses.1862 The ACRS also agreed that SGAP item number 3.2 should be closed.
Crack Unplugging Issue
The ACRS ad hoc subcommittee considered a DPO concern that forces involved with MSLB blowdown and leakage through cracks can cause cracks plugged with corrosion products to leak. In addition, the DPO was concerned that corrosion products in the annular gap between the tubes and TSP holes can be expelled, allowing otherwise occluded cracks to leak. The subcommittee stated that it found no evidence that the “unplugging” of cracks is a damage progression mechanism of concern.1898 The subcommittee made no recommendations concerning any followup study of this issue, and no such work has been included as part of the SGAP. The staff does not believe such work is necessary. Models used to predict leak rate under accident conditions tend to be mechanistic models (based in part on crack geometry) that have been benchmarked against test data (from pulled tube specimens and laboratory specimens) or empirical models such as that used for the voltage-based ARC. In both cases, the test data are expected to reasonably reflect the leakage that would be expected for cracks in the free span under actual accident conditions.
Risk Issues Pertaining to Tube Ruptures or Leakage during MSLB
A central concern of the DPO1936 was that MSLB can lead to primary-to-secondary leakage of tube rupture proportions sufficient to deplete the reactor water storage tank inventory via emergency core cooling system injection lost to the secondary side of the SGs (and therefore not available for recirculation from the containment sump), thereby leading to core damage with containment bypass. This concern relates to primary-to-secondary leakage from one or more tube ruptures or relatively large numbers of tubes that have not burst, such that the total leakage from all tubes is comparable to one or more tube ruptures.
The DPO estimate of core damage frequency and containment bypass frequency associated with SG tube leakage as a consequence of an MSLB was 1.0x10‑4 per reactor year (RY).1937 This estimate is based on assuming (1) an MSLB frequency of 1.0x10‑4/RY, (2) a conditional probability of 1.0 that primary-to-secondary leakage will be of tube rupture proportions under MSLB conditions, and (3) a conditional probability of 1.0 for failure to successfully mitigate the event before core damage occurs.
Staff PRAs considered by the ACRS ad hoc subcommittee assumed that the frequency of initiating secondary side depressurization events is dominated by stuck-open SG relief valves, with a frequency of 1x10‑3/RY estimated from operational event data. The frequencies of MSLB and main feed line break are estimated to be 6.8x10‑4/RY and 1.8x10‑4/RY, respectively, for a 4‑loop plant. The DPO did not appear to have any concerns relative to these estimates, nor did the ACRS ad hoc subcommittee state any concern relative to these estimates.
Conditional Probability of SG Tube Rupture during MSLB
The DPO concern relates to plants with widespread stress-corrosion cracking, particularly those plants with ARC TS that allow many tubes with such cracks to remain in service, and that, because of eddy current limitations in reliably detecting such cracks, leakage of tube rupture proportions is the expected outcome. As discussed earlier, the ACRS ad hoc subcommittee acknowledged that ECT techniques are not capable of 100‑percent accuracy in detecting flaws (though noting the technical advances that have led to improved detection performance). However, the subcommittee stated that this does not degrade the protection afforded to the public health and safety, provided the risk is properly managed.
Staff PRAs considered by the ACRS ad hoc subcommittee assumed the conditional probability of ruptures or leakage from multiple tubes of tube rupture proportions to be equal to or less than 0.05. The ACRS subcommittee did not make specific comments regarding the staff’s assumption, but concluded that, if the risk can be managed properly, it is acceptable to operate plants with known, small flaws as well as undetected flaws in the SG tubes. As an example of managing risk, the ACRS ad hoc subcommittee cited the voltage-based ARC methodology that requires that the conditional probability of rupture be demonstrated periodically to be 0.01 or less (for tubes degraded by ODSCC at the tube-to-TSP intersections). Looking beyond voltage-based ARCs, the performance-based strategy for ensuring tube integrity in the new TS (i.e., ensuring and periodically demonstrating that all tubes satisfy the structural and accident leakage integrity performance criteria consistent with the design and licensing bases) is a risk management strategy. Meeting the performance criteria on a consistent basis ensures that the conditional probability of tube leakage of tube rupture proportions under MSLB is low relative to values assumed in PRAs. This conclusion is supported by operating experience, as discussed earlier.
