ML20217B363

Topical Rept Evaluation of BWR Vessel & Internals Project, Guidelines for Reinspection of BWR Core Shrouds (BWRVIP-07)

ML20217B363
Person / Time
Issue date:	09/15/1997
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20217B360	List: ML20217B355 ML20217B363
References
FACA, NUDOCS 9709240202
Download: ML20217B363 (20)

Text

.

{ RS. NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION DIVISION OF ENGINEERING SAFETY EVALUATION OF EPRI TOPICAL REPORT TR 105747 "BWR VESSEL AND INTERNALS PROJECT. GUIDELINES FOR REINSPECTION OF BWR CORE SHROUDS (BWRVIP-07)"

1.0 INTRODUCTION

\

1.1 Background

By letter dated February 2,1996, the Bolling Water Reactor Vessel and Intemals Project (BWRVIP) submitted the Electric Power Research Institute (EPRI) Topical Report TR 105747, "BWR Vessel and Internals Project, Guidelines for Reinspection of BWR Core Shrouds" (BWRVIP 07, Ref.1), for staff review and approval. The BWRVIP clarified the content of BWRVIP 07 w:th letters dated July 17,1996 (Ref. 2), and May 21,1996 (Ref. 3), and supplemented the content of the guidelines with additionalinformation on October 21,1996 (Ref. 4). The BWR',dP further modified the guidelines in BWRVIP 07 by letter dated January 8,1997 (Ref. 5).

The objective of the BWRVIP in submitting BWRVIP 07 was to provide a uniform Industry approach to reinspection that would ensure the structural and func"onalintegrity of the BWR core shroud whether repaired or unrepaired. This inspection criteria for unrepaired shrouds specify the time interval to the first required reinspection following the baseline inspection, as well as the welds to be inspected, the extent of the inspection, and intervals for subsequent reinspections. For repaired shrouds, the inspection criteria addresses reinspections of both the core shroud and the repair components. The guidelines are meant to be a "living" document and are considered subject to revision as the industry gathers more data after completing baseline inspections.

The General Electric Company (GE) addressed the flaw acceptance criteria for circumferential welds in unrepaired shrouds in "BWR Core Shroud Inspection and Flaw Evaluation Guidelines," GENE 523113-0894, September 2,1994 (Ref. 6) and "BWR Core Ghroud Inspection and Flaw Evaluation Guidelines," GENE 523-113-0894, Rev,1, March 1995 (Ref. 7). The staff's safety evaluations (SEs) for these documents were issued on December 28,1994, and June 16,1995, respectively (Refs. 8 and 9).

The BWRVIP-07 guidelines allow for plant specific analyses to be performed for a given weld or for the entire shroud. These plant specific analyses are not addressed in the scope of this report, and NRC approval must be obtained on a case-by case basis. The staff review of flaw analyses performed according to the more conservative "BWR Core Shroud Inspection and Flaw Evaluation Guidelines,"(GENE 523113-0894, Revs. O and 1) will continue to be evaluated in accordance with the previous SEs. Flaw evaluations conducted according to the 1

l -.

j C

4 d

., more comprehensive but less conservative guidelines of BWRVIP-07 will be evaluated by the staff and will consider the recommendations contained within this SE.

1.2 Purpose The staff reviewed BWRVIP-07 to determine whether its guidance would ensure the structural Integrity of the BWR core shroud.

1.3 Organization of this Report Because BWRVIP-07 is proprietary, this SE was written so as not to repeat information contained in the report. The SE gives a brief summary of the general contents of the report (Section 2.0) and the detailed evaluation in Section 3.0. The SE does not discuss in any deteil the provisions of the guidelines because they are proprietary, nor does it discuss the parts of the guidelines it finds acceptable. Recommendations of issues for the BWRVIP to resolve and conclusions are summarized in Section 4.0. The staff has structured its evaluation according to the o 1anization of the BWRVIP 07 report.

2.0

SUMMARY

OF TOPICAL REPORT The BWRVIP-07 report addresses the following topics in the following order

• Reinspection of BWRs with Unrepaired Core Shrouds

- Category B plants

- Category C plants

• Reinspection of BWRs with Repaired Core Shrouds

- General requirements

- Repair component inspections

- Repair Anchorage inspections

- Horizontal and vertical core shroud weld inspections The BWRVIP 07 report contains appendices on:

• Plant Categories for Core Shroud Inspections Generic Fracture Mechanics Analyses A Statistical Method for Estimating the Cracking in inaccessible Regions of Core Shroud Welds In the staff's evaluation, several terms in BWRVIP 07 are referenced in discussion. For clarification, these terms are defined or described below:

4 9 4 .

n. ,--.. , , - _ . , + ,

I ,

Category B plant: plants operated more than eight hot operating years with core shrouds made with Type 304L stainless steel (Iow carben content) and average conductivity less than 0.3 pS/cm during the first five cycles of operation.

Category C plant: plants operated more than six hot operating years with core shroud made with Type 304 stainless steel (high carbon content) or plants operated more than eigh' hot operating years with core shrouds made of Type 304L stainless steel (low carbon content) and average conductivity greater than 0.3 pS/cm during the l'rst five cycles of operation.

L,,,,  : the minimum required length of uncracked weld, including two cycles of crack growth, to ensure the structural integrity of the weld.

