ML20101A478

Provides Insp Plans for Mitigation of IGSCC During 1985 Refueling Outage.Plans for Weld Reinsp During 1985 Refueling Outage Satisfy Requirements of Generic Ltr 84-11

ML20101A478
Person / Time
Site:	Peach Bottom
Issue date:	12/14/1984
From:	Daltroff S PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC
To:	Stolz J Office of Nuclear Reactor Regulation
References
GL-84-11, NUDOCS 8412180595
Download: ML20101A478 (56)

Text

-.

PHILADELPHIA ELECTRIC.COM PANY 23O1 MARKET STREET P.O. BOX 8699 PHILADELPHIA. PA.19101 SHIELDS L D ALTROFF JEcTEU,o -

December 14, 1984 Docket No. 50-278 Mr. John F. Stolz, Chief Operating Reactors Branch #4 Division of Licensing U.S. Nuclear Regulatory Commission Washington, D.C.

20555

SUBJECT:

Peach Bottom Atomic Power Station - Unit 3 Corrective Actions for Intergranular Stress Corrosion Cracking (IGSCC) during Spring 1985 Refueling Outage l

REFERENCE:

1.

Confirmatory Order, J. F. Stolz, USNRC, to E. G. Bauer, Jr., PEco, September 7, 1983 1

2.

NRC Generic Letter 84-11, " Inspections of BWR Stainless Steel Piping 3.

Letter, S. L. Daltroff, PECo, to D. G. Eisenhut, USNRC, Response to Generic Letter 84-11, June 4, 1984 4.

Letter, S. L. Daltroff, PECo, to D. G. Eisenhut, USNRC, Supplemental Response to Generic Letter 84-11, October 5, 1984 5.

Meeting with NRC Staff, October 9, 1984, Bethesda, MD

Dear Mr. Stolz:

This letter provides our plans for mitigation of Intergranular Stress Corrosion Cracking (IGSCC) in primary system piping during the Peach Bottom Unit 3 Spring 1985 refueling outage in accordance with the Refarence 1 Confirmatory Order.

This inspection plan is submitted for your review and is the basis for our justification to return Unit 3 to full power for a minimum of one additional cycle following the 1985 outage.

During the 1983 refueling outage on Peach Bottom Unit 3, augmented inservice inspection of welds in the recirculation and residual heat removal systems was performed in accordance with d7 the requirements of I.E.Bulletin 83-02, " Stress Corrosion 7I Cracking in Large-Diameter Stainless Steel Recirculation System 8412180595 841214 PDR ADOCK 05000278 P

PDR r

Mr. John F. Stolz D:ctmbar 14, 1984 Page 2 Piping at BWR Plants."

A total of 112 welds or 75% cf all welds in these systems were examined after ultrasonic (UT) indications were discovered in the original sampling.

The welds which were not inspected were of conforming material as specified in NUREG-0313, Rev. 1, or were not practicable to inspect due to high radiation or physical inaccessibility.

The results of the inspections, including crack growth analysis and corrective actions taken by Philadelphia Electric Company, were reported to the NRC in letters dated August 9, 1983; August 22, 1983; August 24, 1983 and August 30, 1983.

The corrective actions taken by Philadelphia Electric Company were found acceptable by the NRC in the Reference 1 Confirmatory Order.

The following describes the inspection plan to be performed during the Spring 1985 Unit 3 refueling outage.

CONFORMANCE WITH GENERIC LETTER 84-11 INSPECTIONS Philadelphia Electric Company will perform weld inspections in response to Generic Letter 84-11, " Inspections of BWR Stainless Steel Piping", dated April 19, 1984.

The Generic Letter was issued as a result of Bulletin 83-02 inspection results, and the NRC concluded that certain actions to be taken by Licensees, in accordance with the Generic Letter, would be considered as acceptable responses to current IGSCC concerns.

These actions were specified in Section 5.4.2 of NUREG-1061 (Dra f t) as an appropriate reinspection program and were derived from SECY-83-267c, " Staff Requirements for Reinspection of BWR Piping and Repair of Cracked Piping" (Commission paper), November 7, 1983.

The Philadelphia Electric Company response of June 4, 1984 (Reference 3) to Generic Letter 84-11 was reviewed by S.

D.

Reynolds of the NRC Region I Staff during Inspection 50-277, 278/84-21, 84-17, and found to meet the requirements of the Generic Letter.

Based on our ongoing effort concerning IGSCC mitigation, certain specific weld locations specified in that response were modified.

These welds were identified along with the technical justification for the changes in the Reference 4 letter.

We believe these changes enhance our conformance to Generic Letter 84-11.

A summary of the welds to be examined specified in the Reference 3 and 4 letters is included as.

The Generic Letter 84-11 inspection plan was presented to the NRC Staff at the Reference 5 meeting.

ADDITIONAL INSPECTION As a result of indications discovered in welds in the Unit 2 reactor coolant primary boundary piping systems during the current Unit 2 outage, Ph'iladelphia Electric Company will perform additional inspections of welds in the reactor

Mr. John F. Stolz' Dacamber 14, 1984 Page 3 nozzle-safe end areas on Unit 3, beyond the requirements of Generic Letter 84-11.

These weld locations are specified in.

QUALIFICATION OF EXAMINERS

- All Level II and Level III personnel utilized for the ultrasonic examination of IGSCC susceptible welds will have successfully completed the EPRI NDE Center's IGSCC UT detection program.

All Level I personnel. utilized for these exams will have received training and familiarization in the techniques and unique requirements of UT scanning for IGSCC.

Additionally, any IGSCC indications detected will be sized by personnel who have successfully completed the EPRI NDE Center's "UT Operator Training for Planar Flaw Sizing."

SUMMARY

Technical Justification for Multi-Cycle Operation of Peach Bottom Unit 3 Recirculation and RHR Piping", October 1984, prepared by the General Electric Company, in conjunction with the reinspection program, supports the deferment of piping raplacement based on the weld overlay repairs and induction heating stress improvement (IHSI) treatment on primary system welds during the 1983 Unit 3 outage.

The reinspection scope will be expanded in accordance with I.E.Bulletin 83-02 if new indications are discovered.

l Section 6.0 of NUREG-1061 (Draft) provides guidance on the L

decision and criteria to evaluate for replacement, repair, or l

continue operation without repair such as leak-before-break criteria and stress corrosion crack growth rates.

Evaluations for continued operation with or without repairs and the criteria for crack repairs must be sufficient to i

provide full ASME Section XI IWB-3640 margin during the operating period.

NUREG-1061 (Draft) acknowledges that

" operating experience and fracture mechanics evaluations indicate that leak-before-break is the most likely mode of piping failure."

During the 1983 outage, appropriate leak detection procedures were implemented and will be continued following conclusion of the Spring 1985 weld reinspection i

program.

This includes moisture sensitive tape installed on selected welds and revised Technical Specifications regarding Limiting Conditions for Operation and Surveillance Requirements on the drywell sump pumpout rate regarding unidentified drywell leakage.

t 4

4 f3

.m,..

..y, y,,,-

m.m.,_.,,,

r

_,...-.,4%.,~r4,. _ _,

,,.y,,..,_._%.,.-

,,-._,,#y

e 9

Mr. John F. Stolz December 14, 1984 Page 4 All crack indications discovere'd during the last outage were weld overlay repaired which, based on the compressive stresses induced by the overlay, should be effective in preventing the initiation-of new IGSCC cracks and inhibiting the: growth of existing flaws.

Permanent utilization of Hydrogen Water Chemistry -(HWC) is being investigated, and an Amendment to the Technical Specifications to allow testing of HNC injection was issued by the Commission on November 14, 1984.

Based on the results of this testing at Peach Bottom (and at Dresden Unit - 2), PECo will make an evaluation on a permanent HWC injection program for Peach Bottom Units.

CONCLUSION The plans for weld reinspection, during the Spring 1985 Unit 3 outage, satisfy the NRC requirements of Generic Letter 84-11, are supported by General Electric Company, and are substantiated by the issuance by NRC Staff of NUREG-1061 (Draft).

Acceptance of weld overlay repair,'the beneficial effects of IHSI treatment and HWC control implementation, as supported by NUREG-1061, have indicated that weld reinspection on Unit 3 is appropriate at this time.

Very truly yours,

,7

?

d Attachments cc:

J. H. Williams, Resident Inspector

e Attactynent 1 Generic Letter 84-11 Category Planned Actions 2.Co) -

Inspect-20% of welds in Recirculation System:

each pipe size not previously 22 inch piping - Inspect 2 welds:

Inspected.

2-BM-2 2-BM-3 28 inch piping - Inspect 4 welds:

2-AS-9 2-BS-8 2-AD-13 2-B0-16 Inspect 20% of welds previously Recirculation System:

inspected and found not cracked 12 inch piping - Inspect 6 welds:

2-AHK-4 2-BHE-3 2-AHH-3 2-BHC-4 2-AHF-3 2-BHA-3 22 inch piping - Inspect 2 welds:

2-BM-4 2-BM-5 28 inch piping - Inspect 5 welds:

2-AS-5 2-BS-6 2-AD-17 2-BD-15 2-BS-4 RHR Shutdown Coolino System:

20 inch piping - Inspect 2 welds:

10-0-3 10-0-2 24 Inch piping - Inspect 4 welds:

10-IA-7 10-IB-4 10-IA-10 10-IB-11 2.(b) -

- Inspect all unrepaired cracked welds.

