ML20155F930

Forwards Trip Rept Provided by Cv Dodd,Documenting Findings as Result of Review of 930512 & 13 Participation in Special Review of Eddy Current Insp Program of SGs at Kewaunee NPP

ML20155F930
Person / Time
Site:	Kewaunee
Issue date:	08/16/1993
From:	Strosnider J Office of Nuclear Reactor Regulation
To:	Wright G NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
Shared Package
ML20155F799	List: ML20155F795 ML20155F930 ML20155G054
References
FOIA-97-488 50-305-93-05, 50-305-93-5, NUDOCS 9811060156
Download: ML20155F930 (19)

Text

- - - - - - - -

y

/  % *s

')' y E UNITED STATES I NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20$55-4001 k P "

• j#

a n no MEMORANDUM FOR: Geoffrey C. Wright, Chief Engineering Branch -

Division of Reactor Safety Region III l

FROM: Jack R. Strosnider, Chief Materials and Chemical Engineering Branch Division of Engineering g gfa

SUBJECT:

KEWAUNEE NUCLEAR POWER PLANT - CONTRACTOR TRIP REPORT FOR l INSPECTION REPORT NO. 50-305/93005 l On May 12 and 13, 1993, a contractor for the Office of Nuclear Reactor Regulation (NRR) from Oak Ridge National Laboratory (ORNL) participated in a special review of the eddy current inspection program of the steam generators l

at the Kewaunee Nuclear Power Plant with Mr. J. F. Schapker of Region III and .

l Mr. K. J. Karwoski of NRR. At the conclusion of the review, the findings were l

! discussed with members of the Wisconsin Public Service Corporation (WPSC) staff. WPSC is the licensee for the Kewaunee Nuclear Power Plant.

I Attached is the trip report provided by the NRR contractor, Dr. C. V. Dodd, documenting his findings as a result of the review. In Dr. Dodd's trip report, he notes that as a result of one of the comments that was made during l

a meeting with the licensee's staff during the site visit, that an error was detected in the mixing algorithm in the eddy current data analysis software used at Kewaunee. Dr. Dodd also notes that the software error was corrected and that the owner of the software, Zetec, has acknowledged that the detection of small signals will be improved as a result of the modification to the software. Furthermore, Dr. Dodd observes that the reduction in the mixing noise as a result of modifying the software is not known.

As a result of these findings and additional discussions with the NRC staff, the licensee has subsequently evaluated the effects of modifying the mixing algorithm as documented by letter dated June 21, 1993, " Additional Information on Eddy Current Data Reanalysis." In the June 21, 1993 letter, theilicensee I concluded that the changes in the signals from indications derived from the  ;

new version of the software program versus the old version of the program were  ;

\ insignificant.) This conclusion was based on the licensee's preliminary l

assessment of the change in the software and a limited scope reanalysis by Zetec of the licensee's 1993 eddy current data using both the old" and "new" (i.e., corrected) version of the eddy current data analysis software.

Zetec, Inc. issued a letter dated June 1,1993, notifying the customers of the affected software program of the anomaly in the mixing algorithm used to calculate and display simple two frequency mixes, and their intent to make available a corrected version of the program. If you have any questions on 10 g 6 981103 k WILLIAM 97-488 PDR s qwoGotSG

• 4 this matter, please contact me at (301) 504-2796.

ORIGIN 8V:

Jack R. Strosnider, Chief Materials and Chemical Engineering Branch Division of. Engineering

Enclosure:

As stated cc: B.D. Liaw T.O. Martin DISTRIBUTION: Central Files EMurphy EMCB RF/PF KKarwoski HConrad GJohnson JStrosnider AHansen RLaufer JSchapker NRC & Local PDRs -

0

• SEE PREVIOUS CONCURRENCE 1

• DE:EMCB *DE:EMCB DE:EMCB ,DE:EMCB KKarwoski:tc EMurphy GJohnson JStrosnider 08/03/93 08/03/93 08//E/93 08/l$/93 0FFICIAL RECORD COPY t

t

i-European Plugging Criteria for

,' Defects at Tube Support Plates NP XXXX Research Project $404 30 February 1991 l l

Prepared by DOMINION ENGINEERING. INC.

6862' Elm Street McLean, Virginia 22101 Principal Investigator ,

J. A. Gorman l

Prepared for Electric Power Research Institute 3412 Hillview Avenue Palo Alto California 94303 EPRI Project Mariagers C. S. Welty and C. L. Williams Steam Generator Reliability Program Nuclear Power Division v

Oraft 2/4/91

--T-i O 3 g &&3 ? > 28 #

legal Notice (by EPRI) .

i 1

)

l l \

i l l

7 i

i I

l

\

r 1

l l

Draft 1/31/91 11

e l t

a ABSTRACT The occurrence of intergranular attack and stress corrosion cracking (IGA / SCC) at tube outer surfaces in tube support plate (TSP) crevices is becoming more prevalent in certain types of PWR steam generators. There are concerns in the industry that this type of corrosion could affect significant numbers of tubes. It is also believed that the plugging criteria currently being used for such defects in the USA are unnecessarily restrictive and, unless revised, will result in the unnecessary repair of many tubes. The criteria ir.

the USA typically require repair of tubes with defects over 40% of tube wall in depth, regarJ1ste of defect location or length.

Plugging criteria for IGA / SCC defects at 1SPs in Europe are, in several cases, much less restrictive than in the USA. For example, this type of defect 4

does not require repair of tubes in Belgium or France, regardless of defect depth. Alternate tube plugging criteria (ATPC) for defects at TSPs have also been developed in Sweden and' Spain. These ATPC are in the process of being reviewed with national safety authorities and will allow, if approved. defects at TSPs up to 70 or 78% of the tube wall to remain without repair.

The purpose of this report is to document European practices and to serve as a reference for an.EPRI coordinated committee in the USA that is developing alternate tube plugging criteria for defects at TSPs.

l l

l l

l 1.

l I

Draft 2/4/91 iii l'

. . . ... . ...- --.... - . = - - - . . . . . . . - . . - . .- . ..- .. .

9 (blank page) .

Draft 1/31/91 iv

. . . . _ . . _ _ . _ .. .._ _ _ _._.,__ .m. . _ . _ . . _ _ . _ _ . . . . . . . _ . . _ . . . . . . _ . _

e

.c 1

i ACKNOWLEDGEMENTS I l

Thanks are extended to Belgatom. Electricite de France and Framatome, the

{

Spanish Utilities, and the Swedish State Power Board for making the information  !

upon which this report is bases available, and for their reviews of this report.

)

L i,

E

! Draft 1/31/91 v l

I 1

1.

(blank Pagai i .

i s.

i.

4 1

t 4

e i

l l

4 i

i

? l e 1 1 <

1 Y

l i l

?

i e

i J

Oraft 1/31/91 vi

CONTENTS C

Section 1.

g Introduction

2. Summary of Information Obtained

3. References Appendices A. Visit with Swedish. State Power Board B. Visit with Belgatom C. Visit with Electriciti de France and Framatome D. Visit with Spanish Utilities' i

-e,u.

4 l

t i

Draft 1/31/91 vii I

l EXECUTIVE

SUMMARY

PURPOSE The purpose of this project is to document European experience with alternate tube plugging criteria (ATPC) for defects due to intergranular attack and stress corrosion cracking (IGA / SCC) at tube support plates (TSPs) of PWR steam generators. This information will serve as a reference for efforts in the USA directed at development and licensing of ATPC for this type of defect.

BACKGROUND IGA / SCC at TSPs has recently become a significant concern to many utilities as a result of the increasing rate of occurrence of this type of defect. For example, it is a major factor in the recent decision by Spanish utilities to replace the steam generators at four PWRs starting in the mid 1990s (1).

Special concern about IGA / SCC at TSPs was generated by experience with a pulled tube from Farley 1 (2). A pulled tube that had, based on eddy current test (ECT), no detectable degradation, was found in laboratory examination to have IGA / SCC defects up to 62% of the tube wall thickness, which exceeded the plugging criteria of 40%. The NRC advised utilities (2) to take a conservative position with regard to interpretation of ECT signals for defects at TSPs, and to plug tubes with with questionable signals. This sequence of events raised serious concerns in the industry because it was believed that such conservative interpretation of ECT signals would result in plugging of many tubes, and was not warranted from either safety or reliability standpoints.

Because of the concerns about possible unnecessary plugging of many tubes due to IGA / SCC defects at TSPs, an EPRI coordinated effort was initiated to develop ATPC for such defects. The intent of these ATPC is to ensure that satisfactory safety and operability margins are maintained while not requiring plugging of small defects that do not pose safety or operability risks. There is confidence in the industry that margin exists such that defects well over the current 40% limit are not a concern. This confidence is based on pulled tube examinations and tests performed by Westinghouse for D. C. Cook (1, 1), and based on tests by Europeans that show that IGA / SCC type defects at TSPs do not reduce tube burst pressures, i.e., that tube burst occurs in the free span, away from the TSP area. The minimal effect of defects at TSPs on burst pressures is Draft 2/4/91 S-1

attributed to the support provided to the tube by the TSP, which prevents

, opening up and propagation of the defects.

Several European countries have performed tests and developed ATPC for defects at TSPs, for similar reasons that they are being developed in the USA.

The intent of this report is to document the European experience, so that the USA effort can build on this experience.

APPROACH A team of EPRI and Dominion Engineering Inc. (DEI) personnel visited utility and contractor personnel in Belgium, France, Spain, and Sweden in August and September 1990 to hold in-depth discussions regarding development and application of ATPC for IGA / SCC defects at TSPs. The results of these visits.

were documented in letter type reports and sent to the visited groups for review. After incorporating the resolution of comments into the reports. the .

letter type reports were issued to EPRI and the visited organizations. The information in these letter type reports is documented in this report. A draft of this report was also reviewed by all of the visited organizations, and al.1 comments were resolved.

CURRENT STATUS OF ATPC FOR IGA / SCC DEFECTS AT TSPS IN EUROPE In two countries, Belgium and France, current plugging criteria (as of September 1990) do not require defects of any depth at TSPs to be plugged. The main bases for these criteria are that tests' have shown that tubes with severe simulated secondary side defects at TSPs do not have reduced tube burst strength, and that the tubes burst in free span areas, rather than at the TSPs. j In two other countries Spain and Sweden', current plugging criteria (as of I September 1990) for defects at TSPs are the same as for free span defects, i.e., i 40% of tube wall in Spain and 50% in Sweden. However, in both of these countries ATPC have been developed'and proposed to safety authorities. These ATPC justify leaving defects up to 70% or 78% depth (in Sweden and Spain respectively) at TSPs without repair. The 70% and 785 values include allowances for defect. growth and for inspection error. The Swedish and Spanish utilities hope to have the ATPC approved in the near future.

OCCURRENCE OF IGA / SCC AT TSPS Examination of pulled tubes has shown the IGA / SCC at TSPs to mainly consist of axial cracks. The IGA / SCC cracks have almost always been located within the edges of the TSP. The one known exception occurred at a flow distribution baffle, not a TSP, where sludge was present on top of the baffle.