Accident Mitigation/Human Factors Issues
The ACRS ad hoc subcommittee concluded that “analyses of human performance errors during design basis accidents appear consistent with current practices.”1898 The subcommittee reviewed the DPO concern that the staff’s estimate of the probability that the operators will fail to perform tasks needed to establish the long-term cooling of the core (i.e., 10‑3 or 1 in 1,000) is overly optimistic. The subcommittee concluded that the staff estimate appears consistent with the state of current understanding of human performance errors when only a single tube ruptures. The subcommittee stated that, in developing assessments of risk concerning these DBAs, the staff must consider the probabilities of multiple tube ruptures until adequate technical arguments have been developed to show that damage progression is improbable.1898
The DPO’s and ACRS ad hoc subcommittee’s concerns pertaining to damage progression were evaluated under item numbers 3.1 and 3.2 of the SGAP. As discussed above, the ACRS has concurred with the staff’s conclusions drawn from the results of these studies and with the staff’s closure of these item numbers. The staff concludes that the damage progression mechanisms cited in the DPO are unlikely to increase the probability of multiple tube ruptures beyond that which has already been considered in staff PRAs.
The ACRS ad hoc subcommittee also observed that the staff needs to develop defensible analyses of the uncertainties in its risk assessments, including uncertainties in its assessments of human error probabilities. The subcommittee noted that, as the staff develops a better understanding of the dynamic processes associated with depressurization during an MSLB, the staff may want to revisit estimates of operator error probability in light of the considerable distraction that might occur during such events. In response to the comments, the staff is developing improved methods for risk assessment under item number 3.5 of the SGAP.1899, 1940 This item number is considered outside the scope of GSI‑163 because it is focused on severe accidents and its completion is not expected (based on early results) to identify needed improvements to the current regulatory framework for ensuring SG tube integrity. With respect to operator distraction that may occur during such an event, the staff notes that the dynamic effects of the event will happen quickly. No mandatory operator actions are needed while the plant is experiencing these short-lived dynamic effects.
Severe Accident Risk Issue
The ACRS ad hoc subcommittee considered a DPO concern that severe accident sequences in which the primary system remains pressurized are more likely to evolve into steam generator tube rupture accidents than the staff predicts in NUREG‑1570, “Risk Assessment of Severe Accident-Induced Steam Generator Tube Rupture,” issued March 1998.1954
The ACRS ad hoc subcommittee concluded that “substantial uncertainties remain in the understanding of steam generator tube performance under severe accident conditions.”1898 The subcommittee stated the following:
The staff has not developed persuasive arguments to show that the steam generator tubes will remain intact under conditions of risk-important accidents in which the reactor coolant system remains pressurized. The current analyses dealing with loop seals in the coolant system are not yet adequate for risk assessments. The treatment of mixing of flows in the inlet plenum to a steam generator under conditions of countercurrent natural convection flow are optimistic and are not substantiated by applicable data from experiments. Sensitivity studies have not explored the plausible ranges of parameter values or the space of uncertainties adequately. Finally, the Ad Hoc Subcommittee notes that analyses of failure of other locations in the coolant system subject to natural convection heating have not included a systematic examination of vulnerable locations in the system.
The ACRS ad hoc subcommittee’s concerns relating to severe accidents were addressed under item number 3.4 of the SGAP.1899, 1940 This item is outside the scope of GSI‑ 163 because, should any action be determined necessary to address severe accident risk concerns, these actions would likely be directed toward accident mitigation rather than modification of the current regulatory framework for ensuring SG tube integrity.