No inspection; all cracked weld locations were weld overlay repaired.

E4 5.^

t.

s,.

(.-

2.(c) -

Inspect all weld overlays on welds RHR System:

where circtrnferential crack length 20 inch piping - Inspect 5 welds:

Is greater than 10% of the' 10-0-5 cI retsnference.

10-0-6 10-0-7 10-0-10 10-0-15::

% te: The circtsnferential crack length at this location is less than 2% of the circtroference.

Nuclear Stearn Supply System:

4 Inch piping - Inspect I weld:

The N8-B det Ptrop Instrtsnent Seal Safe-End to Reducer weld.

3 2.(d) -

Inspect IHSI treated welds which No inspection; all 91 IHSI treated had not been post-IHSI UT acceptance welds were post-IHSI POE.

tested.

y 2.(e) -

If new cracks or significant growth Will address as required.

of old cracks is detected, expand inspection scope in accordance with NRC I.E.Bulletin 83-02.

l r-n e m

-.. - -,, -m-w-,,-,nn,,,--.>n~n, a--ev-.-m---c-e--n,,-----,------n, ew,---e<-

m-

l

{~

s b

e

?. E My -

O I

VSt-OR dh I-18-08 s

[eb U

S-Rf-t N0!!V513N3J z

H E

7 43 0

183MNiv!NO3 A

U O

7*

I E

Ib t-V8-08

<f,+

f g

,=

• Olt

~

\\ \\

m S-dNV-t -

7 0 VB-04

\\ \\

1*

i,-

cr s

7 s*8 x

6-dNV-8 u

g e-t-dNV t 77 o

o w

m a

7

• 2 f'+

E N

• ODE m

S-suv-t -

7 7*

6-sNv-t EI V 4-0 8 "A,

r k*

o t-sNv-t - -

88-V8-04 33

+f'+

m 7E M

• Gl!

3-NNV-t -

"4

-e n

e -V10 30J 7?

A 4 "f E' a; ca g

D-NNV-2 z

A v.g -

p ll-VI 06

/

[

g.N u,#

y;I a

j o

e-

• III o

*8' O.

I. A.

Vit-On c

s-(Nv-t- -

4SA 44 g g 88-07 s o

g D-fkf*l y

=

O t-int-I -

E2

,g

. L V*l*VJI-I a

'Ott

't I

NE IdI#II*II*I P-38-0f-8 S -IN f-t-7 D

?N z

u N

hh

)

3 b ll-318 "k

.$.a B m

t-1tuf-! -

04-3f 8 Z

gngJ ggg333,y, m

9-3-nv-t VSS-0M

• 4 ih_ g-gy.g l

ca l.

m go l

4 W.

E l

m A

L H

1 5

Um

,A.

E2

[

G=

l-tf I * '

.O

=

as

=

.=

z I.ty.t Q

Un A

"7 t-tv I ftD-On M

_J 3 33.g t-t-tf t - i t

U I

l-0-04

/g 4

u A

A380 el 3. 31/l 0 01

[

u

- f t-0-0 6

{ 0'-01 A

t 0-05 a

Sgd f

a d

M S

1. a.O f

Q Hz D

=

@'g'h I

M d

A A

+

I CZ3 m

b A

f

/,a u

M e

/

,<\\v '

e,

~

2%

UE o, '*.,,

+

s.:*si -==

..g

=-

u nn Ef

~

s.

3

&.%.+

O 5

s.o

?

G O

N +,f.o,

.R x

4

m.

a a

I e

30 2

E 3

e4

~

a-.

P

?!

est-on h;se1,et3aja t-et-si o-t-ne.g ess-en g

g L -ne-t

=

g gainesvanes s

0 2

7i

=

o.

=

a 23 t-et-on

1. e'+

H n.

• et

=

~

~

s-ene.t-

\\.a \\

~

e-en-ai 2

m 2

6 The.g N Y

• =4

=

ee8 M

aE

's' t-the-!

"b IY

>=

~:

=

g

• e'+

w cc

• es

=

2 s-ene-t-7R a

- e-et es

=

.z

, gag.g ~

7

,= >= g",,.e,.e. ;;

s t- :;t

=

T

=

es 5 :.

3-3 e-t-

= 1

,Av

. e,s.te.

t

,.gge.g

=

Va

""t e

. 8 cm N

?

== A u

4A 2

DB-el si m

8 3he t n

_/

ii i,

9p'*

~

M J

~ l e

• ett s-eNe-t-

= *,

7 O

ti-se t

/ ggg.en

-4 ao

. w8 o

e eas-s q E

fhI 1"A

/

l=

ti-set-3 o

iE t.e.e-t -

^ A,

=a A,.....e.!

~

g

+.+

~

~

hs

- e m ii se-t

.es i...e.t._ _

y

=

ii.ee-I - >

o m

t 3me t

=

3g

-a,,

A y

g 3

gg.gg.g M

t ane I -

• sit e cm

.nnd 'setsje.e.

g.te t n

m ce s-ne am I

.e cm g

I z

o a

w I

ca.

I

>4 l

5 '

A I

2 e

o I

w 1

J

=>

l c

j cm M

l u

7 W

=

}

=

l

=

CE j

/

Ih s

I-te*t h"

E t le-t j

e.ge.I et> en EC

^

g.te t u

t te t Q.

\\

>=

.J w

w a

a 2

I

?

?

=3 E

b 1

O M

e M

M

• M O

A

.t cze

• M L

u

• E I.

re " a

=m W

Esi:'

~

WB

.o 8'8 4

.,a d

e'0 I A 6L Nke,

~

f l

l 1

'l vi,,,...,,,,,,,,,,. a. _ _,.. _-

4 pg an.

. su.w,..... ~.,.

m 4

g.W:

p<,

m.

ny

-I y

.s:

• fr/mut

~~

..%f3a/c.

h End.fo k...>.~...x'.,.ss~w,.c.os.sssy Pesea, Wall IcT Rf? If5fPJFSK S9L f

e 6

I

i Attactwnent 2 Addltlonal inspections Plamed Inspect lons Beyond Weld Locetions Generic Letter 84-11 Requirenents 1)

N-1 RPV Outlet UT examine mininun of 1 weld location.

Nozzle to Safe-End (2 locations).

!2)

N-2 RPV Inlet UT examine 5 weld locatlons.

Nozzle to Safe-End (10 locations) l3)

Internal Attactinent Perform best effort UT examination weld of Thermal Sleeve at these 10 locations.

4 to N-2 Nozzle Safe-End (10 locations)

4)

Other weld locations in Perform Inspections required to meet containment piping systems.

ISI Program requirements.

4 T

(

9 f

4 s

AtY&%

s eesi no I

l'

.g g

t 4

E I-lR gh I

nl n

o E

8 N-

.E N

• 0 2

< s.

5" i

g

lo n

a ill 8

0 1

aB g2 9

a,i,

n sa E !

$0 b

f !E g

i z

s 12a x

s b.

R N

/

!U

/

sE

/

s e4

[

h,h'

[f e-s u

b B

h g*

s 5:

FIGURE 4

. METALLURGICAL CONDITION OF PEACH BOTT.OM UNIT 3 l

i i

i SAFE END TO NOZZLE ATTACHMENTS i

l l

RECIRCULATION INLET A*Ps e3 Is e

e-i

\\

C.

..,s t 1

e 3'.

r a

s.

i l

D-t i

~

le,*,.t i

x g'

r 6

• 1 t..' '

9;

. _.,i lIs LL, b.

?

i i

n

{

h i

A f

n I

l

U N I T 3' e

RECIRCULATION OUTLET

\\; -

e J*

i se

.C e *I

-B 8

d*

S s

e, e-s h

r^

eg e

e*

D

~

4

{

g

• p' e

4 4

a*.

1 4

e s

• .=

4 J

A)

LOW CARBON STAINLESS STEEL SAFE ENd B)

INCONEL ALLOY 182 WELD CLAD (ID ON NOZZLE SIDE-MAY BE DILUTED WITH TYPE 308 FRCH ORIGINAL FABRICATION) 6 C)

INCONEL ALIOY 192 WELD METAL I

D1 LOW ALICY STEEL NOZZLE A.

4 E)

STAINLESS STEEL N0ZZLE CIAD S

d e

. tt-_

o ATTACHENT 3 GE Report No. 137-0010, >%R 84-21 (Rev. 2)

Technical Justification for Multi-Cycle Operation of Peach Bottom Unit 3 - Recirculation and RHR Piping anm ~ _ J.

(

9 e

i i

F

'f s

kN. ~

la s

a Table of Contents laat 1.

INIt0 DUCTION 1

2.

WELD OVERLAY DESIGN 2

3.

FEAG BOTIUM UNIT 3 VELD OVERLAY CONFIGURATIONS 7

4.

RESIDUAL SMESS IMPROVEMENT 8

3.

CORROSION CRAMING RESISTANG OF TELD MARRIE 11 6.

FRACTURE 3EGANICS MARGIN 16 7.

21'SULE PROM DBORADED PIPE PROGRAM 21 8.

IRL'I 11 RATED WELDED JOINH TIHOUT MAGING 23 9.