L',,,,  : the minimum required length of uncracked weld, including an allowance of crack growth up to a maximum of 6 years, to ensure the structural integrity of the weld. L',,,,is the sum of the minimum required length of sound weld (assuming inaccessible regions are 100 percent cracked) and the crack growth over the desired inspection interval (crack growth rate of 5 x 104 in/hr).

Option A : inspection strategy for category B plants that requires 100% inspection of all accessible areas of the shroud welds H3, H4, H5, and H7. If the cracking is found to be greater than 10 percent in any weld, then full inspections are required in accordance with the requirements for Category C plants.

Option B : inspection strategy for category B plants that allows inspection of a length of L',,,, for the welds specified in Option A.'"

t  : the wall thickness Table 1  : provides core shroud reinspection intervals for Category C plants. The reinspection intervalis a function of the degree of cracking and the stress level in the weld.

Table 2  : provides a summary of the crack length assumption used in determining the reinspection intervals in Table 1.

3.0 STAFF EVALUATION in general, the staff finds that the guidance presented in BWRVIP-07 ensures the structural integrity of the core shroud. The scope of the guidelines is comprehensive, covering circumferential, vertical, and ring segment welds, as well es repair components for both Category 8 and C plants. Industry experience shows that the inspection methodologies -

'" It should be noted that Options A and B vary slightly depending on which weids are being discussed. To more fully 00 scribe these Options would require publicly providing information that the NRC staff has determined to be proprietary in nature.

l

enhanced visual and ultrasonic methods - are appropriate for detecting cracking. In addition, the guidance is consistent with that in earlier staff approved reports on the inspection and evaluation of core shrouds. The guidelines of BWRVIP-07 incorporate many of the recommendations in the referenced safety evaluations and also allow for revision on the basis of new data and lessons learned. The staff, therefore, finds much of the proposed guidance acceptable. However, the staff also finds the guidance complicated and difficult to implement and audit.

Sections 3.1 and 3.2 present the staff's detailed evaluation of the guidance regarding reinspections of BWRs with unrepaired and repaired core shrouds, respectively.

3.1 Reinspection of BWRs With Unrepaired Core Shrouds Sections 3.1.1 and 3.1.2 discuss the staff's findinp regarding the BWRVIP-07 guidelines for the reinspection of Category B and C plants with unrepaired core shrouds.

3.1.1 Category B Plants For Category B plants with unrepaired core shrouds, the staff finds that the guidance is generally acceptable; however, severalissuer that the staff requests the BWRVIP to resolve are presented below.

Issue 3.1.1 1: Use of L. to Determine Reinspection Intervals for Category B Shrouds The guidelines state that if tne limited inspections show that L. is met, more than one cycle of operation may be justified. In earlier safety evaluations of core shroud cracking (Refs. 8 and 9), the staff required reinspection after just one cycle. The staff recommends that L be used to justify only one additional cycle of operat'on.

Issue 3.1.12: Use of Option A to Determine Inspection Intervals for Category B Core Shrouds The staff considers that the recommended reinspection intervals of Option A* for unrepaired Category B core shrouds are not conservative.

The staff recommends the following modifications to BWRVIP-07:

(1) The inspection intervals (stated in years) should correlate reasonably well with cycle lengths.

(2) The 10-year inspection interval should be reduced to 6 years.

4

• e.g., an inspection interval of 120 months if no cracking is found in the weld; 96 months if the weld cracking is less than or equal to 10 percent of the weld length inspected; and, if cracking is greater than 10 percent of the weld length inspected, then full inspections are required in accordance with the requirements for Category C plants 4

(3) The 8 year inspection interval should be reduced to 5 years.

These reductions reflect the rationale with regard to the staff's concems about using Table 1 in BWRVIP 07 to determine inspection intervals. These concerns are addressed under issue 3.1.2 3, below Issue 3.1.13: Use of Option B to Establish the Reinspection Intervals for Category B Core Shrouds The staff's concerns about using the equation for L',,,, given under the proposed Option B are as follows:

(1) In responding to items 0 and 7 in the staff's request for information (ref 3), the BWRVIP stated that the assumption that the uninspected areas are fully cracked is conservative. Nonethelest the responses still do not adequately address the frequency of crack initiations in the "unflawed region." The staff can accept the L',,,,

approach for a relatively short p3riod of operation, (i e., one cycle); however, the staff does not consider this approach adSquate for the long term. Additional evaluation is needed from more rigorous inspections to juttify longer intervals between inspections.

(2) The referenced documents do not justify the assumption that the inspection uncertainty is not to exceed 2t (t= wall thickness). Previous analyses have used other values, including 4t. All values used for inspection uncertainties should be demonstrated and technically justified.

(3) Unlike the equation for L,,,,, this equation does not include a factor of 2 to account for additional safety margin.

Because of these concems, the staff does not consider it acceptable to use L,,,, and L',,,, to justify more than one additional operating cycle. The staff therefore recommends that the BWRVIP revise the inspection strategies citing L,,,n and LL criteria so that they no longer justify multi-cycle operation without reinspection.