Iodine Spiking and Source Term Issues
As part of the voltage-based tube repair criteria,1804 licensees must demonstrate that primary-to-secondary leakage that may potentially occur under MSLB conditions does not exceed values assumed in the licensing basis safety analyses to demonstrate that the associated dose consequences meet applicable regulations (i.e., 10 CFR 50.67 or 10 CFR 100, GDC 19). In accordance with the NRC’s Standard Review Plan (NUREG‑0800, “Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants, LWR Edition”), these dose calculations are based on an initial coolant equilibrium iodine concentration equal to the allowable limit in the technical specifications (typically 1.0 microcurie per gram) and an iodine spiking factor of 500. As part of their license amendment requests for voltage-based tube repair criteria, a number of licensees requested (and the staff approved) reduced limits in the TS on allowable equilibrium iodine concentrations in the primary coolant. This reduction in the allowable equilibrium iodine concentration means that a higher level of primary-to-secondary leakage can be tolerated, assuming the same iodine spiking factor of 500, consistent with the applicable regulatory dose limits, thus enabling additional degraded tubes to remain in service (provided all other requirements of the ARC are met). The ACRS ad hoc subcommittee reviewed a DPO concern that data (primarily from reactor trips, but including SGTR events) indicate that spiking factor increases with decreasing steady-state iodine concentration. Thus, there was a concern that the spiking factor used for the licensing basis accident analysis is too low when the TS limit on the iodine concentration in the primary coolant has been reduced.
The ACRS ad hoc subcommittee recommended the following: “The staff should develop a more technically defensible position on the treatment of radionuclide release to be used in safety analyses of design basis events.”1898 This recommendation was addressed under item number 3.9 of the SGAP and was discussed at the ACRS meeting on February 5–7, 2004. In a letter dated May 21, 2004, the ACRS stated: “The staff continues to treat iodine spiking in a conservative, empirical fashion. We recommend that the staff develop a mechanistic understanding of iodine spiking so that analyses reflect current plant operations and the capabilities of modern fuel rods.”1862 The ACRS continued with the following:
The staff has not accepted our recommendation to develop a mechanistic understanding of the iodine spiking issue. The staff continues to use a conservative, empirical estimate of iodine spiking for accident consequence analyses. This estimate is based on historical data that may not reflect current practices in plant operations or the capabilities of modern fuels to prevent coolant contamination. We again encourage the staff to take advantage of iodine studies available in the literature and develop a mechanistic understanding of the phenomenon.
On the basis of these ACRS comments, the staff proposed a new generic issue, GSI‑197, “Iodine Spiking Phenomena.” This issue was screened1867 by a review panel in accordance with NRC Management Directive 6.4.1858 The review panel found the issue to be of low safety significance and concluded that it should not be continued as a safety issue. The review panel found that there is no evidence that the current regulatory approach is not bounding, even in event of a combined MSLB and SGTR, and that the current regulatory approach to iodine spiking, in spite of its empirical nature, is adequate. Generic Issue 197 and SGAP item number 3.9 are closed.1867 The ACRS stated that it had considered the results of the staff’s screening of GI‑197 and had no objection to dropping this issue from further consideration.1962
To address the DPO concern, the staff evaluated the adequacy and effectiveness of industry practice and regulatory requirements relating to the management of SG tube integrity to ensure that all tubes will exhibit acceptable structural margins against burst or rupture under normal operating conditions and DBAs (including MSLB), and that leakage from one or multiple tubes under DBAs will be limited to very small amounts, consistent with the applicable regulations for offsite and control-room dose. As part of this effort, the staff considered the conclusions and recommendations of the ACRS ad hoc subcommittee, which served as the DPO review panel. The staff’s followup actions taken in response to these findings served as part of its evaluation of the adequacy and effectiveness of regulatory requirements.
As of September 30, 2007, new performance-based TS requirements1943 were in place and being implemented at all U.S. PWRs. These requirements are intended to ensure that all tubes exhibit adequate structural margins against burst or rupture for the spectrum of normal operating and DBA conditions, consistent with the original design basis. In addition, these requirements are intended to ensure that total leakage from tubes at a plant will not exceed values assumed in licensing bases accident analyses, even if no tubes actually rupture under these conditions. In addition, licensees are required to periodically demonstrate that these structural margin and accident leakage criteria are satisfied for all tubes or, if not satisfied, to report the occurrence in accordance with 10 CFR 50.72 and 50.73.