Draft-2/4/91- S-2

The severi%y of occurrence of IGA / SCC at TSPs in the countries visited

-varies. -However, the numbers'of units affected is increasing, and the numbers of tubes affected and'the. sizes of the defe'ts c at individual tubes are also-increasing. In two countries this form of corrosion has been a significant. .

factor in recent decisions regarding steam generator strategic planning: In ,

Sweden,-the operating temperature of one unit has been reduced in order to reduce the rate of increase of IGA / SCC as well as PWSCC. In Spain, IGA / SCC at TSPs was a factor in a recent decision to initiate replacement of steam generators at four units starting in 1994.

AREAS OF CONCERN AND INVESTIGATION TSP Deflection Possible deflection of TSPs during postulated accidents was identified as a significant issue in Sweden. The concern is that the TSP will be deflected during an accident out of its normal position, thus removing the support given by the TSP from the area with the defects. Because lower TSPs deflect less than higher ones, the Swedish intend to address the TSP deflection concern by applying ATPC to only the lower five TSPs.

Evaluation of TSP deflection in France has indicated that, for French Model 51 steam generators, TSP deflections are small (e.g., 2 mm (.08 in.) maximum),

and are not a concern.- It was noted that TSP deflection is dependent on specific design featu'res, such as numbers of tie rods and their diameter, and that TSP deflection at other Model 51s could differ from that determined for French Model Sis.

Effects of the possible deflection of TSPs from hot full power positions caused by differential pressures across TSPs during accidents have not as yet been considered in Belgium or Spain. The work being performed as part of the EPRI technical support effort for ATPC is being relied upon to determine needs in this area. .

Post Accident Leak Rates As a result of work on ATPC for roll transitions in the USA, an issue has been identified regarding the amount of primary to secondary leakage that could occur subsequent to a postulated accident, such as a steam line break or a feedwate- line break. The concern is that leakage through defects at roll transitioc: or at TSPs.could, after the accident, exceed current limits if, as a result of the accident, the defects are opened up by exposure to the accident

. differential pressure of about 2650 psi (183 bar).

Draft 2/4/91 S-3

d e

The approaches being taken regarding this concern differ in the various '

, countries.

The Swedish approach is to limit defect depths at TSPs to values that provide assurance that little post accident leakage could occur . In Belgium. limits on allowed post accident leakage are being increased by analysis. -

In France and Spain concerns about post accident leakage have no been addressed.

The work being performed as part of the EPRI ATPC effort is being followed to determine needs in this area.

Quality Status of TSPs in Sweden, a concern has been raised as to what needs to be done to allow I credit to be taken for TSPs considering that they were not designed and built as safety grade equipment. ,

This concern had not been addressed in any of the other countries visited.

Insnection criteria The NDE methods used to quantify the amount of degradation present at a TSP due to IGA / SCC was reviewed at each of the countries visited.

The groups visited indicated that they are supportive of the EPRI effort to develop an ATPC technical support document based on correlations of tube burst and leakage behavior with ECT signals (amplitude and phase angle), without attempting to measure quantitative depths or lengths of defects. However, until these results

'are available, two of the. countries. Sweden and Spain, will continue to pursue their quantitative defect depth limit (70% and 78%) type of criteria rather than direct ECT signal criteria.

Need for Further Data It was noted during the visits that part of the EPRI effort should inclu'de assembling information regarding the actual morphology of defects observed at TSPs and the correlation of these defects with NDE results, especially bobbin coil ECT results.

REFERENCES 1.

Minutes of Steam Generator Reliability Project Technical Advisory Group meeting, October 23-25, 1990 Palo Alto CA.

2.

NRC Information Notice No. 90-49, " Stress Corrosion Cracking in PWR Steam Generator Tubes," USNRC, Washington, D. C., August 6. 1990 i

- Oraft 2/4/91 S-4

3. WCAP-11056, "NRC Presentation Report on Steam Generator Tube Integrity for ,

D. C. Cook Unit 2, January 1986", Westinghouse Electric Corporation, Pittsburgh, PA, in NRC Public Document Room, ACN 8602130275, Date 860131, POR A00CK 05000316 4 WCAP-11330 "NRC Presentation Report on Steam Generator Tube Integrity for D. C. Cook Unit 2, September 1986", Westinghouse Electric Corporation, Pittsburgh, PA, in NRC Public Document Room, ACN 8612020058, Date 861124, POR ADOCK 05000316 i

d Draft 2/4/91 5-5

Section 1 INTRODUCTION PURPOSE The purpose of this project is to document European experience with alternate tube plugging criteria (ATPC) for defects due to intergranular attack and stress corrosion cracking (IGA / SCC) at tube support plates (TSPs) of PWR

-steam generators.

This information will serve as a reference for efforts in the USA directed at development and licensing of ATPC for this type of defect.

BACKGROUND 1

IGA / SCC at TSPs has recently become a significant concern to many utilities as a result of the increasing rate of occurrence of this type of defect. For example, this' type of defect was cited as a major reason that the Cook 2 steam generators were replaced, rather than repaired by sleeving (1). In addition, it has resulted in over 20% tubes being affected at three Japanese units, and is.a major factor in the recent decision by Spanish utilities to replace the steam i

! generators at four PWRs starting in the mid 1990s (2, 2).

IGA / SCC ca occur in several secondary side locations. These include deep tube sheet crevices, in sludge piles, at eggerate support grids, and in crevices at drilled hole tube support plates (TSPs). IGA / SCC at this last location is the focus of this report.

The reasons for this focus are that IGA / SCC at drilled hole TSPs is of special concern to the industry because it is not as l amenable to control or repair (e.g., Dy sleeving or removal of sludge piles) as

. are IGA / SCC at some of the other locations, and because it has begun to affect a j

large number of plants with high operating temperatures (e.g., over about 605'F l

( 318*C ) , drilled hole TSPs, and mill annealed alloy 600 tubing, Special concern about IGA / SCC was generated by experience with a pulled tube from Farley 1 (1). A pulled tube that had, based on eddy current test

[ (ECT), no detectable degradation, was found in laboratory examination to have

!. IGA / SCC defects up to 62% of the tube wall thickness, which exceeded the

. plugging criteria of 40%.

The NRC advised utilities (1) to take a conservative i-position with regard to interpretation of ECT signals for defects at TSPs, and to plug tubes with with questionable signals. This sequence of events raised i

serious concerns in the industry because it was believed that such conservative L

Draft 2/4/91 1-1 l

interpretation of ECT signals would result in plugging of many tubes that das not' warranted from either safety or reliability standpoints.

Because of the concerns about possible unnecessary plugging of many tubes because of IGA / SCC defects at TSPs, an EPRI coordinated effort was initiated to develop alternate tube plugging criteria (ATPC) for such defects. The intent of these ATPC is to ensure that satisfactory safety and operability margins are maintained while not requiring plugging of small defects that do not pose safety or operability risks. There is confidence in the industry that margin exists such that defects well over the current 40% limit are not a concern. This confidence is based on pulled tube examinations and tests performed-by Westinghouse for D. C. Cook (1. 1), and based on tests by Europeans that show that IGA / SCC type defects at TSPs do not reduce tube burst pressures, i.e. that tube burst occurs in the free span, away from the TSP area. The minimal effect of defects at TSPs is attributed to the support provided to the tube by the TSP.

! which prevents opening up and propagation of the defects.

• I Several European countries have performed tests and developed ATPC for defects at TSPs for similar reasons that they are being developed in the USA.

The intent of this report is to document the' European experience, so that the USA effort can build on this experience.

APPROACH A team of EPRI and Dominion Engineering. Inc. (DEI) personnel visited <

l-utility and contractor personnel in Belgium France. Spain, and Sweden in August j and September 1990 to hol'd in depth discussions regarding development and application of ATPC for IGA / SCC defects at TSPs. The results of these visits were documented in letter type reports and sent to the visited groups for review. After incorporating the resolution of commerits into the reports, the x

letter type reports were issued to EPRI and the visited organizations. The

~

information in these letter type reports is documented in this report. A draft of this report was also reviewed by all of the visited organizations, and all comments'were resolved.

TERMINOLOGY Brief definitions of the terms and acronyms used in this report are given below.

AI.PC. Alternate Tube Plugging Criteria.

ICI. Eddy current test.

Draft 2/4/91 1-2

IGA (Interaranular attack). Intergranular attack is corrosion attack at grain boundaries that tends to occur at all grain boundaries in the affected area. It can occur on either the secondary or primary side but, in this report, refers to attack on the secondary side.

IGA / SCC fintercranular attack / stress corrosion crackino). Intergranular attack / stress corrosion cracking is corrosion attack on the secondary side that involves both intergranular attack and stress corrosion cracking. The stress corrosion cracking is typically intergranular, but is occasionally transgranular.

IGSCC finteraranular stress corrosion crackino). Intergranular stress corrosion cracking is corrosion attack involving isolated cracks at grain boundaries propagating in regions with tensile stress. When it occurs on the pr{ mary side, it is called PWSCC; thus, the term IGSCC usually refers to attack on the secondary si_de.

Model D3 A Model D3 is a steam generator designed by Westinghouse that includes a split flow preheater.

Model 51 A Model 51 is a non preheat steam generator whose original design was developed by Westinghouse, but which was also built, with some design modifications, by licensees (Framatome and Mitsubishi).

RDE. Nondestructive examination.

PWSec f primary water stress corrosion crackina). Intergranular stress corrosion cracking that occurs in alloy 600 on the primary side.

RPc (Rotatino cancake coil). This refers to a specialized eddy current test probe or method often used to obtain an improved characterization of tube degradation.

3321. Swedish State Power Board.

ISE. Tube support plate.

Oraft 2/4/91 1-3

TWD (Throuah wall denth). This indica 8es the portion of the wall thickness of the_ tube that has been penetrated by a defect.

• i l

1 1

4 I

p i

l.

l l

i I

' Draft 2/4/91 1-4 l

Section 2

SUMMARY

OF INFORMATION OBTAINED Detailed reports of the visits made to Sweden. Belgium. France and Spain are contained in Appendices A, B. C. and D respectively. A summary of the information obtained is given below.

CURRENT STATUS OF ATPC FOR IGA / SCC DEFECTS AT TSPS IN EUROPE in two countries. Belgium and France, current plugging criteria (as of September 1990) do not require defects at TSPs to be plugged. The main bases for these criteria are that tests have shown that tubes with severe simulated secondary side defects at TSPs do not have reduced tube burst strength, and that the tubes with TSP defects burst in free span areas, rather than at the TSPs.  !

In two other countries. Spain and Sweden, current plugging criteria (as of )

September 1990) for defects at TSPs are the same as for free span defects, i.e..

40% TWD in Spain and 50% TWO in Sweden. In both of these countries. ATPC have been developed and proposed to safety authorities:

At the time of the visit. September 1990, proposed ATPC had been submitted to, but not yet accepted by. Spanish safety authorities. The proposed criteria would allow defects up to 78% TWO to not be repaired. l The main bases for the proposed criteria are that burst tests have shown 100% TWD defects that extend a short distance beyond the edges of the TSP do not reduce burst strength below accident pressures. For operational reliability reasons, it is desired to avoid development of leaks at TSPs. For this reason, an allowance of 7% for defect growth and 15% for defect deptr. measurement error are subtracted from 100% to arrive at the proposed limit of 78% TWD.

ATPC for TSPs were proposed in mid 1990 to Swedish safety authorities.