Issue 3.1.14 and 3.1.2-4: Inspection of Vertical Welds The proposed reinspection of unrepaired core shrouds for both Category B and Category C plants is limited to horizontal (i.e. circumferential) welds only. This is consistent with the scope of the baseline inspection which is also limited to the inspection of circumferential welds. The reinspection guidelines do not require the inspection of the vertical welds even if the circumferential welds are found to be cracked. This appears to be based on the assumption that the core shroud will still function if the structural integrity of the flawed circurnferential weld can be shown by fracture mechanics analysis and crack growth evaluation. However, the staff has concems about the functionability of the shroud when a vertical weld has completely failed and an intersecting circumferential weld has extensively ciacked. In such a case, the core shroud could open locally during postulated design based accidents. This local opening of the shroud would create a large by pass flow from inside the shroud which could cause inadequate core cooling. Therefore, the staff requests that BWRVIP 07 provide for inspecting the vertical welds to ensure their integrity when cracking of I

. t

• \

e i the circumferential welds, us discussed above, is found. The provision should also be - 3~

incorporated into the BWRVIP document that addresses the baseline inspection of core shrouds. >

3.1.2 Category C Plants l The staff finds the reinspection guidance for Category C plants acceptable with the following t exceptions:

Issue 3.1.21: Use of L,,, to Determine the Reinspection Intervals for Category C Core Shrouds The concem discussed under 3.1.1 1 of Section 3.1.1 of this SE also applies. >

The guidelines stated that if a Category C plant cannot be inspected for L,,,, the acceptability and inspection interval of the weld are to be determined on a cass-by case basis. The staff recommends that such determinations be submitted to the NRC for review and approval.

' t issue 3.1.2 2: Use of the L',,, Equation to Determine the Reinspection Intervals for Category ,

C Core Shrouds This issue is covered by the discussion under issue 3.1.1-3 of Section 3.1.1 of this SE.

In addition, the guidance given by the L',,, method and Table 1 (discussed below) may be inconsistent in terms of conservatism. For instance, for a shroud with 40 percent cracking, ,

the table could require reinspection in about an operating cycle, whereas the L',,,, method may not require reinspection before 6 years, i.e., potentially four or more operstbg cycles.

The L',,, method requires much less inspection than that of the Table 1 method, but may allow a longer reinspection interval. The L',,, method is non-conservative relative to the Table 1 method.

Issue 3.1.2 3: Use of Table 1 to Determine Reinspection Intervals for Category C Core Shrouds The staff has the following concems about using Table 1 of BWRVIP 07 to determine the inspection intervals for Category C plants with unrepaired shrouds.

(1) The proposed 10-year reinspection interval for uncracked circumferential welds is too long because the welds would have been inspected only once since operation began.

Moreover, the original inspections may not have been of a sufficiently high quality.

Visual examinations could have missed some cracks if the surface was not properly prepared, or if the examination was not performed under optimal conditions of lighting and focallength. Similarly, original ultrasonic inspections (using what was then state-of-the-art equipment) may have missed some cract.: because the equipment and techniques were not as good as they are now.

(2)- The statistical model presented in BWRVIP-07 appears to be in error. The number of flaws assumed in Table 2," Crack Length Assumptions Used To Determine 6-

-- .- . _ . : - :. l. ---- . - . - . - . . - _ . - - _ - - - -

0

-l Reinspection intervals," does not bound the industry wide inspection results. The following tables present examples of this discrepancy:

Flaw Indications Reported in the Circumferential Welds of the Peach Bottom Unit 3 Core Shroud (Category C Plant, UT, Fall 1993)

Weld Percent Percent No of Flaws No. of Maximum Depth as a Inspected Cracked in Observed Flaws Percent of Wall Inspected Assumed in Thickness Area BWRVIP-07 HI 84.5 4.8 7 7 30.5 H3 89.5 34.9 10 24 42.5 H4 89.2 32.2 36 24 50.0 HS 90.8 7.5 9 7 11.2 Flaw Indications Reported in the Circumferential Welds of the Susquehanna Unit 1 Core Shroud (Category B Plant, UT, Spring 1995)

Weld Percent Percent No. of No. of Maximum Depth as a inspected Cracked :n Flaws Flawr Percent of Wall Inspected Observed Assumed in Thickness Area BWRVIP-07 Hi 83.2 0.6 3 7 -

H2 83.2 7,9 29 7 -

H4 82.7 20.9 28 20 35 HS 72.7 29.4 27 20 32 H6A 83.9 1.8 8 7 12 H6B 84.2 10.2 19 14 33 The following conclusions can be deduced from a review of the above and the BWRVIP-07 report:

(a) The number of flaws reported in 7 of the 10 welds listed above exceeds the number assumed in BWRVIP-07 Therefore, it appears that the defect density assumed in the statistical model of BWRVIP-07 does not bound the actual inspection results to date with 95 percent confidence as suggested.- According to the proposed methodology,

• the reinspection intervals would have been reduced if a larger number of flaws had been assumed for a weld. To resolve this discrepancy, a sensitivity study is needed 7

5 to show the effect of defect density on the proposed reinspection intervals presented in Table 1 of BWRVlP-07.

(b) Table 2 of BWRVIP-07 appears to contain an inconsistency between the amount of cracking assumed and amount of cracking actually reported. The staff requests that an explanation be provided of how, for example,20 percent cracking observed in 55.3 inches of inspected weld can contain 14 flaws, whereas a statistical model for about the same length of an inaccessible weld assumes only 7 flaws with a 95 percent confidence level.