00N.1.USIONS 24 10.

REFERENCES 25 t

a DEF #137-0010 MAR 54-21 (Rev. 2)

ESMD3.M NGNICAL JUSTIFICATION FOR MULTI-CYG.E OPERATION OF PEAG BOPIWM UNIT 3 BECIRCULATION AND RER FIFING October 1984 Prepared for Philadelphia Electric Company E. 5. Mehta

8. Ranssaath General Electric Company

.m~

l t

1.

INTRODUCTION During the 1983 estage, Peach Botten Unit 3 roeiresistion/RER piplag was h as 1551 and sold

, Anspected for 108CC sad the appropriate remedial actions ass overlays were taken by Philadelphia Slestrie la seaformasse with the zogstrements of 15 Balletin 83-02.

A summary of the member of welds examined cad the remedial actions is given in Table 1.

A total of 149 welds was 1svolve.1, of which 125 welds were oossidered as asseeptible to IGS,C based on l

o esseeptibility matriz evaluation. Minety-one welds were IISI treated, est cf which 15 had IGSCC ladioations. Told overlay repair was performed at all 15 loestions. Also, la Jane 1984, weld overlays were applied on the 4-imah jet pump lastruentation mossle/ safe end.

At the present time, these correstive setions have besa assepted by the NEC for one eyele operatloa.*

I This report documents the teshaisal bases to justify somtlased operation beyond one fuel eyele.

i l

i l

-e

• NOR3G 1041 presently under review la draf t form recommends approval of plant operation with overlays for at least two feel eyeles.

If additional mitigation messares are implemented, entended operation beyond two feel systes may be permitted.

1 E.,(

f.;

i a

r I

2.

WELD OVERLAY BSSIGN l

I l

This seetles first describes the strastarsi design basis for the weld everlays.

Spesifie aspects of Peash Betten 3 weld overlays are thes disens sed.

I 2.1 Strustaral Intenrity Asnests of Wald Overlav i

The disenssies in this seetles has been essentially esserpted from Referesse 1.

Peash Botten Unit 3 resiresisties and RER piping were designed to meet the rogstroseats of ANSI B-31.1.

Spesifiestly, the requirements of this Code are intended to provide design margias agatast oncessive plastie deformaties and l

destile asyture and to assare that fetisse is11sre will set ocear as a result i

f of eyelie leadiass.

Seetles 11 of the ASIE Code applies to la-servios evalsations of these sempements. A weld overlay en a erasked piplag oesponsat des 13aed to meet Bestion II. I53-3640 (Referesse 2) requirements assares that the same minimas seastraation Code margias are malataland af ter assesatias for the laitial overdesign of the eesposest and for the less la strustaral I

sepability dse to the presones of the erask, and seasidering the ersek growth t

daring the projested isture period of operaties.

In the following sections I

the design stress levels la overlays and the assaolated design margias are discussed la greater detail.

2.1.1 Design Stress Levels i

Applied stresses la piplas are elassified late three sategoriest primary, i

ascendary, and peak stress.

Primary stresses are developed by imposed leading j

l (e. g., pro sesse, dead weight, and seismie) and are ne't self-limiting.

(

Seeendary stresses are displacessat-governed stresses ressitias from self oenstralst of the streetsre. They must estisfy the less1 strela esatissity regstrements and are Asherently self-11mittag. As important feature of a assendary stress is that the assestated strela dissentianity som be assammedated by yieldias la destile matorists. Therefore, the seseadary i

stresses de met affect the limit lead of the streetsre.

Feak stress is a i

leest stress, oessering mainly la regless of geometris discontissity or temperatste sea-11asarity, and is of someers saly la fetisse life evalsaties.

2 l

Primary membrano and bending stress limits in piping are:

In f 8, P, + Pb f 1.5 Sh chsre P, is the membrane stress component (due mainly to pressure) and Pb is M

the bendias stress (dse to weight and salmais moments).

S is the ASE desiga

[

h stress intensity equal to the lesser of 0.9 S at 550*F or 1/3 5,.

ANSI B31.1 y

imposes similar limits on primary stresses which assare essentially the same O

design margins. The Bestion III Primary plus Secondary stress (P + 4) limit I

is 3 S,.

The P + 4 limit asesses elastia shakedown and minimises plastic I

deformation. The P + e limit does not directly affect design margins in destile materials. They may be esseeded provided that the effects of plastielty are properly assosated for as they may affest fatismo analysis and lassemental deformatics. There are ao speelfie limits om peak stresses apart frem their effect on the f atisse analysis.

j Based on the above disenssion, it is clear that the important reguirement la cold everlay desiga is to limit the primary stress. The limit on primary l

membrase stress applies mainly to the keep stress.

In general, pipes are

(

sised sash that the hoop stress for normal conditions is less than or ogsal to 8,.

Therefore, the hoop stress for normal senditions is less thaa 8, and the sorresponding salal pressure stress (whish is of interest for sireumferential creeks) is less thea 0.5 8,.

Ja general the mala sostributies to the salal ctress is the pressure stress. Other primary stresses (dead weight and seismis) are assally aseh smaller.

In BTR primary piplas the total asial ctress laeluding contributions of pressure, dead weight, and seismie leads is cell below 5,.

It is elear that weld everlay thiskaess required to malatala is11 Code margins is less than the is11 pipe thiskaess, even if the strastaral i

espab!11ty of the original pipe materlal is not seasidered la detesmiains the stress espability of the everlayed pipe. In feet, for mermal and upset soadations, the P,+ Pb in the salal direction is of ten less than 0.6 S and h

the ' Cede design margins saa be malataland with everlay thiekness less than half the original pipe well thiekness. Speelfically, the overlay thieksess is

. detesmined based on the 2T3-3640 requirements for the everlayed pipe f

3

l r

configuration (thioksess equal to original pipe wall plus the overley) and cath stresses adjusted for the new thioksess. As diseassed later, 175-3640 requirements provide a safety margia of approximately 2.3 based on limit caaly sis. This is comparable to the margias for ascracked piping desismed to ANSI E31.1 requirements and subjected to stress levels sorrespondias to the naziman allowable Code loadings.

The three overlays on welds 10-0-05,10-0-06, and 10-0-07 la the RER system core sized without taking eredit for the uncracked portion of the original pipe. his is defined commonly as a fall strastaral overlay. The ITB-3640 requirements are satisfied assumias a 360* eiroaaforential erack of depth ognal to the original pipe thickness t.

The veld overlay thioksess, h, is determined sneh that the non-dimensional erack depth t/(t + h) meets the coseptamos eriteria of ITB-3640 for a 360' erack under the applied stresses sorresponding to the weld overlayed confisaration. The ITB-3640 margias on limit loads are maintained at levels sorresponding to minimum Code values.

Sinoe the amoracked ligament is not seasidered, the veld overlay thiskaess is independent of both the UT measured length and depth of the original ersok.

Thus, sacertalaties on UT sizing sapabilities do not affest the design.

Crasklas on welds 10-0-10 and 10-0-15 was shallow. With somservative depth catinates it was shova that operation 'as-is' without repair was asseptable, l

ovos af ter assonating for IGSCC growth. Iowever, a weld overlay of 0.25 lash thiekness was applied to provide additional margia.

In fact, based on the cyplied stresses, the 0.25 inch thisk overlay is for all practical purposes a i

fall streetaral overlay.

I l

l Overlays os riser volds have been applied on asial erseks la weld heat effected zones (EAZ). Aslal eraaks la general are short sinee the veld EAZ is limited to a marrow aironsferential band near the veld. Welds with axial f

creeks of this length saa be ahora to retain asseptable design margia eyes if they are assumed to be through-wall. Thus, for axial eracks, seguired strastaral design marglas saa be malatained regardless of the erask depth casse the length is takerently limited by the RAZ width.

For asial erseks, I

overlays (0.25 ineh thiek) were applied mainly to prevent leakage from through-oc11 or nearly thressh-wall erseks.

[f.

I I,

i

~ - - -

-e--

. - - - - - - - -,., - - - ~ ~, - - -

~ - - -, -. - - - ~ ~ -.

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

m -w

-w w r,w -

L e# -g In sedition, a weld overlay of 0.125 inch thickness (ezoladha the first layer) was applied on the 4-inch diameter j et pump instrumentation mossle/ safe cad weld. A sembination of eiroamferential and asial erasking was detooted in this weld.

Fatiane. Then a weld overlay is applied on a piping component with a orack, i

IGSCC erack growth is not espected to be signifloast since the veld overlay prodsees compressive residual stresses on the inside surfeos and la the inner i

portion of the wall of the pipe, thereby preventing new IGSCC initiation or retarding existing IGSCC erack growth. his subject is disonssed in detail in Section 5.

Patigue orack growth saa generally be shora to be small since the roeiroslation and RER piping where overlays have been applied do not omperiemos significant fatisme eyeling.

An alternate approsek which aan be used to demonstrate fatigue margia is to l

perform a Section III f atisse analysis seasidering orack laitiation from the IGSCC defest.

In asoorjasse with ASME Code practice, a masimum fatisse strength redsstion f actor of 3 is used and the fatisse usage is shows to be l

1ess than 1.0.

l l

2.1.2 Marsia to Collapse or Fracture he methodology used la developing the ITB-3640 a'eseptamos flaw size is to first determine the eritical fisw eine for the applied loading conditions.

he asseptamoe fase size is them determined by reguirlag a saltable design cargia on the critical flew oosditions. he eritical fisw sine is determined

'asind limit-load eencepts.