These criteria are that defects up to 70% TWD were acceptable. The 70 value is based on ensuring that defects do not grow through wall during the next operating cycle, and include allowances of 10% for growth and 20% for NDE uncertainty. In 1990. Swedish safety authorities did not grant full approval of the proposed criteria, but did acce'pt them for a limited number cf tubes for one year. The Swedish safety authorities t

wanted more time for evaluation of the criteria. In addition, there were concerns with regard to possible TSP deflection during postulated Draft 2/4/91 21

accidents, and concerns related to the TSPs not being safety grade equipment. Additional work addressing these issues has been performed subsequent to the initial submittal.'SSPB indicated that they intend to resubmit the proposed criteria and hope to have them approved for use during the summer 1991 inspections of Ringhals 3 OCCURRENCE.0F IGA / SCC AT TSPS

! Examination of pulled tubes has shown the IGA / SCC at TSPs to mainly consist <

of axial cracks. The IGA / SCC cracks have essentially always been located within the edges of the TSP. The one known exception occurred at a flow distribution baffle, not a TSP, where sludge was present on top of the baffle. Due to the low elevation of the flow distribution baffle, which promotes accumulation of

. sludge, this experience is not considered applicable to TSPs. In addition, the Spanish indicate that, because of the larger holes in flow distribution baffles, application of ATPC to defects at flow distribution baffles is not justified.

The severity of occurrence of IGA / SCC at TSPs in the countries vfsited is

, as follows:

• In Sweden, the number of pluggable defects detected by ECT to date has been small. However, examination of pulled tubes has indicated that most tubes at Ringhals 3 and 4 have shallow IGA at TSPs and that significant numbers of tubes in Ringhals 3 have SCC at TSPs that was not detectaM ' by ECT (no tubes at Ringhals 4 have been found to have SCC at TSPs even by examination of pulled tubes). Concerns about IGA / SCC. as well as with PWSCC. led to reducing the hot leg temperature of Ringhals 3 from 321*C (610*F) to 308.5'C (587'F) .

. In Belgium two of seven units (Doel 4 and Tihange 1) have exhibited IGA / SCC at TSPs by ECT examination or by leakage. The amount of degradation is limited. Because of the limited extent of the attack, and since plugging criteria do not require such defects to be plugged.

IGA / SCC at TSPs is not considered a serious problem in Belgium.

. In France, about 10 units were affected by IGA / SCC at TSPs as of September 1990, including three plants where it was confirmed by examination of pulled tubes. It was noted by EdF that the. numbers of plants affected is steadily increasing, and the number and size of )

defects in individual plants is also steadily increasing. However. l because plugging criteria do not require plugging of IGA / SCC at TSPs.

it is not considered a serious problem in France.

Oraft 2/4/91 22 I

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

o In Spain e all four units with Model 03 steam generators are experiencing significant amounts of IGA / SCC at TSPs.

The amount of attack is increasing steadily at all four units. Current plugging criteria in Spain for these defects require them to be repaired if over 40% TWD.

The incseasing IGA / SCC at TSPs. together with PWSCC in the roll transition area, was a major factor in the decision to replace steam generators in these four units starting in 1994 AREAS OF CONCERN AND INVESTIGATION TSP Deflection Possible deflection of TSPs during postulated accidents was identified as significant issue in Sweden.

The concern is that the TSP will be deflected during an accident out of its normal position, thus removing the support giv by the TSP from the area with the defects.

Evaluation of possible deflections of TSPs in the Model 03 steam generators in Sweden indicate that only the highest TSPs are deflected enough to be a concern (e.g., up to 13.5mm in.) (

for the top support),

and that the lower five TSPs are not sufficiently deflected to be a concern.

As a result, the Swedish intended to apply ATPC to only the lower five TSPs.

Evaluation of TSP deflection in France has indicated that, for French Model 51 steam generators. TSP deflections are small (e.g., 2 mm (.08 in.) maximum and are not a concern. ,

It was noted that TSP deflection is dependent on specific design features, such as numbers of tie rods and their diameter, and that TSP French Model deflection 51s. at other Model 51s could differ from that determ Effects of the possible deflection of TSPs from hot full power positions caused by differential pressures across TSPs during accidents have not as yet been considered in Belgium or Spain.

The work being performed as part of the EPRI technical support effort for ATPC is being relied upon to determine needs in this area.

Post Accident Leak Rates As a result of work on ATPC for roll transitions in the USA, an issue has been identified regarding the amount of primary to secondary leakage that could occur subsequent to a postulated accident, such as a steam line break or a feedwater line break. The concern is that leakage through defects at TSPs could,-af ter the accident, exceed current limits if, as a result of the accident, the defects are opened up by exposure to the accident differential Draft 2/4/91

, 23

pressure of about 2650 psi (183 bar). For example, 2he accident differential pressure could cause part depth defects at TSPs to penetrate through wall, even -

though full unstable rupture is prevented by the TSP. The through wall defects could increase leak rates significantly, as compared to pre accident leak rates.

Similar potential increases in leak rates have been identified for roll transition defects (I), and the combined leakage from TSP and roll transition defects needs to be considered.

The approaches being taken regarding this concern are as follows:

. Sweden. The SSPB approach regarding leakage at TSPs during accidents is that defect depths at TSPs will be limited to values that assure that the number of defects is minimized that could become through wall during the operating cycle or could propagate through wall during an accident. As discussed earlier, this is the basis for the 705 TWD limit currently proposed for ATPC.

Beloium. Analyses are underway to justify increasing allowed primary to secondary leakage during postulated accidents to higher values, such as 20 gpm. This may require lowering allowed radioactivity levels in the reactor coolant, which is considered practical because of good fuel performance. The leak rate limit would apply to total primary to secondary leakage from all causes, including roll transition cracks and cracks at TSPs.

. France. Leakage during accidents caused by defects at TSPs or roll transitions has not been evaluated to date. Should it be necessary, EdF considers that this evaluation should be based on a probabilistic method such as that developed by Laborelec for the EPRI roll transition ATPC document (1).

. SAAin, The possible contribution of leaks at TSPs to accident leak rates, and whether this might require additional limits on defects at TSPs, has not yet been considered. The work being performed as part of the EPRI ATPC effort is being followed to determine needs in this area.

As can be seen by the variability in the above approaches, resolution of this concern is still in the formative stages, and a standardized approach has not yet been worked out.

Oraft 2/4/91 2-4

. )

Quali?v status of TSPs l i

\

in Sweden, a concern has been raised as to what needs to be done to allow

{

credit to be taken for TSPs considering that they were not designed and built as i Safety grade equipment. For example, documented inspection reports may not be available to show that welds relied upon for strength were made with correct l materials, wereo 'f the correct size, etc. Similarly, TSPs are not subjected to l

periodic inspections to verify that their structural integrity has remained 1 intact.

This concern had not been add.essed in any of the other countries visited.

Insnect on Criteria The NDE methods used to quartify the amount of degradation present at a TSP due to IGA / SCC were reviewed at each of the countries visited. Bobbin coil ECT I is used by all the groups. RPC is used for confirmation in some countries. The I groups visited indicated that they are supportive of the EPRI effort to develop an ATPC technical support document based on correlations of tube burst and l leakage behavior with ECT signals (amplitude and phase angle), without attempting to measure quantitative depths or lengths of defects. However, until these results are available. two of the countries, Sweden and Spain, will l

continue to pursue their quantitative defect depth limit (70% and 78% TWD) type  !

criteria, l

Need for Further Data It was noted during'the visits that part of the EPRI effort should include assembling information regarding the actual morphology of defects observed at TSPs, and the correlation of these defects with NDE results, especially bobbin coil ECT results. This information is needed to continue to confirm that the morphology of. cracks at TSPs is as expected, and to add to the data base I demonstrating that bobbin coil signals provide a reliable way to detect i unacceptable defects.

Swedish personnel indicated a desire for the EPRI work on ATPC for TSPs to include analyses for Model D steam generators, in addition to the work already underway f'or Model Sis.

Oraft 2/4/91 25

.3.~=,..;.=.---.-.--.- .- .._.-. e Section 3 REFERENCES 1.- T. G.-Harshbarger, " Preparing to Replace the Lower Assemblies at Cook,"

Nuclear Engineering International, January 1988 .pp. 26 27.

f

2. H. Takamatsu, " Japanese Steam Generator Operating Experiences and Results of Related R & 0," presented at EPRI Reliability Project Technical Advisory Committee meeting, June 26-28, 1990, Philadelphia, PA.

3. Minutes of Steam Generator Reliability Project Technical Advisory Group meeting, October 23-25, 1990. Palo Alto CA.

4 NRC Information Notice No. 90 49, " Stress Corrosion Cracking in PWR Steam Generator Tubes," USNRC, Washington, D. C., August 6,1990

5. WCAP-11056, "NRC Presentation Report on Steam Generator Tube Integrity for D. C. Cook Unit 2. January 1986". Westinghouse Electric Corporation, Pittsburgh, PA, in NRC Public Document Room ACN 8602130275, Date 860131.

POR ADOCK 05000316

6. WCAP 11330, "NRC Presentation Report on Steam Generator Tube Integrity for D. C. Cook Unit 2. September 1986", Westinghouse Electric Corporation.

Pittsburgh, PA, in NRC Public Document Room, ACN 8612020058, Date 861124, l

PDR A00CK 05000316

7. PWR steam Generator Tube-Pluacina Limits. Technical sunnert Document for Exoansion Zone PWscc in Roll Transitions. EPRI NP-6864 L, December 1990 11 l

1:

Draft 2/4/91 3-1 )

l

I

, Appendix A VISIT WITH SWEDISH STATE POWER BOARD Dates: August 29, 1990 ElArt: Swedish State Power Board (SSPB) offices, V8111ngby, Sweden Subiect- Alternate Tube Plugging Criteria for Defects at Tube Support Plates and Roll TranLitions (ATPC TSP and ATPC-RT)

Person contacted: Jan Engstr5m, SSPB EacL9ns Makino Visit: Jeff Gorman, DEI Al McIlree, EPRI Chuck Welty, EPRI

1.0 Purnose

The main purpose of this visit was to gather information relateo to development of alternate tube plugging criteria-(ATPC) and associated plugging limits for intergranular attack / stress corrosion cracking (IF2A/ SCC) at tube support plates (ATPC TSP). A second purpose of the visit was to review the current situation in Sweden regarding ATPC for primary water stress corrosion cracking (PWSCC) defects at roll transitions (ATPC RT).

2.0 Summary a Alternate tube plugging criteria for TSPs (ATPC TSP) were recently proposed to Swedish safety authorities. These criteria were that defects'up to 70% of through wall depth (TWD) were acceptable. The 70t value is based on ensuring that defects not grow through wall during ,

the next operating cycle, and an allowance of 10% for growth and 20%

for nondestructive examination (NDE) uncertainty. Safety authorities did not grant full approval of the criteria, but did accept them for a limited number of tubes for one year, t

. SSPB's main concern regarding ATPC-TSP is that TSP deflection might occur during a postulated accident such as a steam line break, thus removing the constraining effect of the TSP for defects in the TSP area. Deflections up to 13 mm have been calculated for the top TSP.

One approach being considered to resolve this concern is to limit Draft 1/31/91 A1

application of ATPC-TSP to the bo220m five supports there deflections are smaller. .