(c) In the staffs SE dated December 28,1994 (Ref. 8), the staff recommended that licensees inspect the core shrouds in Category C plants during each refueling outage because these plants are highly susceptible to IGSCC. However, the staff allowed that licensees could reduce the reinspection frequency when the inspection findings support such a reduction. Table 1 of BWRVIP-07 proposes such a reduction; however, the staff finds that additional reinspection data are needed to permit confident prediction of the crack initiation and growth. Actualinspection data from at least two or three reinspections of the core shrouds ranked as Category C* are needed to permit evaluation of the reinspection intervals in Table i to determine the conservatism of the proposed flaw density.

(d) The input to the fracture mechanics analysis still requires careful review to determine its acceptability for generic application. Specifically, the staff has to review the proposed flaw density in inspected and uninspected regions, the assumed defect rate in inaccessible regions, and the use of the Distributed Ligament Length (DLL) computer code for calculating (m.

The staff also has to review the assumption ori which the BWRVIP-07 report's Table 3 is predicated. Data in Table 3, " Defect Rates in inaccessible Regions of Core Shroud Welds," are derived from a probabilistic estimate of the amount of cracking that could occur in the inaccessible region. The analysis assumes that the mean defect rate in the inaccessible region equals that observed in the inspected region and calculates a defect rate that (based on a 95-percent probability) would not be exceeded. This assumption that every region of the weld is equally susceptible to cracking may not be entirely valid but may be acceptable given other margins in the analysis. Moreover, in its response to the staffs RAI , the BWRVIP stated that the number of flaws assumed in the fracture mechanics analyses was based on the UT results of a BWR 4 core shroud inspection. The staff will have to determine the validity of this statement in light of new data from recent inspections.

(e) BWRVIP-07 contains the findings of a new crack initiation study that supported a reinspection interval from 8.4 years to 8.9 years for cracking in no more than 40 percent of the weld circumference. However, this study assumed a crack growth rate

• Includes those core shrouds originally ranked in Category B grouping thut have been reclassified to Category C status as a result of inspections that revealed cracking in excess of 10 percent.

8

l h

J.

of 2 x 10 inch per hour. The staff has this crack growth rate under review. Until the staff can draw a conclusica regarding the acceptability of the proposed growth rate, the bounding crack growth rate currently accepted for fracture mechanics analyses is 5 x 10~5 inch per hour.

e in light of the concems about using Table i to predict the inspectin intervals for Category C

, core shrouds and the time required to resolve them, the staff has oeve:oped an attemative (shown as Table A to this SE) for interim use in determining reinspection intervals for 1 circumferential welds.

\ 3.1.3 Aceptable Alicmative to Table 1 of BWRVIP.07 for Interim UsG in Determining f' Reinspection Intervals (l'able A)

The etafTs alternative (Table A) is acceptable if the circumferential welds have been apected ano ocemined to have a remaining uncracked ligament in excess of the minimum E uncra.;ked ligamon, in addition welds , must have been inspected for at least 50 percent of f the weld circumference The inspection fraquencies given below may be used for the longer term The staff has s

simplified and modified the criteria in Table 1 of BWRVIP-07 to make ther, sufficiently conservative, and have capped the reinspection interval at 6 years.

(1) Fc.r welds with no relevant crack indications found in the inspected length, the

[ reinspection interval should not exceed 6 years or 4 fuel cycles, whichever comes first.

(2) For welds with cracking in no more than 10 percent of the inspected length, the reinspection intervcl shou ld not exceed 5 years or 3 fuel cycles, whichever comes first.

(3) For welds whh cracking in more than 10 percent and no more than 25 percent of the inspected longth, the reinspection interval should not exceed 4 years or 2 fuel cycles, whichever comes first.

(4) For welds with cracking in more than 25 perceni tut not more than 40 percent of the inspected length, the reinspection should be performed during the next refueling n outage.

(5) For welds with cracking in excess of 40 percent of the inspected length, a plant-specific analysis is required to support continued operation for on't fuel cljcle without repair.

3.2 Reinspection of Boiling-Water Reactors with Repaired Core Shrouds With regard to the proposed guMelines for repaired core shroads and the components used in thee rt pair modifications, the staff finds the general guiciance acceptable.

.g.

l

.m_ ,m _________________

l Sections 3.2.1 through 3.2.4 discusses the staffs concems regarding the BWRVIP-07 guidelines for reinspection of components used h core shroud repair modifications, repair anchorages, horizontal and vertical welds in rep rud core shrouds, and ring segment welds in repaired core shrouds, respectively.

3.2.1 Reinspection of Repsir Components The following items summarize tre staffs concems about the proposed guidance on reinspections of components used in repairs of shrouds :

Issue 3.2.11: Recommended Generic Criteria Regarding Reinspection of Core Shroud Repair Components Recent industry experience suggests that the generic recommendations presented in BWRVIP 07 may not be adequate for reinspection of components used in core shroud repa!r modifications. For example, a licensee found that a repair, which was in place for only one cycle, had a failed retainer clip on one assembly and a loose nut on another. After finding the degradation, the licensee examined all the assemblies, in contrast, the guidance in BWRVIP-07 does not provide for expanding the inspection of the sample if degradation of the repair components is discovered. The staff recommends including such a provision in the guideline; of BWRVIP-07.

1. that licensee had followed the guidelines and had selected a different assembly for

,arapling, it could have missed this degradation. The staff requests that licensees provioe a tecnnicaljustification for not performing a 100-percent inspection after one cycle. The staff recommends requiring such an inspection to discover installation errors and other problems.