It is assumed that a pipe'with a circumferential creek is at the point of insipient failure when the met sostion at the orack develops a plastle hinge. Plastic flow is assumed to ocear at a eritical stress level, eg, called the flow stress of the material. h e eriterios is simple to apply and has been shows to be effective is prediotlag fallare of stalaless steel pipes sentaining aircumferential eraeks (Referesses 3, 4).

Using this method, eritical flew parameters saa be represented in the form of a fallare diagram defining the sonbination of eritical erask depth and length 3

a at which sollapse occurs for a given applied stress. Figure 1 shows a typical failure diagram defimias the flaw parameters at fallare for stress level P, + Pb = 5,.

The allowable flaw sine is determined by applying a safety feetor of 2.77 on the failure line. This is comparable to the minimum ASME Code design margia in ascracked piping (Referomoe 5) subjoet to nazimas c11ovable loadings.

The IWB-3640 flaw assessment procedure seed la the veld overlay design caplicitly assures adequate margias against sollapse but implicitly assumes carcia om f ailure by tearing instability. The toughness of the anstomitis etainless base material is high omough to assare that the asseptamos fles sizes la ITB-3640 also provide ognivalent margins on tearing instability.

Simoe most IGSCC eracks are la the base metal near the veld, the fracture properties in the immediate region of eracklag are adeguate. Bowever, resent 3

data os stainless steel weldaents have shown that the toughness of some heats cf ashmerged are wolheats (SAW) saa be significantly lower than that of base total. Parametrio studies using the lower bosad SAW data have shown that the factor of safety on tearing instability saa be lower than that implied by the Code (approximately 2.0 compared to the Code value of 2.77).

Data from the l

name programs, however, show that GTAN weldmaats have substantially higher l

tosshness than the SAW and are therefore not substantially affected by these cold metal toughness concerns.

Simoe veld overlays om ersoked piplag are made using GTAN weiheat, the design based on 1Y8-3640 as described above also provides adequate margia on tearing instability. Detailed fracture ovaluations of the overlays are presented in Section 6.

l 6

3.

PEA 3 30 RID 3 UNIT 3 TELD DVERLAY CONFIGURATIONS Of the total of 15 weld overlays on piping,10 are loested on the roeirealation risers, one om each riser. These were designed as repair measures for the axial flaws wklok were senservatively assumed as through-

)

call. The maximum axial flaw length was 1.0 insk. The weld overlay thioksess uns 0.25 Lash. As noted la section 2.1, the required strastaral margia saa be caintained even if a 1.0 insk long throssh-wall saisi erack is present. Thus the mais faaetion of the riser overlays was to prevent leakage from through-call or essentially through-wall eracks.

The overlays os volds 10-0-05,10-0-06, and 10M7, located on the 20-inch RER smotion piping were also sized without taking eredit for the amersoked ligament in the original pipe. Thus, the IVB-3640 requirements were satisfied casualms a 360* siroaaforential erack of depth ogual to the original pipe thiokse s s.

Thus uncertainties la orack sizing by BT did not affect the weld overlay design. The overlay thioksess at welds 10-0-05 and 10-0-06 was 0.5 inch and that at veld 10-0-07 was 0.35 inch. The eraoking on welds 10-0-10 and 10-0-15 was shallow and shown to be acceptable for operation without repair.

Nevertheless, a 0.25 inch thick overlay was applied on these welds.

This was very slo'se to the thioksess (0.280 insk) required for a fall structural overlay.

l Flaally, is11 strastaral overlays were also applied on the two jet pump instrumentation mossle/ safe end welds.

C a

d i

4 e

7

4.

RESIDUAL SIRESS IMPROVEMENT One of the essential factors la IGSCC initiation is the presease of tensile ctress in excess of the local yield stress.

In the as-welded somdition, the predicted residual stresses on the inside surf ace of a pipe are highly

'r, tensile. Both the IBSI and weld overlays prodsee compressive residual stress et the inside surface of the pipe, thus promoting erack growth retardation or carest. As indicated la Table I, 91 welds la the Peach Bottom 3 roeirealation cad RER 11 ass have been treated with IISI, est of which 15 welds were cabsequently wold overlayed. This section presents the teshaisal support information on the benefiolal residual stresses indseed by IRSI and weld overlay processes.

4.1 Indsetion Beatian Stress Imorovenest (IRSI) a In this process, compressive residual stresses at the inside wall surface (and esbaarface) are produced by induction heating the outer surface of a pipe

{

colheat while sinaltassoasly sooling the inside surface with water [Referesse 7]. Analytical predictions of residual stresses from IISI have been presented la Referesse S.

I A similar analysis has been presented la Referesse 9 for operating plant pipe welds in whiek undetected 105CC erack may already exist. Figure 2 from Referesse 9 shows the predicted IRSI ladsood residsal stress patterm for both an ameraaked pipe and for pipes coatsialas siroamferentisi erseks of depths ogsal to 6% to 40% of thickness.

It is seem that compressive residual stresses are still prodsood, even in the presease of moderately deep eracks.

4.2 Residual stresses Rossitina from Teld Overlava Cath analytical and experimental results on the weld overlay ladseed resideal ctress patterms are disasssed.

9 l

l l

l 8

l

---,---...m_,

m-,

,,-wy-w

---..~----v=--

~ ~ - - - - - * - - * - - - ' - - " ' ' - - - ' - ' - - * " - - - ' "

t J

4.2.1 Analvtical Fredictica of Residsal Stress Distributions Rossitian From fold Overlav Renairs i

The analytical procedazo is essentially similar to that for IESI. h e only

- differesse is in the temperature analysis where heat flow salesistions are now performed either by mains a moving point soares or by ausset area heating method. Figure 3 from Referesse 1 presents a typical salesisted axial stress distributlos thronsh the wall of a 12-inch pipe in the heat affected some, 1

follenlag the application of a 0.25 inch veld everlay with a heat lapst of j

approximately 25 ET/la.

l 4.2.2 Ennerimental Bases for Told Overlav Residual stress. Estensive residual stress meassroments have been made with veld everlays made salag typical field practies. D ese meassroments laslade j

g o Measurements done at Argosas National laboratcry om everlays simulatlas the geometry of the Estok 1 and Eatsk 2 piping (Figures 4, 5).

y..

o Ep11/BWR Owaars' Grosp program on large diameter pipe overlays done at the J. A. Jones Applied Researek Center.

I o GB data om veld overlayed 16-inch diameter pipe.

1 f

the rossits of the different programs somfism that veld everlays prodsee sempressive residsal stresses on the ID surface and throssh a substantial i

portion of the laser pipe wall. D e magnitude and distribution of the I

residual stress is dependent es the speelfie details of the velding process,-

tumber of layers, heat lapst, and pipe thiekness.

fie offset of the favorable residual stress is to prodsee a negative (sempressive) stress latensity factor chich Aahibits ersek growth.

he benefielal effest of the compressive stress cypties even for reistively deep eracks.

Therefore, la addittoa to the streetaral reinfersement from the everlay, there is also addittomal benefit from the residsel stress.

a In overlays which also consider the original pipe wall (the so-ealled mini-overley), the benefit of the compressive stress is laciuded la the orack growth saleslations and subsegssat decisions on ersek asseptability. Iowever, la a is11 structural overlay, the sizing already assumes a 3608 through-wall orack in the orisiaal pipe wall. Therefore, the overlay design does met sonat em the benefit of the residual stresses in limiting erack growth. This is important, since differesses la the welding parameters or other overlay application variables are not significant when the benefit of residual stress is not insisdod la the design basis.

i i

j l

k l

i 10 e

9 i

e noee a e

~+ e

,we-.---------=-,,*-we-

,.mm.

.------.-------.m

- ~ -

-e--,.---,.-,--.---...--.-,-_s

4 a

O 5.

CORROSION CRACKING RESISTAN G OF WELD MATERIAL he application of duplex stainless steel wold overlay to austenitis stainless 4

steel pipe joints results la the deposition of several layers of weld metal over sa existing girth veld containing a defeet. De veld overlay material is selosted to be highly resistaat to IGSCC in 351 envirossents.

In the following section, the eriteria used to seleet the weld overlay material are disenssed, based upon the inherent IGSCC resistamos of sustenitie base j

nieredsplex stainless steels.

l 5.1 IGSCC Resistance of Wald Metal he auseeptibility of austenitio stsialess steels to IGSCC in oxidising I

envirouseats is understood to be the result of a reductiet in the level of shromium at grain boundaries to below approximately 12 wit, the level which somfors passivity spon stainless steels in these environments. his 'shromium depletion' is generally the result of thermal or thermo-mechanical prosessing j

(such as welding) which sauses the material to precipitate shromium-earbides i

at these sentimsons austenitic grain boundaries thereby ' depleting' the I

adjasest matrix material la chromium. Tso straightforward approaches which saa effectively prevent this ehromias depletion at grain bosadaries resulting from the thermo-meskamical processing are (1) imoreasing the shromium level in the alley and (2) redaeing the earbon level in the alloy.