SSP 8's approach for resolving concerns regarding possible primary to secondary leakage during postulated accidents is to limit defect sizes at TSPs to values such that leakage during postulated accidents would be insignificant, even after allowance is made for defect growth and NDE error.

l SSPB intends to resubmit ATPC-TSP to Swedish safety authorities later this year, and hopes to have them approved in time for the mid 1991 inspections of Ringhals 3 and 4 SSPB supports the EPRI SGRP effort on ATPC TSP: however, they would like the current work to include Model 0 as well as Model 51. steam generators.

They noted that items such as TSP deflection are dependent l

on design details, which vary from model to model.

Swedish safety authorities *

^e not willing to accept a recent proposal to increase the length of I owed cracks at roll transitions based on revised estimates for crac. Jrowth rate and for NDE uncertainty.

i 3.0 Discussion 3.1 Occurrence of IGA / SCC Defects at Tube Suncort Plates

!GSCC at TSPs has been detected by eddy current tests (ECT), consisting of bobbin coil inspections and followup rotating pancake coil (RPC) inspections at l

Ringhals 3, but had not been detected as of August 1990 at Ringhals 4 Axial 1 IGSCC up to about 905 through wall depth (TWO) and 12 mm long has been confirm by examination of pulled tubes at Ringhals 3. At Ringhals 4, no cracks have been revealed by examination of pulled tubes but, as also seen at Ringhals 3 IGA up to about 25 microns depth is commonly observed at TSP locations. l The amount of IGA / SCC detected by NDE at Ringhals 3 has been limited, with about i 25 tubes (about 0.2%) having indications exceeding the current plugging limit of 501 TWD.

As discussed later, a few of these tubes have been left in operation to gain crack growth information.

i Oraft 1/31/91 A-2

3.2 Curront Tubo pluccina critoria for Defects at TSPs Plugging criteria for defects at TSPs are that defects shall not have indicated depths exceeding 50% TWD. These criteria are the same as for free span defects.

~

On a temporary basis, Swedish safety authorities have allowed a few tubes to remain in operation for one cycle with indicated defect depths up to 70% TWD.

The purpose of this exception is to allow crack growth rate data to be obtained for a few deeper defects. Defect depths are determined using bobbin coil ECT.

3.3 Procosed Alternate Tube Pluccino criteria for Defects at TSPs.

SSPB has proposed to their safety authorities that plugging criteria for defects at TSPs be changed from 50% TWO to 70% TWD. The bases for the 70% value include the following: (1) burst test results that show that through wall cracks that extend the full height of the TSP 19 mm, or a little more, do not result in tube rupture under accident conditions when supported by the TSP, and (2) a 10%

allowance for growth and a 20% allowance for NDE uncertainty, with an_ objective of assuring no significant through wall penetration occurs during the next operating cycle (this will minimize leakage during normal operation and postulated accidents).

3.4 Status of ATPc TSP Deve1ooment About November 1989. SSPB initiated work at Framatome directed at developing the bases for ATPC TSP. This work included burst tests of tubes with defects located at TSPs. Cold and hat burst tests were performed, including tests using tubes with IGSCC defects generated using caustic. This work has shown that the presence of the TSP prevents rupture of tubes at the TSP for through wall or

~

smaller defects located within the TSP. In addition, it was found that axial defects which extend less than about 10 to 11 mm beyond the edge of the TSP do not reduce tube burst strength.

Evaluation of the possible deflection of TSPs from hot full power positions was also performed. A steam line break (SLB) at hot shutdown conditions was determined'to be the limiting case (causing the most TSP deflection), since initial secondary temperature and stor?d energy are greatest, and since thermal displacement of the TSPs is also the greatest.

Based on burst tests and related analyses, it was proposed to Swedish safety authorities that plugging criteria for defects at TSPs be increased from 50% to 70% TWD. With a value of 70%, which includes an allowance of 10% for growth and 20% for NDE uncertainty, it is expected that the possibility of tube rupture is Draft 1/28/91 A-3

__ _ ____ _ . ._ __ ___ _ __ . . _ _ __.___._ _.~

essentially eliminated-and that leakage due to defects at TSPs would be insignificant following postulated accidents. Thus, this approach avoids the need for quantifying and demonstrating the acceptability of leakage during accidents.due to this cause.

Swedish safety authorities did not accept the proposed change except for a small.

group of tubes which were allowed to have defects up to 70% TWD to obtain growth rate data. The main reason that the criteria were not approved for general use in 1990 was that Swedish safety authorities wanted time to determine what was being done in other countries. In addition, questions were raised with regard to the difficulty in assuring that the TSPs would be in position surrounding the tube defects during and after postulated accidents such as the

.SLB or feedwater line break (FLB), and with regard to quality assurance of the support structure.

1 Subsequent to making the ATPC-TSP proposal to Swedish safety authorities.

Framatome has performed extensive additional evaluations of the potential for TSP deflection. SSPB is currently evaluating the Framatome analyses, and is evaluating what approach should be adopted for handling this issue. SSPB expects to re-apply.for approval of ATPC TSP, with further consideration given ,

to TSP deflection, by early 1991, with the intent of having ATPC TSP in effect for mid 1991 inspections.

3.5 TSP Deflection considerations Analyses performed by Framatome for SSPB have shown that, for the Ringhals 3 and 4 steam generators, the' differential pressure (AP) that can develop across TSPs during postulated accidents varies from low values (about 1 psi) at low TSP elevat:Sns to about 5.5 psi at TSPs at the the top of the tube bundle. The controlling case is for a SL8 or FLB from the hot shutdown condition, where starting temperature and stored energy on the secondary side are at a maximum.

These calculated values of differential pressure (up to 5.5 psi) are large enough to cause deflections of upper TSPs up to 13 mm, if it is assumed that no resistance to deflection is provided by tube-TSP' interaction. If it is also

_ assumed that the TSP could stick in the deflected position, then the TSP deflection could effect tube rupture behavior, and/or have a significant effect on leak behavior..

i One approach for limiting concerns regarding TSP deflection would be to limit application.of ATPC-TSP to lower TSPs, e.g., TSPs 1 through 5, where support Draft 1/28/91 A4 ,

.rw

from the preheater structure is the greatest and predicted defloc2 ions are relatively small.

SSPB/Framatome analyses of TSP deflection have shown that the amount of I deflection is design dependant and is a function of features such as tie rod design (location and diameter), type of welds at wedge blocks used to connect the TSPs to the wrapper (e.g., whether welds are present at both the top and bottom of the joint, or just at the top). flow patterns, presence of pre heater, l etc. Further, in evaluating the question of TSP deflection, it was found difficult to justify assuming that TSPs could not move freely past tubes and quantitative credit for TSP to tube friction was not taken. It was also difficult to develop assurance that the TSP would return to its original position, since slip stick behavior appears to be possible. Thus TSP deflection evaluations are based on assumptions of zero tube to TSP sliding resistance force, together with possible sticking of the TSP in the fully deflected position.

An additional factor of concern regarding TSP deflections and reliance on TSP structural integrity is that the TSP structure was not originally a nuclear safety-(0 list) component, and the amount of OC/0A verification performed during fabrication to assure that the structure was manufactured and installed per design (correct materials used.-welds installed and inspected, etc.) was l

probably limited. I 3.6 Accident Leakace Limits l The SSPB approach regarding leakage at TSPs during accidents is that defect depths at TSPs will be limited to values that assure that the number of defects is minimized that could become through wall during the operating cycle or could propagate through wall during an acc'ident. As discussed earlier, this is the basis for the 70% TWO limit currently proposed for ATPC TSP.

l Burst tests in cold conditions have shown that the presence of the TSP does not affect (i.e., does not prevent) the rupture of the remaining ligament of part depth defects. The test results indicate that cracks of the typical maximum length observed in TSPs of about 12 to 13 mm, and with depths over about 82%

TWD are likely to propagate through wall during postulated accidents, and to result in some primary to secondary leakage after the accident. Larger cracks l

J could propagate at lower pressures. SSPB intends that their proposed 70% TWD i.

Oraft 1/28/91 A5

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

I plugging criteria limit the number of defects that could grow through wall.

~

before or during an accident to an insignificant number. *

.3.7 EPRI ATPC-TSP Technical Suncort SSPB indicated that they are.suppor.tive of the EPRI effort to develop an ATPC-TSP technical support document based on correlations of tube burst and leakage behavior with ECT signals. However, until these results are available, they will continue to pursue the 70% TWO limit discussed above.

The EPRI ATPC-TSP effort is currently focused on Model 51 steam generators.

SSPB would like the EPRI effort to also develop information for Model D steam generators of the type used in Sweden, in parallel with work on Model 51s. SSPB noted that some important parts of the bases for ATPC TSP are dependent on steam generator design details, such as tie rod location and diameter, and that these details can differ from design to design. ,

It was agreed that.part of the EPRI effort should include assembling information regarding the actual morphology of defects observed at TSPs and the correlation of these defects with NDE results, especially bobbin coil ECT results. This is because:

Continuing support will be required for the assumptions that the TSP defects are mainly axial cracks, and do not involve significant general degradation, such as !GA.

[ . Support is needed to demonstrate that use of bobbin coil data provides l reliable detection of defects that are structurally significant (likely l

to burst at or below accident pressures), and also provides reliable detection of defects that increase the probability of developing significant leakage during postulated accidents.

4.0 Discussion of ATPC RT Swedish plugging criteria for roll transition PWSCC cracks currently limit the lengths of cracks extending up from the top of the tube sheet to 7.5 mm (there are additional criteria covering cracks that have both ends above the top of the

( tube sheet: 5 mm if close to the tube sheet, and 3 mm if the top is 16 mm or l more above the tube sheet). Based on correlations of RPC measured defect i

lengths with lengths determined by metallurgical examination of pulled tubes.

SSPB recently proposed reducing the allowance for NDE uncertainty from 3 to 1

' Draft 1/28/91 A6

.. ._ ._ . I 9- l l

mm o and also' proposed that a more realistic basis for setting the allowance for crack growth, to that expected for a 4 mm long crack, be used. The proposed changes would support changing the criteria such that cracks a few mm longer than the current 7.5 mm limit would not require repair. However. Swedish safety authorities took the position that sufficient experience is not available to justify such changes. For example, one expressed concern is that link up of short cracks could result in increased apparent growth rates. SSPB intends to continue pursuing obtaining relaxation of the ATPC RT.

Jeffrey A. Gorman l Dominion Engineering. Inc, l

l l

l l

l l

l l

l l

1 1

Draft 1/28/91 A-7

Appendix B VISIT WITH BELGATOM Recort of Visit DAla: September 3, 1990

{

E1111: Laborelec offices Linkebeek, Belgium Subiect: Alternate Tube Plugging Criteria for Defects at Tube Support Plates and Roll Transitions (ATPC TSP and ATPC-RT)

Person contacted: Paul Hernalsteen Laborelec Persens Makino Visit: Jeff Gorman, DEI Al mci 1ree EPRI Chuck Welty, EPRI

1.0 Purnose

The main purpose of this visit was to gather information related to development of alternate tube plugging criteria (ATPC) and associated plugging limits for intergranular attack / stress corrosion cracking (IGA / SCC) at tube support plates.(ATPC-TSP). A second purpose of the visit was to review the current situation in Belgium regarding ATPC for primary water stress corrosion cracking (PWSCC) defects at roll transitions (ATPC RT). ,

2.0 summarv The occurrence of IGA / SCC at TSPs has been relatively minor in Belgium, with only two units affected (Tihange 1 and Doel 4), and with only a limited number of tubes affected at these units.