It is not clear whether following the provisions of the guidelines would have revealed the loose nut on the degraded repair described above. The staff recommends that the guidelines specify ways of verifying bolt tightness when visual examination is not feasible or adequate.

Issue 3.2.1-2: Recommendation to Review Current Set of Assumptions Used in Analyses of Core Shroud Repair Modifications The staff recommends that the guidelines specify reviewing the design assumptions for the repair. For example, repairs may be installed preemptively. Such repairs may be designed assuming that welds are fully cracked or by assuming that a minimum ligament is present.

The staff recommends a pre-repair inspection for all futt.re repairs to ensure the minimum ligaments assumed Lctually are present.

In light of the above concems, the staff is recommending the attemative below and summarized in Table B of this SE. The alternative specifies more inspections early in the life of the repair to ensure that the repair is performing as designed.

Recommended Altee tive for Reinspections of Components in Core Shroud Repair Designs (1) After installing a repair, licentees should reinspect all repair components during the subsequent refueling outage. If any repair component is found to be degraded, the

. t licensee should perform reinspections of the.same scope during the successive refueling _ outages until no degradation is found. (Degradation includes missing or

degraded parts, niisalignment of repair parts, loosening or relaxation of locking '

devices and deviation of critical gaps or contact areas.)

(2)- After successfully completing the inspections in item (1) above, the licensee may perform a 25 percent reinspection of the repair components during each of the next four refueling outages, if these inspections reveal degradation, the licensee should expand the inspection to 100 percent of the repair components, and inspect 5 accordance with the reinspection scope prescribed in item (1) above.

(3) After successfully completing the inspections of the repair components as prescribed in item (2), the licensee.may perform a 25 percent inspection of the repair components every other cycle, if these inspections reveal degradation, the licensee should expand *he sample to 50 percent of the repair components, and the secpe of inspa. wn m subsequent cycles should follow the scope prescribed in item (1). If thess subsequent inspections also reveal degradation in the expsnded sample, the licensee should further expand the sample to 100 percent.

3.2.2 Reinspection of Repair Anchorages The inspection guidelines for repair anchorages are generally acceptable. Specific issues

, regarding the provisions for reinspection of repair anchorages are summarized in issues 3.2.21, 3.2.2.-2, 3.2.2 3 and 3.2.2-4, below, lasue 3.2.2-1: Justification for Omitting Pre-Inspections of Components Used for the Core Shroud Repair Anchorages.

The guidelines state that, in the absence of a pre-repair inspection, a licensee should perform an inspection in the first operating cycle after the repair is completed. However, the guidelines dio not justify not performing a pre-repair inspection. As discussed in Section 3.2.1, the staff recommends that licensees conduct a pre-repair inspection for all future repairs and then reinspect 100 pc. cent of the repair in the subsequent outage. Welds H8 4 and Hg thould be included in this inspection to ensure that all the repair design assumptions are trA.

Issus 3.2.2 2: Recommended Practices for Reinspections of Gussets, Brackets'and Repair Anchorages

_ in addition to the guidelines in BWRVIP-07, the BWRV!P is currently preparing another document to address inspection requirements for welds H8 and Hg, as well as the shroud support structure (which includes gussets and brackets)._ The staff recommends that licensees reinspect the repair anchorages and set the reinspection intervals according to those of the repair components, because the anchorages must retain their integrity for the repair comporents to be effective. The staff recommends that it would b1 preferable to address welds in anchorages associated with the repair hethe context of these BWRVIP-07 guidelines (rather than in a separate document). An alternative acceptable to the staff is given above under Section 3.2.1 of this SE and summarized in Table B to this SE. Licensees w .,, ..m , ,. , ,-e a.+ ..-, -, ,-

s can also take credit toward the required frequency for inspections done according to other NRC approv;d documents.

Issue 3.2.2 3: Recommendations for Technical Analyses to Verify the Effect of Cracking on Gusset Stiffness The guidelines state that minor cracking in a gusset weld would not have any impact on the effectiveness of a shroud repair end that the stiffness of the gusset is not significantly affected if there is no gross through-thickness cracking. This statement needs to be supported by technical analysis. The staff recommenos that the guidelines of BWRVIP-07 be amended to recommend that technical analyses be performed to verify the validity of this assumption if minor cracking is detected in a gusset weld.

Issue 3.2.2-4: BWRVIP-07 Recommended Intervals for Reinspection of SWR Core Shroud Repair Assemblies Repairs currently made by the industry typically consist of 4 to 10 assemblies. According to the guidelines,10 years can pass before an assembly is reinspected. The staff considers this interval too long. Currently, not enough operating data regarding repaired shrouds has been collected to justify postponing reinspections of the repair assemblies for 10 years. The staff recommends that the reinspection schedule for repair anchorages be set equivalent to

(

the intervals recommended in the staff's attemative in Section 3.2.1 of this SE unless the longer intervals L.an be technically justified. The staff also recommends adding a provision requiring sample expansion similar to that recommended in Issue 3.2.12 of Section 3.2.1 to this SE if degradation is found.