In general, the IGSCC resistaneo of misredsplex stainless steels of the Type

{

308 or 316 stainless stool types (or interdendritis stress sorrosion erasking resistasse) is derived from the fact that these steels sontain ferrito, a phase rish in ehromium. The ehromium level in this' phase is generally of the

(

order of 25 wt%, while la the eastsatte phase the ehromium level is typically of the order of 16-18 wt%. Unlike the fully anstenitie Types 304 or 316

}

stainless stool pipe in which the shromium is depleted from the austoalte gasta boundary vielaity, la the microdsples stainless steels the latergranular ehromium earbido proelpitation eseurs predominantly along austeatte-ferrite grain boundaries darias thermal prosesslag. The ferrite grain typically provides the additional shromium required for the precipitation sisee this i

l l

11

g I

N x

(

phase he s cap significantly higher compositica of thromian them austenite and the skromium dif(asivity at 1100*F is approximately three or&ers of msgsitada, greaterlaferriththanin'austomite. A small anoast of proeipitation may cesar from the austenite" side'of tha f'errite-asstenite grain bosadary, thus

.~ rdsalting in some 'anall smosat of sensitization. The potential for

'n sensitization and susceptibility tog sterdendritto stress oor:6sion erse'<1ag i

cre dependent primarily on the earbon oostent and the ferrite eostent of the wols metal. These two factors are interdependent since ferrite sostent

, 3::mera11y amoreases'as the earbon oostemt is rednoed. Thus, both factors fcwor IGSCC resintance. A member of investigations, laboratory tests and field servios evaluations have been performed and used to establish the basis 7

,'3' af requirossats.f or earbon level and ferrite levels of weld overlay materials if that will provide 'the required IGSCC resistamos of the veld overlay buildsp.

This work will be disonssed below.

)

5.2 Carbon and Ferrite Levels Reas(red to Provide IGSCC Resistance I

i s

s In the extensive pipe test programs ssed to evaluate the behavior of' Type 304 M

atalalese steel a.s well as the dif ferent mitigation techniques (suck as IRSI, and the alternate materials 316Mi and 3'C4NS),

' LFEST, solution heat treatment, the weld metal used to join the test pieces was made cit Type 308L with 0.035 i

l peroomt maximum earbon and a minimum of 8 Ferrite Number (FN). The weld metal in those tests was always resistant to IGSCC even with the high residual stresses which generally peak near the weld senter11as.

Secondly, la tests on

^

shop applied sorrosion resistant e1 adding (CRC) with maximum 0.0350 earbon and claimar, ferrite 8 FN, no IGSCC was observed (Referemos 10). The IGSCC s

l resistasse of weld metal is sonsistent with laboratory studies perf.used by l

Devine (Reference 10). He evaluated the infisease of, earboa level, ferrite p

sad ferrite distribution in duplex stalaloss steel alloys and weld

{

content, l

cetals on IGSCC resistance in oopper-sopper sulphate solution. Eis results l

seafism that if the earboa level is less than 0.015E IGSCC imanalty was ossared and that Type 308L with 8 FN weld metal would also be espected to be I

- sesistsat to IGSCC.

i l

11 o

I

p Laboratory experience also has established that veld metal with low carbon and i

l 51 < :

high ferrite sostent will arrest propagating intergramalar stress corrosion

[

eraeks as well. As part of as EPRI/ General Electric study lavestigating erack growth rates in Type 304 55 pipe material, preersoked fracture moohanies

. specimens costalaims welds of Type 308 or 308L stainless steel with different l,

ferrite levels were tested la a laboratory simulation of the BTR environment under slow syclio loading (Referomoe 11). The high earbon Type 308 SS welds exhibited significant intergrasslar penetration into the veld metal at ferrite levels up to 3.5%.

Arrest was observed at a ferrite level of 11.5%.

Bovever,

\\

the Type 308L SS welds, fabricated from 0.025 wtt earbon material and containing from 5.5 to 11.5% ferrite always exhibited arrest of the IGSCC cracks, which had initiated in the adjaoemt sensitized host affected some.

The intergramalar branches of the primary crack contiated la the wrought, sensitised Type 304 SS along the bass metal weld heat affected some parallei to the weld / base metal interface, demonstrating that the IGSCC mechanism coatissed to be active la the specimea.

In another RPRI-sponsored study, lavestigators at Ishikavajima-Barina Beavy Industries (IBI) fabricated and tested girth welded Type 304 SS pipe processed to prodsoo nearly through-wall IGSCC (Reference 12). One intergramalar erack penetrated the weld metal and extended several millimeters into the weld. The weld metal erack which penetrated la high earboa Type 308 SS of approximately

'n 55 ferrite appeared to terminate where the ferrite level had imoreased to approximately 95 ferrite.

l l

One additional study sponsored by EPRI and performed by General Blootric l

provided striklag evidemos of the ability of microdsplex stainless steel weld metal to resist IGSCC in BTR-like environments (Refeiemos 13). In that study, fatigue procracked, plate welded fracture mechanies specimens were bolt loaded sad tested for an extended period of time la simulated BTR water. All of the austealtic materials, isolading Types 304M and '316M stainless steel l

exhibited significant intergramalar erack growth in this very severe test.

In two asaples where the f atigue preorack inadvertently terminated la the weld metal, ao erack extension was observed in the weld. Type 308L stalaless steel l

l sostaining SPN minimum and 0.035 wt% earboa nazimum was specified for these tests.

13

/

j Finally, in tests run as part of the degraded pipe study, Type 308L high ferrite (8 PN minimum) weld overlays demonstrated complete ersok arrest at the f

weld interface (Referemos 14).

he field data is similar to the laboratory data.

In general, all weld metals, whether Type 308 or 308L, have exhibited excellent IGSCC resistance.

here are only a few incidemoes where grorlag IGSCC oracks have penetrated low ferrite high earbon veld metals.

In all cases the cracks appeared to arrest chen the ferrite level increased to - 65 In most esses arrest oosarred la weld metal with a ferrite content as low as 35.

Details of the field data are described here.

Evaluation of the IGSCC failure at ERB showed that the IGSCC always appeared to arrest at ferrite levels of 35 when the cracks penetrated the weldsent (Reference 15). A second instance of weld metal eracking was in a a sostion cf the sore spray line at Quad Cities Unit 2 (Reference 16). Metallurgical analysis of the weld revealed the cracking to be interdendritic in 0.064%

aarbon Type 308 stainless steel contalaims approximately 55 ferrite. Finally, ct Nine Nile Point Unit 1 selected metallurgical analyses were performed on pipe specimens removed from the recirculation system (Reference 17). The caviroasentally assisted cracking had grown through the base metal (Type 316 SS) isto the wold metal in tvo of the pipe joints. De erasking la the wold cetal appeared to be interdendritic, although the photomicrographs provided did not sentain sufficient detail to conclusively document that observation.

(

The veld filler la believed to have been Type 308 or 316 stainless steel and was deternimod to sostain between 35 and 65 ferrite. This result is similar to the GB/RPRI results.

All these field observations provide clear evidence that in instances of weld cetal IGSCC erasking, there was either high carbon or lor ferrite sostemt or a oombination of both.

3 14

l 0

5.3 First Laver Teld Overlav Dilution Due to weld penetration and mixing with the pipe base material, some redsotion la the ferrite level and imorosse in carbon oostent will oosar in the first Icyer. Basept for very localized microstrsotural and fusion line diffusion cffoots, each overlay weld layer is a homogeneous fased structure of salform

)

carbon and ferrite composition. The autossatic GTAT process applied to weld overlay will rossit in lase metal to total fused weld metal dilstlos of 20-305. This low level of dilution vos1d still prodsos a first layer overlay that vos1d be highly resistaat to IGSCC when a 308L high ferrite veld matorist I

is used for the overlay. The second or subsequent layers are essentially nadilsted, so that the bulk overlay material has IGSCC resistamos equivalent to that of the veld material. This is confirmed by the rossits of the GE/EPRI degraded pipe program, Referomoe 14, (described in Section 7) where orack arrest was observed at the base metal interface with overlays made of 308L SS with 8 PN ferrite content (Figure 17).

5.4 Peach Botton Unit 3 Overlav Material Soesification Because of the importance of ferrite to the IGSCC resistance of 308L material, GE specifications require ferrite detezzination using onlibrated magnetic ccassroment of weld deposits for each heat of material. This technigne has been fosad to be more reliable and conservative than ehemical composition (Schaffler or DeLong diagrams) or microstructural meassroments of ferrite.

Magnetic measurement using the actual weld deposit eliminates potential errors due to estimated weld material altrogen oostent which must be assumed for the ehemical analysis methods.

Using the magnetic measurement procedsre, a alminum ferrite content of 5.0 PN was applied for the Peach Bottom weld overlay welding material.

Carbon content requirements were in accordamos with ASIE SFA S.9 requirements for Type BR308L. This welding material specification requirement provides a weld overlay (including the first layer) that is highly resistant to IGSCC. The actual heats of material applied at Peach Bottom 3 were typically 0.0205 maximum earbon and 10-12 PN, providing even farther ISSCC margia for the overlay.

i 15 w

g

,-----w---

as-. - -_,+-..

yn l

6.