Current plugging criteria in Belgium do not require defects located j within the TSP edges to be plugged. I Analyses are underway to justify increasing allowed primary to secondary leakage during postulated accidents to higher values, such as 20 gpm. This may require lowering allowed radioactivity levels in the reactor coolant, which is considered practical because of good fuel performance. The leak rate limit would apply to total primary to Draft 1/28/91 B-1 l

h. '

secondary leakage from all causes, including roll Bransition cracks and V .

cracks at TSPs.

~

The EPRI-SGRP effort for ATPC TSP is being relied upon to develop guidance relative to questions regarding TSP deflection during accidents, and correlation of bobbin coil signals and the leakage perfoamance of tubes at TSPs.

+

No significant circumferential cracks have been found at roll transitions in Belgian plants, despite extensive inspections using sensitive methods. Plugging criteria have been' approved that allow circumferential cracks of limited sizes to remain in place. However, in practice, tubes with circumferential cracks will probably be plugged when circumferential cracks are combined with multiple axial cracks.

3.0 Discussion 3.1 Occurrence of IGA / Scc Defects at Tube Sunoort plates

-IGA / SCC defects at.TSPs have been detected by bobbin coil inspections and by examination of pulled tubes at Tihange 1. and by leakage and subsequent bobbin coil inspections and follow-up ultrasonic inspections at Doel 4 IGA / SCC at TSPs.has not been detected by bobbin coil inspections or by leakage at Doel 1.

2. or 3. or at Tihange 2 or 3. The IGA / SCC at TSPs in Tihange 1 has been increasing slowly for several years, and about 69 tubes are now affected.

Examination of a pulled tube indicates that the IGA / SCC consists of axial intergranular and transgranular cracks that are fully within the TSP edges. The Doel-4 IGA / SCC has only recently been detected, in about 5 or 6 tubes.

Ultrasonic inspections and leak tests indicated that the largest defects consist of through wall axial cracks 11 to 12 mm long.

I 3.2 Current Tube Pluacino Criteria for Defects at TSPs a

Belgian tube plugging criteria permit operation with axial through wall cracks with lengths completely within the edges of the TSP. i.e.. do not r n.utre tubes i

-with such cracks to be plugged. These criteria are based on results of burst g tests which show that such axial cracks do not reduce the burst pressure: 1.e..

the tube bursts in the free span area, away from the TSP.

i i

i Drafi i128/91 B2 J

_ _ . _ _ ~ . . _ __ . _ . . . _ _ _ . . _ _ _ _ _ _ _ . . _ _ . _ . _ _ . . _

a 3.3 Status of Tube Suonor2 Plate Alternate Tube pluacina Criteria (ATPC TSP)

[ Develonment and Review

• As noted above. ATPC-TSP already exist which permit through wall axial cracks contained within the TSP to remain without repair. However, the possible effects on leakage during a postulated accident such as a steam line break (SLB) or a feedwater line break (FLB) are recognized and are being addressed by Belgatom, This leakage could add to the leakage from roll transition cracks, and may lead to the need to limit the number and/or depth of cracks at TSPs.

Belgatom is not doing any separate research regarding quantification of possible leakage at TSPs. but rather'is participating in the EPRI-SGRP ATPC-TSP effort, which is being relied upon to develop needed technical support for the alternate criteria.

3.4 TSP Deflection considerations Belgatom has no separate work underway at this time regarding evaluation of possible TSP deflection during postulated accidents and, as above, is participating in the EPRI ATPC-TSP effort, which is being relied upon to develop i

needed informction.

3.5 EPRT ATPC-TSP Technical suonort Belgatom is cooperating with EPRI in the development of an ATPC-TSP technical support document and, as noted above, is relying on the EPRI project to develop any needed information regarding TSP deflection and leakage due to defects at TSPs. The Belgian perspective with regard to one of the main thrusts of the EPRI effort, correlation of ECT signals with structural capability, was discussed. They indicated:

They agree with the general concept of correlating bobbin coil ECT signals at TSPs to leakage' behavior, without attempting to measure quantitative depths.or lengths of defects, subject to the following reservations.

They are not confident that meaningful information can be obtained from bobbin coil phase angle data. In addition, they noted that there are large effects of probe location (wobble) on the amplitude of bobbin coil signals generated by a defect at one circumferential location.

For example, the variation in amplitude of the signal generated by an ASME defect as the probe is moved from one side to the other of a tube is 68% to 1975, where'r 100% signal is developed by the centered probe. I Draft 1/28/91 B-3 1

Despite these difficulties, they concur that no better approach is presently available, since bobbin coil inspections are considered to bt the only practical way of monitoring tube conditions at TSPs.

1 3.6 circumferential cracks at Roll Transitions No significant circumferential cracks have been detected at rcia transitions in Belgian plants. One very short (0.5 mm) crack was seen in a pulled tube from Doel 2 in 1983, and several short segments (maximum length of 3.5 mm) were seen in a pulled tube from Doel 3 in 1986. The absence of significant circumferential cracking has been confirmed by RPC examinations of all tubes in Doel 3 and Tihange 2 for several years, and by examination using sensitive A

ultrasonic test (UT) methods of 250 tubes in the sludge pile region of one steam generator in Doel 3 for two years. The UT method is considered capable of detecting circumferential through wall cracks of 2 to 3 mm length. The RPC method is considered capable of detecting circumferential through wall cracks of 6 to 8 mm length if not combined with axial cracks at the same location.

Burst tests have been performed that indicate that circumferential cracks 6 mm or shorter which intersect axial cracks do not significantly affect tube burst behavior, i.e., do not excessively reduce burst pressures below the values set by the axial cracks.

ATPC for circumferential cracks have been developed, and have been accepted by Belgian safety authorities (16 mm limit for 7/8* tubes and 14 mm limit for 3/4* j tubes). If circumferential cracks are detected. then inspections with increased l

sensitivity will probably be performed, e.g.,100% inspection using an axially oriented UT signal. In addition, if necessary based on the types of crack patterns observed, plugging criteria would be re-evaluated and revised as appropriate.

3.7 Accident Leakaos Limits Belgatom is in the process of re evaluating allowable leakage during postulated accidents such as a steam line break (SLB) or a feedwater line break (FLB).

Site dose rate limits are more stringent in Belgium than in the USA, which tends to limit the allowable leakage. However, fuel performance has been very good, and reactor coolant activity limits may be reduced to take advantage of this good experience. It is anticipated that total accident leak rates of 20 gpm Draft 1/28/91: B-4

. . .. . m..~, - , . - _ . _ - . _ . . . _ . - . . _ . . . . . _ . _ _ . _ . . - .

will be shown to-be permissible. These leak rates cill apply to the sum of i leakage from roll transition and TSP defects.

Jeffrey A. Gorman Dominion Engineering, Inc.

)

I e

t 4

I 8

l.

Draft'1/28/91 85

, 1 b

Appendix C VISIT WITH ELECTICITE DE FRANCE AND FRAMATOME Recort of Visit Dates: September 4 and 5, 1990 l (

Places:

Electricite de France offices in St. Denis anr la Cefense and Framatome offices in La Defense Subiect: Alternate Tube Plugging Criteria for Defects at Tube Support Plates and Roll Transitions (ATPC-TSP and ATPC RT)

Persons contacted: Libin Bernard, Framatome Jean Paul Billoue, Framatome .

l Robert Comby, EdF John Esposito Westinghouse Breno Flesch, EdF i

Angel Gelpi, Framatome Bob Gold, Westinghouse Don Harrod, Westinghouse Pierre Nicot, EdF Francis Nordmann, EdF

' Jean Marie Pecout EdF l Georges Slama, Framatome Persons Makina visit: Jeff Gorman, DEI Al McIlree, EPRI Chuck Welty. EPRI l 1.0 Purnose: The ma'in purpose of this visit was to gather information related to development of alternate tube plugging criteria (ATPC) and associated plugging limits for intergranular attack / stress corrosion cracki.;g (IGA / SCC) at tube support plates-(ATPC-TSP). A second purpose of the visit was to review the current situation in France regarding ATPC for primary water stress corrosion

! cracking (PWSCC) defects at roll transitions (ATPC RT).  !

Oraft 1/31/91 C1

d 2.0 Summarv

. The number of units experiencing IGA / SCC at TSPs, and the numbers of tubes affected in these units, are gradually increasing as plants age.

About 10 units are now affected, and the number of tubes affected increases by about a factor of 2 to 2-1/2 per fuel cycle in those plants affected to the greatest degree. However, there is no requirement to plug or repair indications at TSPs, and the sizes of the eddy current test (ECT) indications seen in plants are much less than the sizes that laboratory tests indicate could significantly affect tube burst strength. Accordingly. IGA / SCC at TSPs is not considered a serious problem in France.

. TSP deflection during accidents has been evaluated and determined to not be a problem for French steam generators.

Leakage during accidents caused by defects at TSPs.or roll transitions has not been evaluated to date. Should it be necessary. EdF considers that this evaluation should be based on a probabilistic method such as that developed by Laborelec for the EPRI ATPC-RT guidelines.

EdF has recently adopted a length based criterion (13 mm limit) for I PWSCC at roll transitions. as opposed to the previous leak before risk of break (LBRB) criterion. EdF plans to perform rotating pancake coil (RPC) inspections of 100% of the tubes in affected steam generators, with initial inspections of all affected steam generators (over 5%

tubes affected) being completed over a two year period.

3.0 Discussion 1 3.1 Occurrence of IGA / SCC Defects at Tube Sunnort Plates IGA / SCC defects at TSPs were first discovered in French plants at Fessenheim 1 in 1986. Since'that time. TSP defects have been identified in about 10 other units. Because these defects are not considered to be significant with regard to safety, inspections to identify and trend such defects have generally been limited to the relatively small inspection samples used to monitor the general condition of the steam generator and to monitor special areas, such as row 1 and 2 U bends and areas subject to antivibration bar (AVB) wear. Also, the threshold used to determine whether signals at TSPs should be reported has been subject to some change with time. As a result, limited information is available Draft 1/31/91 C2

I l

on the occurrence and progression / growth of defects at TSPs. Nevertheless, some general observations can be made:

. The number of plants affected has been increasing slowly but steadily

'as plants age. About'10 plants are now listed as having experienced

.some degree of detectable secondary side attack at hot leg TSPs.  ;

l

• At plants with measurable / reportable amount of attack', the number of '

TSP intersections with ECT signals over a given threshold voltage'tends to increase by a factor of about 2 to 2 1/2 per annual cycle.

l

! 3.2 Results of Pulled Tube Examinations l

Three tubes with IGA / SCC defects at TSPs were removed from Fessenheim 1 in 1986.

one from St. Laurent B1 in 1987, and one from Bugey 5 in 1989. Examination of these tubes showed the defects to be primarily axial IGSCC. All of the defects were confined within the borders of the TSPs.

l 3.3 Results of Burst Tests An extensive series of burst tests was performed subsequent to the discovery of l IGA / SCC at'TSPs in Fessenheim 1 in 1986. These tests were performed using specimens with IGA generated by sodium tetrathionate attack of sensitized tubing. The tests showed that even very severe degradation of the tube (about 60Z of tube wall severely attacked for its full circumference) does not cause i

reduction of the tube burst strength, as long as the degradation is fully-I confined within the borders of the TSP.

l 3.4 Pluaaina criteria for TSPs Based on the' results of the burst tests described above, it has been concluded that the type of IGA / SCC degradation observed in French plants does not impact tube burst properties and is therefore not a safety concern. Accordingly', there l

are no requirements for plugging tubes as a result of detection of IGA / SCC defects at TSPs. These criteria are not documented in a formal report. Rather.

they are based on an agreement with French safety authorities and, in the event that significantly different types of defects are discovered, the criteria will be re evaluated.- However, it was noted that the voltage signals from the severely. degraded test tubes, that did not experience reduced burst pressure as a result of the degradation, weri about a factor of ten greater than the largest 4

Thus, there is a p

signals detected in plants (e.g.,18 volts versus 1.8 volts).

l l

'Oraft1/31/9y C3

large margin between the observed conditions and those that would require reevaluation of the need for plugging criteria for IGA / SCC defects at TSPs.