3.2.3 Reinspection of Core Shroud Horizontal and Vertical Welds For horizontal welds structurally replaced by the repair, the staff concludes that BWRVIP-07 provides acceptable guidance. As ,.,evioutly approved in the staff's SE dated September 29, 1994 (Ref.13), the horizontal welds structurally replaced by the shroud repair do not need to be reinspected. Those not replaced by the repair will be reinspected consistent with the requirements for an unrepaired shroud (Ref.13). The staff's evaluation for unrepaired horizontal welds has previously been discussed in Section 3.1 to this SE. The staff has some concems, however, about the guidelines proposed for vertical shroud welds as discussed below.

Issue 3.2.3-1: General Issue Regarding the Proposed Guidelines for Reinspection of Vertical Welds in Repaired Shrouds The staff recommends reevaluating vertical welds needed to ensure structural integrity to confirm that the der:gn essumptions remain valid. In the past, it was not considered necessary to inspect vertical welds, and such inspection was not requ: red. Since then, some licensees have made repairs (some of them preemptive), without inspecting for L. in vertical welds.

Cracking is now being reptrted in vertical welds with increasing frequency. Operating experience has now shown that a combination of residual stresses, fluence, and environment

) '

~

can cause extensive cracking in vertical welds. These welds warrant inspections that are more extensive than those proposed. The staff recommends that licensees perform baseline inspections of the vertical welds (100 percent of accessible areas, a'. least 50 percent of the total weld length) when the repair is made or at the next outage if the repair has already been made. This recommendation replaces the existing guidelines which specify using Option A or Option B.

Issue 3.2.3-2: Use of Option A or Option B for Reinspections of Vertical Welds in Repaired Core Shrouds Option A for vertical welds is acceptable except for its provision regarding the reinspection interval determined by L'.. As stated earlier in this evaluation, the staff recommends that the L',,,,, method be used to support additional operating cycle.

In contrast, the staff considers Option B inadequate. The staff recommends that licensees successfully complete a baseline inspection and a 25 poicent sampling of the total vertical weld length for every cycle. Welds should be inspected from both the ID and OD or volumetrically. Recent industry experience confirms the need for this type of inspection, having shown that cracks initiate on either ID or OD and do not extend through the wall.

The staff recommends that, in addition to the provisions of the guidelines regarding sample expansion, licensees inspect the same location of some other vertical welds at the same elevation.

The staff also recommends that licensees' samples include the vertical welds near the beltline exposed to high fluence. in selecting welds for reinspection, licensees should emphasize welds wnh high susceptibility to IGSCC or irradiation assisted stress corrosion cracking (IASCC).

3.2.4 Reinspection of Ring Segment Welds The staff has determined that the provisions in BWRVIP-07 for reinspect:ans of core shroud ring segment welds are generally acceptable. However, the staff has some concems as

discussed under issues 3.2.4-1 and 3.2.4-2, below.

Issue 3.2.4-1: Use of Option A for Reinspecbons of Ring Segment Welds Option A for ring segment welds is based on an L',,,,, approach to determine inspection scope and intervals, as distinguished from the definition in Section 2.0, above. Cption A is acceptable except, as diswswo earlier, it should be used to support continued operation for only one additional fuel cycle without repair. The staff has not declared the crack growth rate to be acceptable as referenced (Ref.11); i+ is still under review. The bounding crack growth rate of 5 x 10in/hr should be used in flaw evaluations.

Issue 3.2.4-2: Use of Option B for Reinspections of Ring Segment Welds The staffs evaluations of the methods have been discussed in Section 3.1.1 of this report.

l i

l

Very little data exist regarding the integrity of ring segment welds. Therefore, the staff '

- recommends that, to acquire baseline data, licensees should inspect all ring segment weld surfaces, including heat affected zones, unless there is adequate technical justification for not inspecting them. 'The inspection method should be capable of detecting flaws initiating from

. all surfaces (e.g., the OD, ID, bottom surface, and top surfaces), and should not be limited to inspecting just the OD or ID.

The staff considers that Option B requires examining too small a sample over too long a period. The accessible areas of these welds are very limited. Therefore, it is desirable that licensees apply thls sampling only after performing a baseline inspection of all accessible regions at each ring segment weld, with the total inspected length at each weld encompassing at least 50 percent of the weld length and then sampling 25-percent of the total number of ring segment welds every cycle.

Finally, the staff recommends that licensees' samples include the ring segment welds near the beltline exposed to high fluence, in selecting welds for reinspection, licensees should emphasize welds with high susceptibility to IGSCC or IASCC.

4.0 RECOMMENDATIONS AND CONCLUSIONS As orsviously stated, the staff finds much of the guidance in BWRVIP-07 9 .eptable, it is consistent with guidance in previous submi'tals and incorporates recom .dations in previous safety evaluations. However, the staff also finds the guidant implicated and not sufficiently conservative. The staff recommends that the BWRVIP resoNe the issues raised in this SE. These issues, discussed in Sections 3.1.1, 3.1.2. 3.2.1, 3.2.2, 3.2.3, and 3.2.4 of this SE, are restated (in bullet form) in Sections 4.1,4.2,4.3,4.4, and 4.5 below for clarity.

4.1 Guidelines for Unrepaired Core Shrouds in Category B and C Plants issues 3.1.1-1, 3.1.1 3, 3.1.2-1, and 3.1.2 2

. Inspection strategies that cite L,,,n and L', to justify multi-cycle operation (for Category B and C plants) without inspection should be revised to justify operation for only one cycle.

Issue 3.1.12 The 10-year reinspection interval used in Option A for Category B plants should be reduced to 6 years and the 8-year reinspection interval should be reduced to 5 years.