FRACITRE ECHANICS MARGIN i-Told overlay designs for the oracks at the riser and RER welds, and the Jet

- Pump instrumentation mossle at Peach Bottom Unit 3 were based om Paragraph ITB-

{'

t 3640 of Section II, ASE Code. The allowable flaw size tables in ITB-3640 are based on the met section oo11 apse theory. With implied safety margin of 2.8 frr normal (Level A) and upset (Level B) somditions and 1.4 for emergency g.

(Level C) and faulted (Level D) somditions.

M, Maximum load and allowable flaw size predictions of this theory have been shown to be la good agreement with both the esperimental results (Referesses 13 and 18) and the salesistions using the carrently available elastle-plastic fracture mechanies (EPFM) techniques (Reference 19). Since most of the attention was directed toward stainless steel pipes with IGSCC, the caperimental studies were focused om eracks la the heat affected some (RAZ) or the base material. EPFM evaluations of the tests used the base metal fracture toughne ss properties. The implied assimiption was that the stainless steel veld metal in all esses vos1d have essentially the same toughness as the base cetal and that due to its ferrite sostemt requirements, the IGSCC eracks are salikely to propagate in the weld metal. Bowever, ressatly available tosshness data on the staialoss steel welds seem to suggest that the welds produced by some processes such as the submerged are welding (SAT) may have considerably lower toughness than that of the bas's metal or the RAZ. On the YM sther hand, the wolds prodsood by TIG, the process used la the veld overlays, ahows sonsiderably higher tosshness.

1 l

The fracture mechanics analysis presented la this sociion addresses these weld tosshness comoeras as they relate to Peach Botton Unit 3.

It is shown that l

the safety margias implied in ITB-3640 are exceeded even when the low weld tosshness properties are factored in. The evaluation was sondssted using EPFM technigues.

The 1rold overlay tosshness was skaracterised in terms of a lower bound J-resistamos surve based on a review of the available teobaioal literature.

16

/

F:r esok distimot overlay somfiguration, an allowable stress magnitude was salsalated sorresponding to the elastle-plastic instability. The design or tho metaal primary stress, F, + P, at the same loostion was obtained from the b

overlay design report. The ratio of these two stresses determlaes the factor

- cf safety based om RPFM saloslations. These are compared with the ITB-3640 fessors of safety.

6.1 Wald Tosahness Data Evaluation Resent data from Referomoes 20 through 23 indicate that in some cases the stainless steel weld metal tosshness may be lower than the base material. The differosos appears to be a strong inaction of the welding process. The welds pradsoed by the SAW process appear to show the lowest toughassa. On the other h=d, the welds prodsood by TIG, the process used in veld overlays, shows sensiderably higher tosshness, almost approaching that of the base metal.

Figure 6'shows somservative representations of the (J,,g, T, g) earves for two types of wolds.

Carve 1 from Reference 23 is for a submerged are weld. Carve 2 caso from Reference 23 is for gas tsagsten are weld (GTAW) siellar to the overlay welds.

Curves 1 and 2 may not be the absolute lower bosads for the espective veld eategories, but were considered as representative lower bosada cxd were used in the EPFM evaluation.

6.2 BPFM Calosistion Nethodoloav and Parameters The J-integral and the applied tearing modulas, T, were evaluated as a fanation of applied loading using the estimation scheme procedure given la l

Elastle-Plastic Fracture Handbook (Referesco 24). The latorsection of (Japplied,Tapplied) earve and the appropriate (J,,g,(g) earve gave the valse cf J-integral at instability. The applied stress at instability was determined sorresponding to the J value (see schematic la Figure 7).

A key input in the evalsation of J,, and T,, is the Ramberg-Osgood skarac-terisation of the material stress-strain behavior:

O a

i

.. ~. - -

a a

E = (E-) + a (E-)

o o

o

• n s.

o a

There e,, e, a are parametera determined by fitting the equation to the true stress-strala earve.

i Yariations la the stress-strala behavior of austenitio stainless stools may oscar due to steh factors as welding process, heat impat rate, filler metal composition, thermal boundary conditions, etc.

Sinos a detailed study of these aspects was beyond the scope of this report, the evaluations were t

l performed for one set of these parameters.

He fo11orlag Ramberg-Osgood parameters from Reference 23 based on an emperimentally-determined tras stress-tras strala carve of a GTAT weldsent were used:

1 l

l s = 2.83, m = 11.84, e,= $3,900 psi i

l 6.3 BRD Wald Overlav Evalmation no RER supply line veld overlays 10-0-05,10-0-06, and 10-0-07 are full struotaral type; i.e.,

the overlays were sized without taking eredit for the amorsoked portion of the original pipe. he ITB-3640 requirements were satisfied assualms a 360* earcumferential orack of depth equal to the original pipe thickasas. De orerlay thicknas: at uolds 10-0-05 and 10-0-06 is 0.5 inch and that at veld 10-0-07 is 0.35 isok.

For the assumed 360' erack geometry, the J-integral and T,, values were salenlated using the fo11erlag estimation schose formulas:

J=fg (a,, E /1,)

+ae,s,ehh

[

g y

o

1. L M

2

• da o

18

I e

he estimation scheme parameters for bending loading are earrently unavailable and, therefore, only the pare tension loading ease was evaluated.

Figure 8 shows a plot of J-integral versus applied axial stress for overlay

- oomfigurations on weld 10-0-05.

Sinos the orack tip for this asse vos1d cdvance essentially in a GTAT weld metal, GTAT J-T (Carve 2) was used to i

determine the value of J-integral at instability. The instability stress was l.'

saloslated as 31.1 kai and is shown la Table 2 along with the notaal primary The membrane (P,) and bending (P ) stress based on the piping stress report.

b (P, + P ) stress at veld 10-0-05 is 5.96 kai and, thus, the salculated f actor b

of safety is 31.1/5.96 or 5.2.

Results of similar saiculations for overlays j,

on welds 10-0-06 and 10-0-07 are shown la Table 2.

[',W Indications in veld 10-0-10 and 10-0-15 were relatively shallow and were asseptable as is.

Nevertheless, a 0.25 inch thick overlay was applied. The

{,

fracture margia evaluation for velds 10-0-10 and 10+15 was sonducted based l

I on the following conservative assumptions:

(1)

The flaw depth was taken as two times that reported by UT laspection (ii)

Submerged arc (J-T) properties, Carve 1, was used in the lastability stress evaluation.

(iii) The reinforcing benefit of the 0.25 imah thick overlayed was senservatively neglected.

The results for these two welds and the sorresponding factor of safety are f

also listed in Table 2.

1 l-6.4 Riser Wald Overlavs All of the UT-detooted eracks near the riser welds in the Peach Botton Unit 3 roeirealation line were short (G inch) and were oriented in the axial l

direction. The applied minimum weld overlay thickness was 0.25 insk. This overlay thieksess was based on the conservative asseption that the orack

(

depth is equal to the original pipe thioksess.

Figure 9 shows the dimensions 1

of this overlay.

19

-c.

- Currently, the J-integral estimation scheme formulas for part-through wall axial flaws in sylinders are not available and, therefore, the RPFM ealesistions were performed assanias a through-wall flaw. he J-integral was estimated salas the solution for a orack la an inflaite plate by Shik and Estekinsom [25] and a sarvature ooriestion f actor from Referomoe [26]:

2 2

= a[1 + f (

)(f) ](f) a,s,a o

o I

a+1

+ a [3.85 m (1 )+I](f) for e i e, o

= a[1 + 2 (el)](E-)2 l

a, a+1 m(1-f)+I](f) for a J,e,

+ a [3.85 a

where 22 = (1 + 1.25 h )

for h 1 1.0 2

2

= (0.6 + 0.9 h ) for 1.0 I h 1 5.0 Pisure 10 shows a plot of J-latogral as a function of acminal hoop stress.

he instability hoop stress was determined to be ~52 ksi, he nominal hoop ctress in the riser pipes for the Level A and B sonditions is 10.9 kai. he corresponding factor of safety is 4.8 and is shown la Table 2.

6.5 Discussion on Frsetare Marains he calculated factora of safety in Table 2 for the riser and RER weld overlays are well in excess of the implied safety margias in ITB-3640. Even higher margias are espected for the jet pump instrumentation mozzle overlays I

sisee the applied loads on the nozzle are low. Based on these results, it is ooselnded the Code-implied safety margins are maintained at the subject l

loestions in the Peach Botton Unit 3 reciroslation and RER 11 ass, even when the, lower bound material toughness properties are factored in.

i 20 dp

'x

/

7.

RESULTS PROM DEGRADED FIPE PROGRAM Eder EPRI sponsorship, GE is conducting a test program om degraded piping ccbjected to remedies like IRSI or veld overlay. The purpose of this program is to provide the experimental basis to define the design life of the remedies. The initial goal was to confirm weld overlay life for 1-2 eycles.

Rosalts of this test program to date are described here.

The first result om veld overlayed piping was on a 4-inch Schedule 80 pipe cath a fall structural overlay. The veld overlay specified 308L stalalors steel material with ferrito number 8 FN. The specimens were preersoked by IGSCC mechanism sad were subsequently overlayed. Testias was at a acminal stress level of 16.9 ksi (8,) in 8 ppe oxygenated water at $50*F.

To obtala an early assessment of life for the veld overlay, specimen ESP-14 was removed l

citer 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> esposare in the test environment to perform a destructive ovaluation of erack growth under the overlay. Told 'D' of the specimen was ekosen for ovaluation. EAZ 'D-2' had a through-wall erack, which sansed a small blowhole when penetrated by the veld overlay and required a repair Since procedure similar to procedures used in the field for this oscarremos.

this orack was of known depth, terminating at the veld overlay interface, it efforded an opportunity of ovaluating erack growth sador the overlay.

i Since the weld overlay provides f all structural reinforcement of the pipe, the i

primary conceras to be resolved were:

(1) whether the crack would penetrate

(

through the veld overlay, and (3) whether the orack would extead in a sireamferential direetion.

l A dye penetraat osamination was made on the ID of tho' pipe at veld

'D' to determine the aircumferential length of the orack and compare this to the length of the orack before the remedy was applied as determined by UT l

osamination. No PT indication oos1d be obtained. This sould be due to the high compressive stress on the ID surface which eassed the ersek to close up cad prevent absorption of the dye.

21

l

' The siroaaferentisi length of one of the ersoks was measured af ter bending the pipe wall to open up the eraok.

Comparison of this length to the original length as measured by UT before overlay application showed no growth.

~

Stro longitadiaal metallographic pipe sections were made through veld RAZ

'D' ct two different locations where blowholes osourred when penetrated by the veld overlay. Both of these sostions showed complete arrest of the crack at the veld overlay metal interface.

Figure 11 shows metallography of the orack costion oomfisming erack arrest. Based on the observation of no messarable crack growth is any direction, a lifetime of several fuel eyeles saa be predicted. This provides definitive proof, under sinalated field conditions, f.',

that the veld overlay provides orack arrest even ander 8 pga ozygensted water

{'

ot 550*F.

Exact determination of the factor of improvement is not possible sinos the stress and environment in the pipe test were mora severe than that i

W.

in the field.

One measure of the improvement may be dedsood from the fast that the average time to f ailure in pipe testa on manitigated 304 stainless

'd steel welds in 8 pga osygenated water is 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> (compared to erack initiation time in the field of 2-3 years). The fast that no erack growth

(

socarred in 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> suggests that the wold overlays should be good for D

several fuel eyoles.

Similar results were also obtained on preeraoked IRSI pipe tests.

Ssooessful cperation of up to 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> has been confirmed. This shows that la welds with IRSI, even if undstooted oracklag exists, erack propagation would not cesar and the benefit of IRSI would be malatained.

l l

l l

f(

i 22

8.

IRSI REAIED WELDED JOINTS WIIBOUT MACKING Simoe IRSI la ocasidered am IGSCC mitigating measure, a reexamination of these welds may not be necessary. As an alternative, a sample of such welds may be

• reezamine d.

The membez of welds to be inspected and the fregnamoy of inspection will depend as the regulatory requirements at the time of inspection.

O e

23 w--

---w

-<:a

.,---__-w~.

?

9.

CONQ.USIONS The report presents the technical basis to justify continued operation of Peach Botton Unit 3 recirculation and RER 11 ass for more than one smal cycle.

Various elements of the technical justification include:

Structural margin ovaluation and disonssion of IGSCC resistanse of weld s.

overlays.

b.

Residual stress improvement due to IHSI and weld overlay prosesses.

Results of GE experimental programs for veld overlays.

o.

Based on the results presented here, sostianed operation with overlays can be extended well la excess of two fuel eyeles.

{

l 1

9 o

I l

34

- -, n r.

--,r--

10.

REFERENCBS 1.

' Continued Servlee Justification for Veld Overlay Pipe Repairs,' report prepared by RPRI, GE. Nutsch, SIA for BWR Owners Group and RPRI, May 1984.

2.

ASME Boiler an.1 Pressure Yessel Code,Section II, Division 1, Nuclear Power Plant Composants, Americam Society of Mechanical Ragineers,1983 e

Edition (Winter 1983 Addendam).

3.

' Mechanical Fracture Predictions for Sensitised Stainless Steel Piping with Circumferential Cracks,' RPRI Report NP-192, Flaal Report, Electric Power Research Institute, Palo Alto, CA, September 1976.

1.

t ',

4.

' Review and Assessment of Research Relevant to Design Aspects of Nuclear Power Plant Piping Systems,' NUREG-0307, Nuclear Regulatory Commission, Tashington, DC, July 1977.

W j

5.

Ranganath, S. and Mehta, R.

S., ' Engineering Methods for the Assessment

{

of Dsctile Fracture Marsia in Naclear Power Plant Piping,' Riastic-Plastic Fracture:

Second Svanosium. Volume II-Fracture Re sistance j

Carve s and Enalmeerina Annlications. ASTM STP 303. C. F. Shik, and J. P.

h. _

[

Gadas, Eds., American Society for Testing and Materials,1983, pp. II-309 i

to 11-33 0.

6.

' Inspection of BTR Stainless Steel Piping,' USNRC Generic Letter 84-11, from Eisamhat to BTR Licensees of Operating Reactors, etc., April 1984.

7.

' Induction Beating Stress Improvement,' EPRI Report No. NP-3375, November 1983, prepared by General Electric Company, San Jose, CA.

8.

Eerrera, M. L., Lange, C. R., and Ranganath, S.,

'Amalytteal Evaluation of Residual Stresses in Piping Subjected to Induction Eesting Stress Improvement Process and Application to Operating Plants,' presented at the 1981 Pressure Yessel and Piping Technology Confereacc held June 21-

' tw 25, 1981 in Denver, Colorado, Paper No. 81-PVP-19.

25

a 9.

Eerrera, M.

L.,

Mehta, E.

S., and Ranganath, 8., ' Residual Stress Analysis of Piplag with Pre-Existing Cracks Subj ected to the Induction Beating Stress Improvement Treatment, ' ASIE! Paper No. 82-FYP-60.

10.

' Evaluation of Near-Tern BTR Piping Remedies,

• EPRI NP-1222, Vol.1, 2, 3, November 1979.

11.

'The Growth and Stability of Stress Corrosion Cracks in Large Dismeter i

BTR Fiping,

• EPRI NP-2472, Vols.1, 2, July 1932.

I 12.

' Assessment of the Feasibility of Producing Pipe Samples with Tight Through-Tall IGSCC, ' EPRI NP-2241-LD, February 1982.

13.

' Alternative Alloys for BTR Pipe Applications,' EPRI NP-2671-LD, October 1982.

14.

A. E. Pickett, ' Assessment of Remedies for Degraded Piplag,' EPRI/GB Program Progress Report T301-02 (to be published).

15.

J. C. Catt, private somaanisation.

i 16.

' Weld Metal Cracklag la Nine Mile Point Unit 1 Recirculation Piping Joints,' Letter, R. E. Smith to D. Norris (RPRI), February 23, 1984.

I

17. Analysis of Cracked Core Spray Piping From the Esad Cities Unit 2 Boillag Water Reactor,' D. R. Diercks and S. M. Gaitonde, Materials in l

Naelear Emerav. 1983.

18.

Ianalmen, M.

F.,

et si, ' Mechanical Fracture Fredictions for Sensitised Stalaless Steel Piping with Circumferential Cracks,' Final Report, Elsetrio Power Research Institute, NP-192, September 1975.

i 19.

Ranganath, S. and Mehta, I.

S., 'Eagineerlag Methods for the Assessment of Duet 11e Fracture Marsia la Naalear Power Plant Piping,' Elastle-26 f

p 9.,.

7,

_--9,_

.gy,,-__.g----wm w-9

---m-

-*aw g-3---w----"-"S'--P*"'*--'#""

Plastic Fracture:

Second Svanosium. Volume II-Fracture Resistance Carve s and Ramiseerian Aeolicat ions. ASTM STP-803.1983, pp. 3 09-33 0.

20.

Paris, P. C., Brune tti,' J. V., and Cotter, E. E.,

' The Effect of Large Crack Extensica on the Tearing Resistance of Stainless Stool Piping Materials,' presented at the CSNI Specialist Meeting on Leak-Before-Ereak in Nuclear Reactor Piping System, September 1983, baterey, CA.

21.

Gadas, J. P. and Anderson, D. R., 'J-R Carve Characteristies of Piping Material and Telds,' paper presented at the USNRC 9th Water Reactor Safety Research Information Meeting, Washington, DC, October 1981.

22. Vassilaros, M. and Rays, J.

P.,

'DRTNSRDC, ' presentation at EPRI/NRC Research Coordination Meeting, December 1983.

23. Landes, J.,

Presentation at the Section II Pipe Fisw Evaluation Task Group Meeting, San Antonio, TI, April 1984.

24. Kumar, V., German, M. D., and Shik, C. F., 'An Engineering Approach for Elastic-Plastic Fracture Analysis,' Electric Power Research Institute Report NP-1931, Palo Alto, CA, July 1891.

i 25.

Shik, C. F. and Entchinson, J.

T.,

'Fally Plastic Solutions and Large-Seale fielding Estimates for Plaae Stress Crack Problems,' Transactions of the ASME: Journal of Enmineerina Materisis and Technolomv. Series I, Vol. 98, No. 4, October 1976, pp. 289-295.

26.

Paris, P. C. and Johnson,1.

E.,

'A Method of Application of Elastle-Plastic Fracture Mechanics to Naslear Yessel Analysis,' 51sstle-Plastic Fracture Second Svanosium. Volume II-Fracture Resistsace Carves and and Ramineerina Annlications. ASTM STP 303. C. F. Shik and J. P. Gadas, Eds., Amerloan Soolety for Testing and Materials,1983, Pp. II II-40.

27

--s-r-w------w--

w - - - -

---e

.,.-m..

---

,---------------e-

,w

---

+, - -

Table 1 Peach Bottom 3 EER/Recire'ulation Piping Repair Summary Total number of welds to first isolation 149 (Note 1) valve outside primary containment (Recirculation System and RER Shutdown Cooling Section and Retara Pipias)

Number of weld locations thought not 24 (Note 2) susceptible to IGSCC Number of weld locations susceptible to IGSCC 125 Number of weld locations IRSI treated 91 Number of weld locations where IGSCC was 15 detected Number of weld locations weld overlay repaired 15 c,

2 Number of weld locations with IGSCC and not weld overlay repaired Number of weld locations act exmained and not 17 l

IESI treated Number of weld locations examined but not IESI 17 treated 4

1.

This total does not include 3 assoeptible RfCU weld locations which were not inspected.

2.

4 sweepolets to manifold locations were examined and found acceptable; 4 other sweepolet to manifold v' eld locations were not inspected.

r S

V e

28

4 Table 2 EPPM Based Safety Faetor Evaluation for Teld Overlays Stresses Aceounting i

Pipe Diameter /

Told Overlay for Overlaw Tkloksess EPPN Calonisted Factor of Failure Stress Safety-Told ID Thiekne ss Creek Goemetry Thickness (la.)

P, P

P, + Pb b

I 1

10-0-05 20 la./0.85 is.

.35 la. deep 0.50 4.3 1.66 5.96 31.1 5.2 172*

1 i

16 20 in./0.95 is.

.4 la. deep 0.50 4.03 2.76 6.79 29.2 4.3 1H*

l J

1 3.5 i

10-0-07 20 in./0.95 la.

.35 is, deep 0.35 4.2 2.5 6.7 23.7 1098 2

2 10-0-10 20 la./0.90 la.

.20 la. deep 0.25 5.83 2.2 8.03 38.8 4.8

{

132*

2 10-0-15 20 la./0.95 in.

.3 is, deep 0.25 5.5 2.41 7.91 27.1 3.4 298 10.9 52.0 4.8 j

Riser Yelds 12 1a./0.69 is.

1 lash long, 0.25 10.9 IBased on a 360* eireamferential flaw with a depth equal to the orialsal pipe thickness.

2 Welds 10-0-10 and 10-0-15 eostained shallow flaws and were shown to be aseeptable for eontissed operation and repair.

Nevertheless, am 0.25 inch thick weld overlay was applied to provide additional margia.

Safety factors shown here Lower assumed twise the UT measured fler depth and did not oomsider the additional relaforcement due to the overlay.

l boasd toughness properties oorresponding to submerged are weldments were used.

l l

j t

l

r E

y.

I J

i s

i I

^* a ls u

1

\\

\\

\\

\\

\\

g

\\

..g we u e.

\\,

u s

,~~~~

v.

4 Igu

.s.

s

.j wm ece..um g

mee 6 e u

u

\\

e.. a sr e

i I

e u

u u

u a

w. e.

a v.

e FIGDRE 1.

Detemination of al' ible flaw sizes with a safety factor of 2.773 for normal iditions.

i 30

l 1

N I

ao l

?:

I M

=

ah*0A ah = 0.3s4 J

ds = 8 2 4

40

= =

I 30

!I i

=

' i 10 =

l 2

s 0

e

/

-i0 -

/

/

p=

.se

/

I

/

40

% UNCRACKED

"""*ami.. p 4o _

l dt = 0.064

-80 9

-40 W e = 0.2 4

e f

-N i

/ a/e = 0.4 I

/

-ee -

/

/

f l

e I#'

1/4 THICKNESS 1/2 THICKNESS 3pe THICKNESS eA.

R ADIAL POSITION 9

Figure 2.

IHSI Through-Wall Axial Stress Distribution at Weld Center Line for Preexisting Cracks 31 G

._,_-,--n.

nn.,

y_

m-so-Eo i

i i

i i

e

=

e u

u u

alt e

$y -,o-OJ5" OvtRL AY "E"

Q

• 2M2 IU/IN 18 = 75%

- = -

( t = original well thickness )

r 4

PIPE PIPE 10 00 4

l Figure 3.

Typical Axial Weld Residual Stress Distribution After Overlay h

32

-m._._,,.____._,-.__

j

.ar oviRt.AY WH.D ON 17 SCH.100 MPE amura m u m erens

.rm

,Crw0 m.

m a.

e.

e.

a.

a.

a w

e w

a, m, / =.

a.

a.

m,

~

" ^ * " *

/

Nr

.ma i

f-aause crwi

-- tas*

O g

1

........ m e /

s me

\\

]

a-g -. y

\\

~'***...g l de -

y

-l

, % @w* @

O I

"/ '

g g/

\\

7

\\

\\

/

I 2 q',ja

=

/

4-I e-l

\\

g/

I V

t

\\

.a -

I t.

V l_,,,

%!k..@r

\\\\

%.!W

\\'

g.

~

t NATcM 2 NATtH.i SHORT wel.a PRO tons wel.D map Le 4

i

,S. -

Figure 4, ID Surface Axial Residual Stress 33

,,,.--,--n.v--.--,

,,n,n.-

,,,-,,-er-

P.-

(

l

/

l i

WELD OVERLAY TESTS l

l l

OUTSIDE DIAMETER

^^^ ^

O o.9- -

g g,

o.s - -

O O **

q e

o.7 - -

e 8ee g# #e G

l e **

l t

O 9

o.g-ee l

(

\\

o W*

FAsM Dom o,3 O

AwAL /=estnsWS j g#e L

/

8 99 3

o.2--

e

%e o.s..

.ao do aio do 4o 'do do 1'o o

1'o 2'o 3Io 4'o so' eIo 7'o 30

{

STRESS (Ksil INSIDE DIAMETER I

.20" OVERLAY HATCH-1 (LONG) WELD PREP i

CALCULATED AXIAL RESIDUAL STRESS THROUGH-WAJ.L OF 12" SCH.100 PlPE

~

Figure 5.

Through-Wall Residual Stresses 34 i

L.,

30 D

O

%q 20

~

~

h

• f se

\\

~

g

\\

~

N N

8

\\ s g

N s N 9 10 GTAW Weld (Curve 2) 8 g

k ubmerged Arc Weld (Curve 1) 100 2bo 3b0 40O TEARING MODULUS. T

Figure

6.

• Selected (J,g, Teat) Curves For Welds b

35

• =* -

• +=-oem.

o o

9 1

~

J J

Instability Point I

Instability Jaat.Tsat Stress l

I 1

i I

(JappT,,,)

Stress T

Figure 7.

Illustration of Metho' d For Instability Stress Determination i

I 1

e l

36

~.

o :.

. 6, i

7 6

5

~

w

%n.

14 e

0 ac i.o wy3 4

2 1

t t

e I

e I

a 9

m 10 20 30 40 50 NOMINALSTRESS(ksi)

Figure 8.

J-Integral Versus Axial Stress for Weld 10-0-05 1

37

/ wa.(& Dve.1in

.A.

o a

N O.1.5

/

i 4...

g b

I 1

hA i

a 1

4

+

+ e. fm Pipe Side Elbov Side t=

0.7 in.

t=

0.8 in.

g t=

0.95 in.

t=

1.05 in, a=

0.7 a=

0.8 j, =

= 0.74 0.76

=

FIGURE 9.

Schematic of Riser Axial Flav with Weld Overlay

)

I 38 a---------w.w---

,-w,,wwwy.we,e_

,..,e,e.m-e-_,,-....,e-..,_.e..,

-m m.m -.

7 b

n e

I a.

5 g

I

?v J

4 m:

ec o

J ul l

P 2

3 h

i 2-i l-(

i to so 3o go e

howwA L s7 4tss (wsx)

Figure 10 J-Integral versus nominal Stress for Axial Flaw in 12 Inch Sch. 90 Pipe 39 e*

-v-

- - - - - - ~ ' ' ' * ' ' ' ~ " ' '

rr l

L i

e' t'

j l

8

-] z

.d y

N

. <.fs: %C*

fy'!

?

~;~~

g Y f.f..

z... c; $**.

• f.'l t

. s.

\\. :/;.,:+;,7,p..-

7.

N 5.

f. ' ~11.. ;*?

.f.*. '

g s

.l(,,

y.'~

? Nh N,.

h'.

! " ' 'l..,

r~

g.M.

'g.

. w,- i.

Q d

.x. %y'-

-e.:..

e.,

9-u c.

5

% ~ -'

' 3 'E bc.,,

CP,^

' i)-g.)fj..-

.W u

%..;3,a qd,e g

y,, -

s. s V

-7.:p dw-4

.y

?.

.iv a.

-Y I..

.h' '.

L g

i ;$ {; i g i g # s; ?

200X p.

Figure )),

Crack, Under Weld Overlay, Specimen RSP-14 After 1000 Ifours Exposure i

!l I 40 I

1 t

P, l

---

- -