The possible effects of IGA / SCC defects at TSPs on leak rate following postulated accidents such as a steam line break (SLB) or a feedwater line break (FLB). were discussed. EPRI indicated that the need to limit the leak rate due to its impact on site boundary doses was considered in the USA as a probable reason for limiting the number and size of defects at TSPs. as well as at roll transitions. EdF indicated that there are no plugging criteria in existence in France as a result of this particular effect; however, the subject is currently being evaluated by EdF. In this regard. EdF noted that:

. Previous evaluations of acceptable primary to secondary side leakage during postulated accidents associated with defects at antivibration bars (AVBs) indicated that maximum leak rates in the neighbor, hood of 15 cubic meters per hour (66 gpm) are acceptable, and that similar limits are expected to apply to the sum of leakage from all defects, including TSP and roll transition defects. It was noted that the leak rate rate limits for defects at AVBs were established by core flow considerations, rather than by site boundary dose considerations, which generally control accident leak rate limits in the USA.

. The presence of the TSP is not expected to limit the growth through wall during a postulated accident of defects that are nearly through wall prior to the accident. Such growth is the result of the high differential pressure expected during postulated accidents, and because the support provided by the TSP does not become effective until the tube experiences significant plastic deformation. Thus, the occurrence of a postulated accident could result in an increase in the number of through wall defects and an increase in the primary to secondary leak rate. EdF considers that such possibilities of crack growth and increased leakage should be addressed using statistical models for crack growth and leakage behavior, similar to that developed by Laborelec for roll transition cracks.

3.5 ISP Deflection Evaluations of possible TSP deflections during postulated accidents have been performed. Detailed results were not available during the meeting, but maximum deflections of about 2 mm were determined for the highest TSP for French Model Draft 1/28/91 C-4

_ - . - - . _ . ~ - - . _ _ - . -

51 steam generators: these values reflect the specific design features of the j ,

French units, e.g.. number and size of tie rods. These deflections are not

} considered to be significant.

s

3.6 EPRI ATPC TSP Technical Suncort Effort EPRI indicated that the EPRI technical support document is directed at l establishing a correlation between bobbin coil signals at TSPs and tube

structural and leakage performance. EdF described some of their experiences l

with correlation of bobbin coil ECT signals with defects at TSPs.

l-As described above regarding burst tests. EdF has found that tubes with j

e bobbin coil voltage signals about ten times as large as the largest  ;

i '

signal seen in plants have not significantly reduced burst strengths.

EdF provided EPRI with their data correlating TSP signals a~no defect size at TSPs based on examination of pulled t'ubes. Correlation of defect depth with signal amplitude has not been good, but correlation with phase angle has been better. However, the number of pulled tubes has been limited, and EdF considers that it is not possible to draw firm conclusions regarding the ability of ECT to monitor defect depths.

3.7 Discussion of ATPC-RT l

EdF has recently adopted a length based plugging limit of 13 mm for I axial cracks for all tubes. Previously, this limit applied only to the sludge pile region, and leak before risk of break (LBRB) criteria were unid for non sludge pile regions. EdF considers'LBRB to still be valid for non sludge pile regions but has adopted the inspection based approach for regulatory reasons.

i EdF has started performing RPC inspections of the roll transition region of 1005 of the tubes in affected steam generators, i.e., steam

. generators with over 5% of the tubes affected by PWSCC at roll transitions. This is a change from the previous practice of performing 100% RPC inspections of sludge pile areas and limiting RPC inspections l of tubes in other areas to tubes with helium leaks or unusual bobbin coil signals. The 1005 inspections of all affected plants will be completed during a two year time frame.

Oraft 1/31/91 C-5

. - _ _ . _ - _ - _ . . .. _ ~ _ . . . - - ~_ _ _ - - _ - - - - . - - -

i

. A leak rate limit of 5 L/h per steam generator continues to be used for

• 900 MWe plants that are significantly affected by roll transition PWSCC. despite the adoption of a length based limit and 100% RPC inspection. A leak rate limit of 3 L/h is being used for 1300 MWe plants. These leak rate limits have not been found to be unduly burdensome. N-16 monitors on steam lines are used to provide rapid on-line indic'ation of primary to secondary leak rates.

. About 70 tubes in the sludge pile region are being preventively plugged in steam generators in which one or more circumferential cracks in the sludge pile region have been detected (about 13 steam generators in 5 units).

. The growth rate of PWSCC cracks in many roll transitions in France appears to De very low or zero. However, recent measurements in one steam generator of Dampierre 3 (fourth cycle after peening) indicate that average crack growth rates of about 1 mm per fuel cycle occurred, similar to values measured in Belgium. Growth rates decreased as crack l 1ength increased, with essentially zero crack growth for the longest cracks. The decision to use 100% RPC inspections of roll transitions is expected to provide additional useful data on this subject in the

( near future.

l

- EdF's position regarding circumferential cracks at roll transitions was discussed, and includes:

EdF plugs all tubes in which circumferential cracks are detected.

EdF considers that the inspection sensitivity of RPC is such that any significant circumferential crack (i.e., a crack that would have a marked impact on burst strength) will be detected by the 100% RPC examination that has recently been instituted.

Draft 1/28/91 C6 1

=

The situation regarding effects of roll transition PWSCC on leakage during postulated accidents (e.g., SLB and FLB) is the same as l

previously discussed for TSP defects, i.e., EdF has not had to meet specific requirements up to the present, but is currently evaluating this concern.  ;

Jeffrey A. Gorman Dominion Engineering. Inc.

l l

l l

h l-1 V

4 1.

! Draft'1/28/91 C-7 l

i

Q Appendix-D ,

VISIT WITH SPANISH UTILITIES Reoort of Visit Dates: September 18 and 19, 1990 Places: Offices of Centro Investigaciones Energeticas, Medioambientales Y Tecnologicas (CIEMAT) and Central Nuclear de Almaraz (CNA). Madrid, Spain Subiect: Alternate Tube Plugging Criteria for Defects at Tube Support Plates and Roll Transitions  !

Persons contacted: Gustavo Bollini, Tecnatom-PWR Project Dolores Gomez Briceno, CIEMAT Julian Gorosarri, CNA Anna Marie Lancha Hernandez. CIEMAT Emilio Lopez Toribio Asco Jauquin Lozano, Asco Jose Pelaez Martinez, CIEMAT

-Jose Maria Zamarron. CNA Persons Makina Visit: Jeff Gorman, DEI Al McIlree, EPRI Chuck Welty, EPRI

1.0 Puroose

The main purpose of this visit was to gather information related to development of alternate tube plugging criteria (ATPC) and associated plugging limits for intergranular attack / stress corrosion cracking (IGA / SCC) at

. tube support plates-(ATPC TSP). A second purpose of the visit was to review the current situation in Spain regarding ATPC for primary water stress corrosion cracking (PWSCC) defects at roll transitions (ATPC RT).

2.0 Summarv IGA / SCC defects are occurring in increasing rates at Almaraz 1 and 2 and at Asco 1 and 2. The defects are mostly axial in orientation. can be up to 1005 in through will depth (TWD), and generally do not extend beyond the edge of the TSP. In one case, defects at the flow Draft 1/28/91 0-1

=

distribution baffle (FDB) did extend beyond the edge of the F08:

however, this was associated with sludge being present on top of the F08 and to sodium from initial operation of polishers, and is not '

considered applicable to IGA / SCC at TSPs. In another case, bobbin coil i

inspection indicated that a primary side defect at a dented TSP -

extended beyond the edge of the TSP: however, this defect was not confirmed by RPC.

l Current tube plugging criteria for defects at TSPs are the same as for l free span defects, 40% TWD. Proposed ATPC TSP have been submitted to, but not yet accepted by. Spanish safety authorities. The proposed criteria would allow defects up to 78% TWO to not be repaired. The main bases for the criteria are that burst tests have shown 100% TWO I defects that extend a short distance beyond the edges of the TSD to not reduce burst strength below accident pressures. For operational reliability reasons, it is desired to avoid development of leaks at TSPs. For this reason, an allowance of 7% for defect growth and 15%

for defect depth measurement error are subtracted from 100% to arrive at the proposed limit of 78% TWD.

The EPRI-SGRP ATPC-TSP effort is being relied upon to identify any i needed actions on'ATPC TSP regarding effects of TSP deflection and '

possible primary to secondary side leakage during accidents.

l The Spanish utilities are supportive of the EPRI-SGRP ATPC TSP approach of correlating bobbin coil signals with tube burst and leakage behavior, without quantifying defect depth or length. However, until the EPRI criteria are available, the Spanish will continue to pursue the 785 TWO limit discussed above.

During the summer 1990 inspection at Asco 1, 41 tubes were found to have 10 circumferential cracks at the roll transition. In earlier tube pulls at Asco and Almaraz, 00 circumferential cracks have been found.

The occurrence of circumferential cracks has required re evaluation of the ATPC-RT for Almaraz 1 and 2 (they have never been accepted for Asco 1 and 2). Based on burst tests which show that circumferential cracks to not degrade burst test performance if they are located below the top of the tube sheet (TTS), and based on experience which indicates that circumferential cracks occur at roll transitions, revised criteria have Draft 1/28/91 02

o I been proposed to Spanish safety au%horities that allow current ATPC-RT to be applied to tubes with roll transitions located below the TTS. but require plugging of tubes with roll transitions located above the TTS if any axial or circumferential defects are detected above the TTS.

3.0 Discussion 3.1 occurrence of IGA / SCC Defects at TSPs

!GA/ SCC defects at TSPs and flow distribution baffles (FOBS) have been detected ,

for several years by bobbin coil inspections in Almaraz 1 and 2 and Asco 1 and

2. The defects have been confirmed by examination of pulled tubes from all four units. They have also been confirmed in many cases using rotating pancake coil (RPC) methods. The number of IGA / SCC defects has been increasing at a significant rate at all four units. The bobbin coil indications at TSPs have been growing in amplitude by about 0.5 volts per fuel cycle, and in indicated ,

depth (by phase angle) of about 7 to 9% TWD per fuel cycle. -

Examination of pulled tubes has shown the defects to mostly be axially oriented

-IGSCC. Defect depths at TSPs have ranged up to 1005 TWD. Most of the IGA / SCC has been confined within the TSP.