The inspection intervals (stated in years) shoulc correspond reasonably well to operating cycle lengths.

Issues 3.1.1-4 and 3.1.2 4 -

The reinspection guidelines do not require the inspection of vertical welds even when inspection reveals cracking of circumferential welds. The staff has concems about the functionability of the shroud when a vertical weld has completely failed and the --

_J _ _ - - ___-___- __-__-_.

1 e

circumferential weld intersecting it has extensively cracked. The staff requests that BWRVIP-07 provide for inspecting vertical welds to assure their integrity when cracking in the intersecting circumferential welds is found. The provision should also be incorporated into the relevant BWRVIP document for baseline inspection of the core shrouds.

Issue 3.1.21

. The guidelines in BWRVIP-07 state that if a licensee cannot inspect Category C core shroud for L,,,,, the acceptability and inspection interval of the weld are to be determined on a case by-case basis. The evaluation addressing these issues should be submitted for NRC review and approval.

Issue 3.1.2 * !ssue 3.1.2 3 lists a number of concems about Table 1 that Thould be resolved. The staff has provided an ae.eptable attemative to Table 1 of the BWRVIP-07 guidelines in its summary of Issue 3.1.2 3. This attemative is also summarized in Table A of this SE.

4.2 Guidelines for Reinspection of Components in Repaired Core Shrouds issues 3.2.1 1 and 3.2.1-2 The staff recommends incorporating the following changes into the existing BWRVIP-07 guidelines:

• Require a 100 percent inspection after one cycle to find installation errors.

. Specify other ways of verifying bolt tightness where vi5 Jal examination is not feasible or adequate.

• Require expanding the sample if degradation is found in any of the repair components.

= ' Specify a review of the design assumptions regarding the repair.

Having considered the need for baseline data on how repairs are performing, and in view of recent operating experience, the staff finds that the scope and frequenP S proposed in BWRVIP-07 are inadequate. Consequently, the staff recommends an attemative for reinspecting repair components, as discussed in Section 3.2.1 of this SE and summarized in Table B of this SE.

4.2 -Guidelines for Reinspection of Repair Anchorages issues 3.2.2-1, 3.2.2-2, 3.2.2-3 and 3.2.2-4

The staff recommends incorporating the following changes into the existing BWRVIP guidelines:

• Specify a pre-repair inspection for all future repairs and reinspection of 100 percent of the repair components after the first operating cycle following the installation of the repair modification.

• Address reinspections of the welds in anchorages associated with the repair in these BWRVIP-07 guidelines (rather than in a separate document). An attemative acceptable to the staff in scope and frequency is given under Section 3.2.1 and summarized in Table B of this SE.

• Require that, if minor cracking is detected in a gusset weld, technical analyses be performed to verify the assumption that the cracking at the gusset weld would not have any impact on the effectiveness of the shroud repair.

• Reduce the interval between inspections of the anchorages unless the longer interval can be technically justified.

• Add a provision requiring that licensees expand the sample if degradation is found.

4.4 Guidelines for Reinspection of Horizontal and Verlical Welds in Repaired Core Shrouds issue 3.2.3-1:

The staff recommends incorporating the following changes into the existing BWRVIP-07 guidelines:

. State that vertical welds needed for structural integrity be roevaluated to confirm that the original design assumptions remain valid.

- State that a baseline inspection (100 percent of accessible areas, at least 50 percent of the total weld length) of the vertical welds should be performed when the repair is installed, or at the next outage if the repair has already been made. (This inspection should be specified in place of those specified by Options A and B.)

Issue 3.2.3 2:

The staff recommends incorporating the following changes into the existito BWRMP-07 guidelines:

• Specify that, for Option A for vertical welds, the inspection intervals detem1ined from the L'm methods prescribed in Section 4.4.1 of BWRVIP-07 be revised to justify only one additional cycle of operation.

Require licensees to inspect 25 percent of total vertical weld length every cycle after the baseline inspections of the vertical welds are complete.

k

• Specify inspecting welds visually from both the ID and OD, or volumetrically.

• State that in expanding the vertical weld inspection sample, licensees should inspect the same location of some other vertical welds at the same elevation.

• State that licensees' sampling should include the vertical welds near the beltline which are exposed to high fluence.

4.5 Guidelines for Reinspection of Ring Cegment Welds in Repaired Core Shrouds issue 3.2.41:

The staff recommends incorporating '.he following changes into the existing BWRVIP-07 guidelines:

• The NRC approved crack growth rate of 5 X 104 inches /hr should be specified in lieu of the growth rate specified in EPRI Topical Report TR-105873 (Ref.11) .

Issue 3.2.4 2:

The staff recommends incorporating the following changes into the existing BWRVIP-07 guidelines:

• Use the staffs recommendations regarding the calculation methods for the unrepaired core shrouds for Category B plants.

+ State that the inspriction method should be capable of detecting flaws initiated from all surfaces (the OD, ID, bottom surface, and top surface), and not be limited to just the OD or 10.

• State that, after successfully completing a baseline inspection (100 percent of accessible areas, at least 50 percent of the weld length), licensees may apply a 25-percent sampling of the total number of ring segment welds during each subsequent refueling outage .

State that licensees' sampling should include the ring segment welds near the beltline exposed to high fluence.

In additiory the staff recommends that the slowest crack growth rate used for analyses should be 5 x 10 inch per hour (the crack growth rate currently approved by the NRC).