U.S. PWR licensees have used the basic elements of the required performance-based approach since 2000 as part of the industry’s initiative under NEI 97‑06.1939 NEI 97‑06 itself was an evolutionary development because tube inspection technologies, inspection practices, and tube integrity management practices had been undergoing significant improvement since the mid-1970s. These improvements contributed significantly to improved SG tube integrity performance during this period. Improved water chemistry practices and the increasing number of PWRs with SGs of improved design and more stress-corrosion crack-resistant tubing have also contributed to this trend. Since adoption of the NEI 97‑06 performance-based strategy in licensee SG programs and the corresponding availability of more complete information about instances of failure to satisfy SG tube integrity performance criteria, actual incidences of failure to meet these criteria have been infrequent. This experience provides strong evidence that the potential for one or more tube ruptures, or leakage from multiple tubes totaling tube rupture proportions, under normal operating conditions or DBAs is well within that assumed in NRC risk studies to date.
The staff completed all SGAP1899, 1940 tasks that were opened to address the ACRS ad hoc subcommittee’s conclusions and recommendations stemming from its review of the DPO concerns relating to voltage-based ARCs, damage progression mechanisms, and iodine spiking. On the basis of the results of these tasks, the staff concluded that the DPO concerns relating to these issues were not substantiated and that no changes to existing requirements were needed to ensure public health and safety. The ACRS concurred with the closure of these issues. In response to ACRS ad hoc subcommittee conclusions and recommendations, the staff continued to evaluate risk issues associated with accident sequences involving ruptured or leaking SG tubes as part of SGAP1899, 1940 item numbers 3.4 and 3.5. These studies are primarily focused on severe accidents and are not expected to identify needed changes to existing requirements for managing SG tube integrity; therefore, they are outside the scope of GSI‑163.
On the basis of the above, the staff concluded that current TS requirements1943 relating to SG tube integrity provide reasonable assurance that all tubes will exhibit acceptable structural margins against burst or rupture under normal operating conditions and DBAs, including MSLB, and that leakage from one or multiple tubes under DBAs will be limited to very small amounts, consistent with the applicable regulations for offsite and control-room dose. Thus, the staff concludes that the GSI‑163 principal assertion and related concerns in the DPO are not substantiated, that no changes to existing regulations or guidance are needed, and that actions for the GSI are completed.
In accordance with Management Directive 6.4,1858 the GSI closeout process includes an endorsement by the ACRS. The staff met with the ACRS on May 7, 2009, to discuss the staff’s technical basis for resolution of GSI‑163. In a letter dated May 20, 2009, to Gregory B. Jaczko, Chairman, NRC, the ACRS concluded that GSI‑163 can be closed as proposed by the staff.1963 On July 16, 2009, the staff issued a memorandum to the EDO to indicate the completion of actions for GSI‑163.1948
681. NUREG‑0844, “NRC Integrated Program for the Resolution of Unresolved Safety Issues A‑3, A‑4, and A‑5 Regarding Steam Generator Tube Integrity,” U.S. Nuclear Regulatory Commission, September 1988.
1031. Memorandum for T. Murley from E. Beckjord, “A New Generic Issue: Multiple Steam Generator Tube Leakage,” June 16, 1992.
1091. Memorandum for D. Morrison from H. Thompson, “Generic Issue Management Control System,” January 17, 1997.
1804. Generic Letter 95‑05, “Voltage-Based Repair Criteria for Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress-Corrosion Cracking,” U.S. Nuclear Regulatory Commission, August 3, 1995. 
1858. Management Directive 6.4, “Generic Issues Program,” U.S. Nuclear Regulatory Commission, November 17, 2009.
1862. Letter to W. Travers (U.S. Nuclear Regulatory Commission) from M. Bonaca (Advisory Committee on Reactor Safeguards), “Resolution of Certain Items Identified by the ACRS in NUREG‑1740, ‘Voltage Based Alternative Repair Criteria,’” May 21, 2004. [ML041420237]
1870. Memorandum for L. Reyes from C. Paperiello, “Completion of Generic Safety Issue 188, ‘Steam Generator Tube Leaks/Ruptures Concurrent with Containment Bypass from Main Steam or Feedwater Line Breaches,’” December 16, 2005. [ML052150154]
1899. Memorandum for W Travers from S. Collins and A. Thadani, “Steam Generator Action Plan Revision to Address Differing Professional Opinion on Steam Generator Tube Integrity (WITS Item 200100026),” May 11, 2001. [ML011300073]
1900. SECY‑03‑0080, “Steam Generator Tube Integrity (SGTI)—Plans for Revising the Associated Regulatory Framework,” U.S. Nuclear Regulatory Commission, May 16, 2003.
1936. Memorandum for J. Hopenfeld from T. Speis, “Your Differing Professional Opinion Dated 12/23/91,” February 19, 1992. This memorandum encloses (Enclosure 1) J. Hopenfeld’s Differing Professional Opinion, dated December 23, 1991. 
1938. ASME Boiler and Pressure Vessel Code, Section III, “Rules for Construction of Nuclear Power Plant Components,” Division I, American Society of Mechanical Engineers, New York, NY
1943. Federal Register Notice 70 FR 24126, “A Notice of Availability of Model Application Concerning Technical Specification; Improvement To Modify Requirements Regarding Steam Generator Tube Integrity; Using the Consolidated Line Item Improvement Process,” May 6, 2005.
1944. Draft Regulatory Guide DG‑1074, “Steam Generator Tube Integrity,” U.S. Nuclear Regulatory Commission, December 1998. Issued for public comment in Federal Register Notice 64 FR 3138; January 20, 1999. [ML003739223]
1948. Memorandum for R. Borchardt from E. Leeds, “Completion of Actions for Generic Safety Issue 163, ‘Multiple Steam Generator Tube Leakage’ (Steam Generator Action Plan Item 3.11),” July 16, 2009. [ML091540192]
1949. Letter to A. Blind (Consolidated Edison Company) from W. Lanning (U.S. Nuclear Regulatory Commission), “NRC Special Inspection Report—Indian Point Unit 2 Steam Generator Failure—Report No. 05000247/2000‑010,” August 31, 2000. [ML003746339]
1950. Letter to C. Terry (TXU Energy) from D. Chamberlain (U.S. Nuclear Regulatory Commission), “Comanche Peak Steam Electric Station—Special Team Inspection Report 50‑445/02-09,” January 9, 2003. [ML030090566]
1952. Licensee Event Report 50‑302/2004‑004‑00 for Crystal River Unit 3, “NUREG‑1022 Clarification Required Reporting of Previous Steam Generator Tube Inspection Results,” November 22, 2004. [ML043340228]
1953. Letter to D.E. Young (Florida Power Corporation) from B.L. Mozafari (U.S. Nuclear Regulatory Commission), “Crystal River Unit 3—Issuance of Amendment Regarding Probabilistic Methodology for Tube End Crack Alternate Repair Criteria (TAC No. MC5813),” October 31, 2005. [ML052940179]
1954. NUREG‑1570, “Risk Assessment of Severe Accident-Induced Steam Generator Tube Rupture,” U.S. Nuclear Regulatory Commission, March 1998.
1955. Letter to G. Rueger (Pacific Gas and Electric Co.) from G. Shukla (U.S. Nuclear Regulatory Commission), “Diablo Canyon Power Plant, Unit Nos. 1 and 2—Issuance of Amendment Re: Permanently Revised Steam Generator Voltage-Based Repair Criteria Probability of Detection Method (TAC Nos. MC2313 and MC2314),” October 28, 2004. [ML043140452]
1958. Letter to M.V. Bonaca (Advisory Committee on Reactor Safeguards) from L.A. Reyes (U.S. Nuclear Regulatory Commission), “Resolution of Certain Items Identified by the Advisory Committee on Reactor Safeguards in NUREG‑1740, ‘Voltage-Based Alternative Repair Criteria,’” August 25, 2004. [ML042400055]
1963. Letter to G.B. Jaczko (Chairman, U.S. Nuclear Regulatory Commission) from M.V. Bonaca (Advisory Committee on Reactor Safeguards), “Proposed Resolution of Generic Safety Issue‑163, ‘Multiple Steam Generator Tube Leakage,’” May 20, 2009. [ML091320055]