In a few cases defects have been noted extending beyond the edges of tube supports and flow distribution baffles, or the cracks have had an off-axial component. For example:

. A large leak occurred at Almaraz 1 in July 1988 and was found to be

^

caused by IGSCC above the FDB. The IGSCC appeared to be associated with the presence of sludge deposited on top of the FDB and was caused by sodium input during initial operation of polishers; thus, this IGSCC is not considered to be rep'resentative of IGA / SCC at TSPs. Also, as noted later. ATPC TSP in Spain are not considered applicable to the FDB area, so that the occurrence of these defects above the FDB is not relevant to the question of ATPC-TSP.

. Bobbin coil examination in 1990 at Asco 1 identified a possible primary .

side crack extending beyond the edge of the TSP (the defect was not confirmed by RPC). The tube was dented at the TSP. Since this cracking occurred on the primary side. it is not directly applicable to ATPC-TSP for secondary side IGA / SCC. However, it illustrates the need Draft 1/28/91 D-3

_ . ._ ___ _ _ _ - ~ - _ _ - _. _

v l

' for ATPC TSP to consider the possibility of primary side attack at TSPs if dents are present.

. A pulled tube from Asco was found to have IGA / SCC at a TSP with cracks .

l running at a small angle to the tube axis. No denting was detected, and the reason for the off-axial orientation is not known.

3.2 Current Tube Pluacina criteria for Defects at TSPs Current plugging criteria for defects at TSPs are that defects shall not have indicated depths exceeding 40% TWD. and are the same limits as for free span defects.

3.3 Prooosed ATPC-TSP The Spanish utilities have recently proposed to their safety authorities that l The l plugging limits for defects at TSPs be changed from 40% TWO to 78% TWD.

bases for the 78% value are:  !

A 1005 TWD axial !GSCC defect at a TSP is not considered to be a safety concern from a tube rupture standpoint. This is based on burst tests that have shown that defects that are 1005 TWD and extend the full width (thickness) of the TSP, and up to three mm beyond the edge of the TSP. do not reduce tube burst pressure to values close to accident values.

. For operational reliability reasons, it is desired to avoid significant leakage at TSPs during the operating cycle following a given inspection. Accordingly, an allowance of 7% for defect growth and an allowance of 15% for NDE uncertainty are subtracted from the 100% TWO The value that is considered acceptable from a safety point of view.

7% growth value is an average one for IGA / SCC at TSPs, and the -15% NDE uncertainty comes from comparison between ECT data and pulled tube results.

l 3.4 Status of ATPc TSP Develonment A significant amount of work has been performed to develop and support ATPC TSP This work has included development of specimens with OD IGSCC for use in leak tests, performance of an array of burst tests using specimen; with of leah tests, and electrodischarge machined (EDM) notches performance supporting analyses. This work indicated that:

l l

! Draft 1/28/91 04

b

. l

. Clearances in FDB holes are so large that the FOB does not provide i effective support against rupture. Thus. ATPC TSP cannot be applied to

~

defects at the F08.

. Tube burst pressures with through wall defects that are fully contained within the TSP are much higher than the burst pressures for the same defect located in free span areas, and are much higher than accident pressures.

l

. Extension of defects 3 mm beyond the edges of the TSP only causes a 10 to 201 reduction in burst pressure, and the resulting burst pressure is still much higher than accident pressures.

The proposed ATPC-TSP described in 3.3 were recently submitted to Spanish safety authorities. A rapid response is not expected, because of attention currently being given to resolution of questions regarding roll transition alternate tube plugging criteria (ATPC RT) raised by detection of circumferential cracks at Asco.

Effects of the possible deflection of TSPs from hot full power positions caused by differential pressures across TSPs during accidents have not as yet been considered. The work being performed as part of the EPRI ATPC-TSP technical support effort is being relied upon to determine needs in this aaea.

The possible contribution of leaks at TSPs to accident leak rates, and whether this might require additional limits on defects at TSPs. has not yet been considered. The work being performed as part of the EPRI ATPC-TSP effort is being followed to determine needs in this area.

3.7 EPRI ATPC-TSP Technical Sunnort Effort

-The Spanish utilities indicated that they are supportive of the EPRI effort to develop ATPC-TSP based on correlations of tube burst and leakage behavior with ECT signals. However, until these criteria are available, they will continue to pursue the 78% TWD limit discussed above.

4.0 Discussion of ATPC RT Spanish plugging criteria for roll transition PWSCC cracks currently limit the lennths of axial cracks extending up from the last point of contact between the Draft 1/31/91 05

1 w.

l 1

tube and tube sheet to about 7 mm. These criteria are based on the assumed absence of circumferential cracks. Recent ECT inspections and pulled tube results have revealed the presence of significant numbers of primary side initiated circumferential cracks at Asco 1. In addition, a number of pulled tubes (6 of 12 pulled tubes) from Asco and Almaraz over the past three years have revealed.the presence of OD circumferential cracks. Because of the possible presence of circumferential cracks, revised ATPC-RT were recently prepared and submitted to Spanish safety authorities for use in the refueling outage just started at Almaraz 1. In summary. these criteria are as follows:

A 1005 bobbin coil inspection and a 25% RPC inspection (plus RPC of any tubes with bobbin coil indications) will be performed of the hot leg  ;

roll transition area of each steam generator. If no circumferential cracks are detected, then the normal ATPC RT will be used (axial crack length limited to about 7 mm).

. Steam generators in which circumferential cracks are detected will be  !

inspected 1005 by RPC. All tubes with defects above the top of the tube sheet will be identified. For these tubes Tubes with axial cracks only (no circumferential cracks detected) and with the roll transition below the top of the tube sheet will be treated using the normal ATPC RT.

Tubes with the roll transition located above the top of the tube sheet and with either longitudinal or circumferential defects located above the top of the tube sheet will be plugged. .

l Some of the considerations that went into the above criteria are as follows:

i

. Analyses based on experience at Spanish and foreign plants indicate th'at isolated circumferential cracks do not grow sufficiently in one fuel cycle to grow from an undetected size (e.g., 7.5 mm long) to an unacceptable size (critical length of 45 mm).

. Recently performed burst tests show that circumferential cracks below the top of the tube sheet do not degrade the rupture behavior of critical length axial cracks located above the top of the tube sheet.

4 Oraft 1/31/91 06

w Recently performed burst tests show that circumferential cracks located

• above the top of the tube sheet can significantly reduce the rupture j strength of tubes with intersecting axial cracks, and can lead to large leakage configurations, before rupture, as a result of flaps or corners bending where axial and circumferential cracks intersect. Because the presence of roll transitions located above the top of the tube sheet increases risks of undetected circumferential cracks above the the top of the tube sheet, tubes with roll transitions above the top of the tube sheet are subject to more strict criteria than tubes with roll transitions below.the top of the tube sneet.

Jeffrey A. Gorman Dominion engineering, Inc.

I 1

l l

i l

Draft 1/28/91 0-7 ,

4 Specifically, the following issues have been evaluated:

4 TSP degradation related to tube denting.

j - l TSP degradation related to chemical cleaning.

TSP degradation in the area of anti-rotation devices.

i -

Distorted TSP signals related to TSP Geometry.

]

Tube Bundle Wrapper Dislocation.

TSP Degradation Related to Tube Denting:

Comed is aware of TSP cracking at a number of units associated with tube denting due to TSP corrosion at the tube to TSP crevice.

Byron Unit 1 and Braidwood Unit I have not experienced corrosion induced

! tube denting. It is shown in WCAP 14273 Section 12.4 that a diametral .

i reduction of 65 mils is required to develop stress levels above yield in the TSP ligaments at dented intersections. For this reason, a requirement that a tube must be capable of passing an appropriately sized bobbin probe j

through a dented TSP intersection for 3.0 volt IPC to be applied is' q j

i incorporated. If an appropriately sized bobbin probe can not pass through a yf j

dented repair intersection, criteria applied to the all cold surrcunding leg TSPs. tube locations must be repaired to

TSP Degradation Related to Chemical Cleaning

4 During a 1995 steam generator eddy current inspection at a foreign plant, i distortions were observed in one area at the top TSP in two of the three j

steam generators. Visual inspections discovered a small section of the TSP in an area next to the wrapper was missing. The missing parts were found 1

t resting on the TSP below the top TSP. The support plate also showed 2

indications of metal loss estimated as high as 0.25 inches of the 0.75 inch thick TSP. The edge of the TSP near the outer wrapper also appeared to have some metal loss. A review of the 1994 eddy current inspection data

indicated that support plate distortions had been present, but not identified

~

during the data review process. A root cause investigation was performed by the utility which concluded that the damaged TSPs were a result of a

{ chemical cleaning that occurred in 1992. The chemical cleaning process was developed by the utility and was significantly different than the modified i

EPRI/SGOG chemical cleaning process performed at Byron Unit 1. The foreign process used a Gluconic Acid / Citric Acid solution at a pH of 3.2

recirculated at a high velocity for 150 hours0.00174 days <br />0.0417 hours <br />2.480159e-4 weeks <br />5.7075e-5 months <br />.

i k ala bybodipeasup3 wyf31 1

- -w ,v - - - - - - ,,~,w -, ,,- -- , - - - .--- - - - - - - ,, --n-

i1 1 l

6  :

$I l

The solution was recirculated from the blowdown pipe through a pump 4

which discharged through three hoses placed above the tube bundle. The ,

1 j flowrate through each hose was required to be less than 1 meter /second, but i

actual Dowrates were up to 2 meters /second in the two affected steam i generators. The solvent recirculation rate was 200 cubic meters / hour in the two steam generators and 50 cubic meters / hour in the unaffected generator.

In addition, one of the hoses in the affected generators was mispositioned 4 and was too close to the top support plate, thus creating a higher solution velocity at the top TSP.

1 3

The higher recirculation velocities is thought to have stripped the cleaning l solution inhibitor from the base metal allowing excessive corrosioti iates to

' occur at the top support plate and the blowdown line. This phenonemon was confirmed through testing and also demonstrated by EPRI during the qualification testing of chemical cleaning processes. The metal loss due to j corrosion at the top TSP was estimated to be at least 250 mils.

The Byron Unit 1 chemical cleaning that was performed in 1994 used a modiSed EPRI/SGOG qualified process conducted at high temperature and was significantly different from that performed in the foreign unit. At Byron, the solvent contained a 30% EDTA solution with hydrazine and an inhibitor. The pH of the solution at Byron was 8.0, as contrasted by a 3.2 pH for the foreign unit. The total contact time at Byron was approximately 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />, compared to 150 hours0.00174 days <br />0.0417 hours <br />2.480159e-4 weeks <br />5.7075e-5 months <br /> for the foreign unit. The Byron chemical cleaning used a fill and drain technique, instead of the recirculation l technique. The Byron process involved passivation with an ammonia /hydrazine solution. The overall corrosion measured by actual material coupons was 2.16 mils, well within the allowable limit of 11 mils.  :-

Eddy current inspection of the steam generator tubing was performed aner \

the Byron chemical cleaning with no TSP distortions reported. \

Due to the differences in the chemical cleaning processes and the root cause analysis that was performed by the foreign utility, Comed has determined that the TSP damage mechanism at the foreign unit does not apply to Byron Unit 1. No chemical cleanings have been performed at Braidwood Unit 1.

The specific details of the damage experienced at the foreign unit were discussed in proprietary meetings between Comed and the NRC held on July 20,1995 and August 17,1995.

- - ,em

1 s

t TSP Degradation in the Area of Anti rotation Devices:

Cracked tube support plates have recently been reported by one foreign i

utility, Cracked TSPs were discovered when distorted eddy current '

indications were visually inspected.

The degradation seen has been limited to the top TSP, in the area of the anti-rotation device. This area is unique to the steam generators in that it is where the wrapper anti rotation device connects to the ID of the shell.

Preliminary review of baseline eddy current data indicates that the degradation may have been present prior to plant operation, indicating that it may be the resuh of a manufacturing problem. The degradation has been seen in at least one of the three steam generators in each plant inspected thus far.

The steam generator internals of the affected foreign units are similar to the design of the Westinghouse Model 51 steam generator, in particular there is one anti-rotation device of similar debign at the top TSP. The single anti-rotational device carries the full load associated with the shell to wrapper motion. The rotational load is believed to be transferred to the TSP, which may increase susceptibility to TSP cracking. However, due to the design differences of the wrapper internals, the Westinghouse design is less susceptible to cracking at this location.

4 The steam generator internals of a Westinghouse Model D-4 design have three similar anti rotation devices at the top support plate. The three anti-rotation devices distribute the rotational loads at the three locations which result in lower loads at each individual device, as compared to the foreign units design. Again, there are design differences internal to the wrapper that make the Westinghouse design less susceptible to'eracking at this

location.

Comed believes the TSP cracking seen in the foreign steam generators is unique to their design and the fabrication of their internals. Eddy Current inspections at Byron and Braidwood have been performed with no missing support plate signals being reported. Comed is working closely with the foreign utility in understanding the root cause of the TSP degradation.

Comed is also working closely with the industry to develop improved TSP inspection technique,s.

d

! -- .,m J

.-v . _ , , , _ _ . _ . _ . . - _ .. _,_ . - - , - _

i j- .

I a

j The specific details of the damage experienced at the foreign unit were discussed in proprietary meetings between Comed and the NRC held on July 20,1995 and August 17,1995.

Although Comed considers the TSP cracking experienced by the foreign steam generators to be unique to their design, an inspection of this TSP

! along with other SG internals which would be in the MSLB load carrying path as a result of application of the provisions of this amendment will be 3

conducted in accordance with the Byron /Braidwood SG internals inspection i

plan. Results of this inspection will be reported to the NRC in accordance

• with the provisions of the reporting requirements section of this amendment j request, i

! Distorted TSP Signals Related to TSP Geometry:

i j

Recently, a domestic utility reported potential TSP. degradation unrelated to

' denting. The utility had reported a small number of distorted TSP signals over the past three refueling outages. Upon further review of the data, and j

mapping the signals on a tubesheet diagram, it was shown that these j

signals were associated with TSP geometries. Specifically, the area where the access cover is welded into the TSP produced distorted signals due to permeability variations. In another location, a distorted signal is believed to be associated with the weld between the wedge and the TSP.

These signals have been reviewed by both the utility and Westinghouse and are not believed to be the result of TSP cracking.

1 j Tube Bundle Wrapper Dislocation j

4 One foreign unit has observed a dislocation of the tube bundle wrapper.

Apparently, this was the result of the wrapper not being appropriately attached to its support structure. This was discovered when they were j

unable to pass sludge lancing equipment through a handhole in the wrapper j due to mi -lignment of the wrapper and handhole of the shell.

Byron Unit I and Braidwood Unit I have performed steam generator sludge lance operations each refueling outage and have not observed misalignment of the wrapper and the shell. The foreign unit was of a different design and manufacturer than the Byron Unit 1 and Braidwood Unit 1 Westinghouse Model D-4. The wrapper support design is significantly different from that j' used on the Model D-4 steam generator.

I i

i j k als bytmdtasamapa wet.34 i

i i

'i i

! INSIDE NRC ARTICLE - FEB. 96 l (IN 96-09, Damage in Foreign SG InternalsD i

i l SG BUNDLE WRAPPER COLLAPSE AT i BLAYAIS 3 j SG1 20mm bundle wrapper collapse discovered j during sludge lancing in June 1994.

All 6 vertical support welds cracked.

j -

Upper TSP moved (5-13mm) l

\

SG2 l

Smm collapse j -

3 of 6 vertical support welds cracked i

! l i SG3 l No Collapse l

i 1

j Potential Safety Concerns identified l Concerns with total collapse include loss of l feedwater and damage of biggest radius U-bends.

)- Concerns with tearing away of a vertical support 3

include loose part interaction and SG tube rupture.

4

) Other concerns regard earthquake and LOCA.

L

N I

i 1

q SG BUNDLE WRAPPER COLLAPSE AT BLAYAIS-3 Ccon't)

EDF is still investigating the causes of wrapper collapse and has postulated the following; Unanticipated axial restraint against differential thermal expansion between the wrapper and the SG pressure vessel shell led to significant loading of the vertical support welds. The restraint may have been due to thermal ,

expansion between the 7th TSP and the shell.

1 These units use a special operating procedure to .

I accelerate the transition from hot to cold shutdown by injecting AFW into still-hot SGs.

Poor quality of wrapper vertical support welds may also have been a contributing factor.

EDF has installed displacement sensors with strain gauges to allow inservice surveillance of the wrapper position.

EDF has also initiated a major design review of SG Internals that is expected to be complete by Fall 1996.

r

.]

TUBE SUPPORT PLATE (TSPD DAMAGE IN MODEL 51 A & 51M FRAMATOME SGs SGs similar but not identical to Westinghouse model 51. TSPs are drilled hole and fabricated from carbon steel. i FESSENHEIM-2 (51 A) l l

Initially discovered in April 1995 after anomalous eddy current reading was visually inspected.

Pieces of 8th TSP were found resting on next l lower TSP. A similar problem was found in Unit 1.

1 Chemical cleaning performed in 1992 identified as cause. Pipes used to direct chemical solution was installed incorrectly, too close to 8th TSP. High impingement velocity of the solution rendered corrosion inhibitor ineffective.

TSP damage eliminated lateral support of tubing within affected region, raising concerns regarding tube stresses due to flow induced vibration, earthquakes, and LOCA. 224 unsupported tubes plugged.

Planned repair is to insert anti-vibrational cables within insufficiently supported tubes, replug those tubes, and install sentinel plugs on the cold leg tubes.

,e , - .., ,- ., -..,- - - -. . ,, -- - - - -. -,,-.c., , .,m,-- ,- - - - -.

1 $

l TSP DAMAGE IN MODEL 51 A & 51B l FRAMATOME SGs (con't?

) ST. LAURENT-B2 (51M)

Although less severe, large cracks in 8th TSP, near li seismic stop, were found. 20 tubes plugged.

l -

No chemical cleaning performed and damage could i be traced back to staitup. Differential expansion

of SG parts is possible cause.

l l

Similar indications found at Tricastin-2 &

j Dampierre-4 both with model 51M SGs.

l 4

GRAVELINES-2 (51M) i 1 -

Cracking found at various TSPs, not just 8th TSP.

l 274 tubes pluggged.

1 l:

Damage thought to have evolved over time and possibly caused by errosion-corrosion from

seawater.

t

{

TSP damage justifies early replacement of SGs.

1 i

l Unit 4 found damage on SG flow distribution plate.

I i

)

i

)

i i ;

l DAMAGE IN FOREIGN SG INTERNALS i

1 i TSP damage was not detected earlier since NDE concentrated on tubes.

l Inspection difficulties include ECT and camera j results are not detailed enough, time consuming and labor intensive. EDF is hoping to develop j better inspection methods.

l l EDF/DSIN trying to understand phenomena i involved.

! EDF plans to conduct destructive testing on SG l TSPs with similar degradation from Gravelines-1 l model 51M SGs (removed from service in 1994).

~

EPRI has initiated an effort to develop qualified j procedures for detecting TSP cracks in the US.

I NRC IN 96-09 highlights that other mechanisms j may lead to TSP damage and notes that APC is i based on TSPs being capable of locally l constraining tubes against rupture.

l l The staff continues to monitor information on TSP j and wrapper damage as it becomes available.

i 1

l i

y GENERIC LETTER 96-XX DEGRADATION OF STEAM GENERATOR INTERNALS CRGR REVIEW i NOVEMBER 19,1996 Stephanie M. Coffin Materials and Chemical Engineering Branch Division of Engineering t

PURPOSE

1) Notify addressees of damage to steam generator internals at foreign PWR facilities

2) Require addressees submit information to determine:

A) What initiatives have been taken in response to the foreign experience l

B) If similar degradation has been identified l

Information will be used to verify compliance with 10 CFR Part 50, Appendix B 2

DAMAGE TO STEAM GENERATOR INTERNALS AT FOREIGlV PWR FACILITIES Plants similar to Westinghouse Model 51 Steam Generators

1) Wastage of the uppermost support plate--

Misapplication of a chemical cleaning process

2) Broken tube support plate ligaments at uppelmost TSP--

Excessive stress during thermal treatment

3) Wastage of tube support plates at various elevations--

Cause is undetermined

4) Tube bundle wrapper drop and cracking of the wrapper--

Excessive stress during cooling transient, poor welds; l

cause of cracking unknown i

3 )

l n -

~

SIGNIFICANCE OF EVEN TS j

> Degrada~ tion of the tube support plates and the tube bundle l wrapper directly affect the integrity of the steam generator l tubes

)

• Tube support plates prevent lateral displacement, vibration, and minimize bending moments in the tubes i

i

> Fall of the tube bundle wrapper can lead to a loss of feedwater, damage to steam generator tubes i 4 ,

FOREIGN RESPONSE TO STEAM GENERATOR 4 INTERNALS DEGRADATION To detect wastage and cracking of tube support plates:

• Eddy current testing of the steam generator tubes j i

> Visual and video camera inspections of the secondary side

Response

4

> Tubes without adequate lateral support are stabilized and plugged  ;

5

FOREIGN RESPONSE TO STEAM GENERATOR INTERNALS DEGRADATION i

4 To detect tube bundle wrapper drop:

i Visual and video camera inspections of the secondary side j

• Online instrumentation of the bundle wrapper

Response

• Repairs completed on damaged bundle wrappers 4

6

APPLICABILITY TO US PWR FACILITIES Damage mechanisms at foreign PWRs are generic; not obviously foreign practice or model/ design-specific problems

> Cracked or deformed tube support plates found at dented intersections in U.S. steam generators (Surry 1&2, Turkey Point 3&4, San Onofre 1, Indian Point 2, Millstone 2)***

> Possible tube support plate ligament cracking at Diablo Canyon (not necessarily associated with denting)

> Eddy current and visual inspections performed at ,

Braidwood Unit 1 and Byron Unit 1 found no evidence of tube support plate or tube bundle wrapper degradation

• • • Except for IP-2, all SGs replaced or plant was shutdown 7

___- ______-_= _ __ _

_ =__ _ _ _ _

_ _ _ _ _ . _ _ = = - _ _ .

___.= --_ ]

REQUIRED INFORMATION

1) Discuss the program you have in place to detect degradati~on of steam generator internals A) Discuss reviews of past eddy current data B) Discuss secondary side visual / video camera inspection results -

2) If no such program is in place, describe plans to determine

if degradation of steam generator internals has occurred or discuss why no such plans are needed 4

8