5.0 BIBLIOGRAPHY AND REFERENCES

1. February 29,1996 - Letter from J.T. Beckham, Chairman - BWRVIP, to the U.S.

Nuclear Regulatory Commission Document Control Desk, ~BWR Vessel and intomal Project, Guidelines for Reinspection of BWR Core Shrc,uds (BWRVIP-07)' EPRI Report TR-105747, February 1996."

~

c --

-y f I,

- 2. July 17,1996 - Letter from V. Wagoner, Chairman - BWRVIP Integration -

Committee, to C. Carpenter, U.S. Nuclear Regulatory Commission: " Clarification to Core Shroud Reinspection Guidelines."

3. -May 21,1996 - Letter from C.E. Carpenter, Lead Project Manager, U.S. Nuclear Regulatory Commission, to J.T. Beckham, Jr., Chairman - BWRVIP, submitting

" Request for Additional Information - Review of BWR Vessel and Intemals Project -

Proprietary Report, 'BWR Vessel and Intemals Project, Guidelines for Reinspection of BWR Core Shrouds (BWRVIP-07).";

4.- October 21,1996 - Letter from Larry Stewart, Acting on behalf of Robin Dyle, BWRVIP Technical Chairman, to the U.S. Nuclear Regulatory Document Control Desk,

- submitting the "BWRVIP Response to Request for Additional Information Regarding Proprietary Topical Report EPRI TR-105747."

5. January 8,1997 - Letter from W. Bilanin, Acting on behalf of Robin Dyle, BWRVIP Technical Chairman, to the U.S. Nuclear Reguictory Document Control Desk, submitting the " Modification to BWRVIP Response to NRC Request for Additional Inforn;ation on BWRVIP-07."

6. September 2,1994 - General Electric Company Topical Report No. GENE 523113-0894, "BWR Core Shroud Inspection and Evaluation Guidelines."

7. March 1995 - General Electric Company Topical Report No.

GENE 523-113-0894, Rev.1, "BWR Core Shroud Inspection and Flaw Evaiustien Guidelines."

8. December 28,1994 - Letter from B.W. Sheron, Director -Division of Engineering, U.S. Nuclear Regulatory Commission, to J.T. Beckham, Chairman - BWRVIP:

" Evaluation of 'BWR Shroud Cracking Generic Safety Assessment', Rev.1, GENE-523-A107P 0794, August 5,1994 and 'BWR Core Shroud Inspection and Evaluation Guidelines,' GENE 523113-0894, September 2,1994."

9. June 16,1995 - Letter from B.W. Sheron, Director - Division of Engineering,-U.S.

Nuclear Regulatory Commission, to J.T. Beckham, Chairman - BWRVIP: " Evaluation of 'BWR Core Shroud Inspection and Evaluation Guidelines,' GENE-523-113-0894, Rev.1, dated March 1995, and 'BWRVIP Core Shroud NDE Uncertainty and Procedure Standard,' dated November 22,1994."

10. October 1995 - Electric Power Research Institute Topical Report No.~ TR-105692,'" Guide for Format and Content of Core Shroud Repair Design -

Submittals (BWRVIP 04).",

11. November 1995 - Electric Power Research Institute Topical Report No. TR-105873, 2nd Draft, " Evaluation of Crack Growth in BWR Stainless Steel RPV Intemais."

12. September 13,1994 - Letter from J.T. Beckham, Chairman - BWRVIP, to

. ~ . . - . . . -- __ -- .

k D.S. Brinkman, Senior Project Manager, U.S. Nuclear Regulatory Commission:

" Response to NRC Request for AdditionalInformation (RAI) Regarding Boiling Water Reactor (BWR) Core Shroud Repair Design Criteria," including Attachment 1,

" Response to NRC Questions," and Attachment 2, " Revision 1 of the Core Shroud Repair Design Criteria."

13. September 29,1994 - Letter from B.W. Sheron, Director - Division of Engineering, U.S, Nuclear Regulatory Commission, to B. McLeod, Chairman - BWRVIP Repair Technical Subcommittee: " Safety Evaluation on Boiling Water Reactor BWR Core Shroud Repair Design Criteria."

1 h.

Table' A: Staff Alternative to Table 1 of the BWRVIP-07 Guidelines -

Cracking (%) Inspection Interval

.%=0 6 yrs or 4_ cycles, wb!chever comes first 10 h % > 0 5 yrs or 3 cycles, whichever comes first 25 h % > 10 4 yrs or 2 cycles, whichever comes first

. 40 m % > 25 Next outage

% > 40 p! ant specific analysis needed to operate 1 cycle without repair

-Table B: Staff Attemative for inspecting Repair Components and Anchorages What When Frequency Expansion (1)lnspect to ensure Pre-repair inspection N/A N/A all design for all future repairs assumptions are met 100% of repair inspect after If no degradation is if any degradation is components and completion of repair found, go to (2) found, inspect 100%

anchorages or after one cycle if each outage until no repair already done. degradation is found (2) 25% sampling After (1) is Each cycle for 4 If degradation is successful. cycles. If no found, expand to all degradation is found, repair components go to (3) and follow (1)

(3) 25% sampling After (2) is Every other cycle if degradation is successful found, expand inspections to 50%

<. of the repair components. if further degradation is found, expand to 100 %.

_ - - _ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ .