ML20207J570

Non-proprietary Rev 1 to WCAP-14955, Probabilistic & Economic Evaluation of Rv Closure Head Penetration Integrity for VC Summer Nuclear Plant

ML20207J570
Person / Time
Site:	Summer
Issue date:	10/31/1998
From:	Bamford W, Brad Bishop, Duran J WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17292B802	List: ML20207J566 ML20207J570 NSD-NRC-98-5775, Responds to Request for Withholding Info from Public Disclosure Re VC Summer Nuclear Station.W Concludes That WCAP-14932 Listed Matl Can Be Released as Requested
References
WCAP-14955, WCAP-14955-R01, WCAP-14955-R1, NUDOCS 9903160324
Download: ML20207J570 (71)

Text

,

r, mun m,#

aggl

' ' fR ',,

,,, ' a Qg s

S

_r' h

4

)

y,q.

y?

^

~

.f

, 3 ep l ).

1'

,n v..t o, n 1

'V vs m'

4 9

3 o

t 4;

A

(

h q,, '

a <

}

,d v

t,-

4 w

ajy

. p.

.t l ("

]

4 c

r

,h.W.

h h1;'

c

\\'

)

es ayi

g

+

n '{k, '

py.,aw'! y g

. f " '1 k -lhh h k)g(

.h ed v

~~

y, vn

.u. :,

r "fj k

.f t

l

. A f,\\

~

3 g

l$;w a

s..

w'q 3 a, s

' p5 m

e k '

r.

.. 4

>k 7

r.

b f'f.Jg fl, m L: ; g

%!.pp s?

gy PDR ADOCK 05000395 4

PDR

- b, sf

/

y

~

g y

C

~'

h* ', #

,' $fk J$]W{ Wy 4 904

,f,,

l4,

\\

Lw j.

a,'% % p.

7-.. -.

-~

m..,

'g o

s jj, - E ;.? @-.

,~

j

+

' W~.

P, m

vy.y cy,

~.

~.,.

s., +

s-p" 3.,

a;,.

+,

" LWestin; house NonyProprietap Class 3L

.. q,.

d(

r.., '. '

,4, g,

(

<. 7,?.

i-(..

t

,,,}-

i.

'f If.

w 4

t.,

1

};)!

, ;_'f

'U 1

(, l5

~,y.

l, 8

h.3 4,.,, &..... w..,.. %.s. +4 n

" " ~ "s"

=., t Revision T

. ~

m 1-4:

t

.go t

v y 7,y"

f.^

r

.7 1, :

l

,{_

g p

.a 4

d T.

['

er

,y-m

,.jj i

^

1 i r s r

r

^

\\

5 y.

_.), abilisticiand!Ecoriomic 1

Prob,_..-

r

,s'

/

,,g A valuationtof? Read or e

i MesselsClosurejHead:

Renstra ionilntegrity 6r

~

. ~ ? Virgil!COSummer: Nuclear:

J Planth ',

+

U{,

}

r f_

yj..

"2[

.i '

'x hf L

4y*

x, y,,

m

)(-

+

. w;

,r

' m.,

g,

. Ll.

c.

.q-p,., 4y,.

m,

r i

r u

Y

-.?'

i

}

g y :,

+

.; r i4

+

s g

y 7

r<.

_-y

\\G_

c..~

=.,.:

w.. : :

a w

e jf lti?in g'nLo' u s eJE n e r g y :S y s t e m s.

e W

t i

g y, * ;,

PDP. I603;.'4 99Ci308 t<DOCy 050c;o3q5 I

e V

PI)H

[. M L-fi$

s r

_f

WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-14955 i

Revision 1 Probabilistic and Economic Evaluation of

~

Reactor Vessel Closure Head Penetration Integrity for the Virgil C. Summer Nuclear Plant W. H. Bamford B. A Bishop J. F. Duran October 1998 Work Performed Under Shop Order STGP-105 Prepared by Westinghouse Electric Company for the South Carolina Electric & Gas Company Reviewed by:

0'W G.V.kao' Engineering & Materia s Te&nology Approved: d

/*A 8Vf#

D. M. Trombola, Manager Mechanical Systems Integration

'n Westinghouse Electric Company Nuclear Services Division P.O. Box 355 Pittsburgh, PA 15230-0355 C1998 Westinghouse Electric Company All Rights Reserved

TABLE OF CONTENTS 1.0 Introduction - Summary of Safety Case 1-1

' 2.0 Development of a Crack Growth Rate Model for Alloy 600 Head Penetrations 2-1 3.0 Technical Description of Probabilistic and Economic Decision Models 3-1 4.0 Results of Probabilistic and Economic Decision Models 4-1 l

5.0 Summary / Conclusions 5-1 6.0 References 6-1 Appendix A - Output Files from VHPNPROF program A-1 Appendix B - Output Files from VHPNECON for the Economic Decision Analysis B-1 I

l l

i I

i October 1998

SUMMARY

OF SAFETY CASE Primary Water Stress Corrosion Cracking (PWSCC) in Alloy 600 reactor vessel head penetrations is a relatively new issue in me nuclear industry. The issue was first brought to attention in 1991 when, afte 10 years of operation, a leak was detected during a hydrotest of the reactor coolant system at the Bugey Unit 3 plant in France. Since that time a significant number of research programs have been funded by the industly to determine the causas of the

' probeem and develop strategies for repair and management. Through these studies is was

~

concluded that the reactor vcssel head penetration cracking is a thermally activated stress corrosion process in primary water environments. The process is a slow one that causes no immediate safety concem. Based on conservative evaluation results, the NRC and industry concluded that cracks were most likely to initiate from the inside surface of the penetrations, in the axial direction, and would take at laast six years to propagate through the wall under typical j

plant operating conditions. Fracture mechanics eva!uations have determined that the crack is j

non critical until its axial length reaches 8.5 inches to 20 inches, degnding on plant design.

Extemal circumferential cracking is less probab!e. It may occur only in the presence of an i

above the attachment weld through-wall crack, with active leakage. Assuming coo! ant is present on the outer diameter of the penetration, one conservative analysis estimated that it would take more than 90 years before penetration failure would occur. In the presence of reactor coolant, corrosion of the alloy steel reactor vessel head is possible. Conservative

)

evaluations estimate that it would take longer than six years after a through-wall crack occurs before the ASME Code structural integrity margin for the reactor vessel head would be impacted by corrosion. It was concluded that periodic visual inspection of the reactor vessel head in accordance with Generic Letter 88-05 is adequate and sufficient to detect leakage prior 1

to significant cracking and vessel head corrosion.

Based on the above, evaluations using 10CFR50.59 requirements concluded that head penetration cracking is not an unreviewed safety question.

l On April 1,1997 the NRC issued Generic Letter 97-01, " Degradation of Control Rod Drive Mechanism Nozzle and Other Vessel Closure Penetrations". The purpose of the letter is to request licensees to describe, in writing, their program for ensuring timely inspection of vessel closure head penetrations. This description is to include programs / plans to der.1 with PWSCC of vessel head penetrations and to perform an assessment of any resin bed ingress into the RCS.

The purpose of this report is to provide South Carolina Electric and Gas wit an analytical basis for developing a response to Generic Letter 97-01 relative to PWSCC of the vessel head V

penetrations.

Rev.1 11 October 1998 oMG94non. doc:1b:1000/98 y

v

---y

2.0 DEVELOPMENT OF A CRACK GROWTH RATE MODEL FOR ALLOY 600 HEAD PENETRATIONS Crack growth rate testing has been underway since 1992 to characterize the behavior of head penetration materials. The modified Scott mode!, as described below was initially used for safety evaluation calculations in submittals made in 1992 and 1993. The goal of this work is to review the applicability of that model in light of the past five years of testing, during wnich over

' forty specimens have been tested representing 15 heats of material. The original basis of the model will be reviewed, fellowed by all the available laboratory results, and finally a treatment of the available firsid results.

The effort to develop a reliable crack growth rate model for Alloy 600 began in the Spring of 1992, when the Westinghouse, Combustion Engineering, and Babcock and Wilcox Owners Groups were deve!oping a safety case to support continued operation of plants. At the time there was no ava!!able crack growth rate data for head penetration materials, and only a few publications existed on growth rates of Alloy 600 in any product form.

The best available publication was found to be that of Peter Scott of Framatome, who had developed a growth rate model for PWR steam generator materials [1]. His model was based on a study of results obtained by McIlree and Smiatowska [2] who had tested short steam generator tubes which had been flattened into thin compact specimens. His model is shown in Figure 2-1. Upon study of his paper there were several ambiguities, and several phone conversations were held to clarify his conclusions. These discussions indicated that reference 1 contains an error, in that no correction for cold work was applied to the McIliree/Smiatowska data. The correct development is given below.

An equation was fided to the data of reference [2] for the results obtained in water chemistries that fell within the standard specification for PWR primary coolant. Results for chemistries outside the specification were not used. The following equation was fitted to the data for a temperature of 330EC:

b = 2.8 x 10-" (K 9)"' m/ sec dt where K is in MPa[m]". This equation implies a threshold for cracking susceptibility, K,cc = 9 MPa[m]". Correction factors for other temperatures are shown in Table 2-1.

The next step described by Scott in his paper was to correct these results for the effects of cold work. Based on work by Cassagne and Gelpi [3], he concluded that dividing the above equation by a factor of 10 would be appropriate to account for the effects of cold work. This step was inadvertently omitted from Scott's paper, even though it is discussed. The crack growth model for 330'C then becomes:

da

= 2.8 x 1042 (K-9)"8 m/sec dt Rev.1 21 October 1998 c:\\3694non. doc:1b:10/09/98

This equation was verified by Scott in a phone call in July 1992.

Scott further corrected this model for the effects of temperature, but his correction was not used in the model employed here. Instead, an independent temperature correction was developed based on service experience. This correction uses an activation energy of 32.4 kCal/ mole,

. which gives a smaller temperature correction than that used by Scott (44 kcal/ mole), and will be discussed in more detail below.

Scott's crack growth model for 350*C was independently obtained by B. Woodman of ABB-CE

[4), who went back to the original data base, and did not account for cold work. His equation was of a slightly different form:

- = 0.2 exp [A + B in (in (K-Q)}]

Where A = -25.942 i

B = 3.595 O=0 I

This equation is nearly identical with Peter Scott's original model uncorrected for cold work.

This work provided an independent verification of Scott's work. A further verification of the i

modified Scott model used here was provided by some operational crack growth rates collected by Hunt, et al [5].

The final proof of the usefulness of Peter Scott's model will come from actual data from head penetration materials in service, as will be discussed further below. To date 15 heats have been tested in carefully controlled PWR environment. One heat did not crack, and of the fourteen heats where cracking was observed, the growth rates observed in tweh a were t

bounded by the Scott model. Two heats cracked at a faster rate, and the e@lanation for this behavior is being investigated.

A compilation was made of the laboratory data obtained to date in the Westinghouse laboratory tests at 325'C, and the results appear in Figure 2-3. Notice that much of the data is far below the Scott model, and a few data points are above the model. These results represent 14 heats of head penetrations.

The effect of temperature on crack growth rate was first studied by compiling all the available crack growth rate data, for both laboratory and fieid cracking of Alloy 600. This information is summarized in Figure 2-2, where the open symbols are for steam generator tube materials, and the solid symbols are for head penetration materials. The results are presented in a simple format, with crack growth plotted as a function of temperature. The effect of different applied stress intensity factor values has been ignored in this presentation, and this doubtless adds to the scatter in the data. The remarkable result is a consistent temperature effect over a temperature range from 288*C to 370*C, more than covering the temperature range of PWR plant operation. The work done originally in 1992 results in a calculated activation energy of 32.4 Kcal/ mole, which has been used to adjust the base crack growth law to account for different operating temperatures.

Rev.1 2-2 October 1998 o:0694non. doc:1b:1000/98

i A series of crack growth tests is in progress under carefully controlled conditions to study the temperature effect for head penetration materials, and the results are shown in Figure 2-3.

Sufficient results are available to report preliminary findings. The tests were performed with an applied stress intensity factor of 23 Ksi 6 (25.3 MPa[m)"), periodic unload / reload parameters of a hold time of one hour and a water chemistry of 1200 ppm B + 2 ppm Li +

25 cc/kg Hz. The results are consistent with the previous steam generator and head

, penetration material work. In the case of heat 69, the three results in the middle of the temperature range,309'C,327*C and 341*C have the same trend as the scatter band, almost exactly, while the high temperature and low temperature results are both lower than would be predicted by the activation energy, as shown in Figure 2-2. The results for heat 20 show a similar behavior, with the results at 325'C and 340*C also with the scatter band and nearly parallel to the heat 69 specimens, but at a lower crack growth rate, as shown in Figure 2-2.

The effects of several different water chemistries have been investigated in a closely controlled series of tests, on two different heats of archive material. Results showed there is no measurable effect of Boron and Lithium on crack growth.

The key test of the laboratory crack growth data is its comparison to field data. Crack growth from actual head penetrations has been plotted on Figure 2-2 as solid points. The solid circles are from Swedish and French plants and the solid stars are from a US plant.

Figure 2-4 shows a summary of the inservice cracking experience in the head penetrations of French plants, prepared by Amzallag [6), compared with the Westinghouse laboratory data, corrected for temperature. This figure shows excellent agreement between lab and field data, further supporting the applicability of the lab data.

Therefore it can be seen that the laboratory data is well represented by the Scott model corrected for temperature using an activation energy of 32.4 kcal/ mole. Also the laboratory results are consistent with the crack growth rates measured on actual installed penetrations.

Therefore the use of the Scott model in the safety evaluations is still justifiable, in light of both laboratory and field data obtained to date.

k l

E Rev.1 2-3 October 1998 o:0694non. doc:1b:10f09/98

___= -_.

1 TABLE 2-1 TEMPERATURE CORRECTION FACTORS FOR CRACK GROWTH: ALLOY 600 Temperature Correction Factor (CF)

Coefficient (Co) 330C 1.0 2.8 x 10

325 0.798 2.23 x 10'"

320 0.634 1.78 x 10

310 0.396 1.11 x 10

300 0.243 7.14 x 10

290 0.147 4.12 x 10"'

b = Co (K - 9)"' m / s dt where K is in MPa[m)"

t Rev.1 2-4 October 1998 o:\\3694non. doc:1b:10/09/98

crack 48 drw 1 E.09 330*C M

E UJ 1E.10 4

E I

O E

O V

g., s.,,

(

g g

l It-12 3

0 20 40 60 80 100 K - MPa SQRT(m) l Figure 2-1 Scott Model for PWSCC of Alloy 600 at 330*C, as modified from reference [1]

Rev.1 2-5 October 1998 oA3694non. doc:1b:10/09/98

TEMPERATU.9E. DEG. C 372 352 333 36 298 282 c*f i

I i

i i

I G.

h 4

g 7

's coo m

w g>

tj k'

W E

W v

Nn j

i x

a O

y 4

E 1*

N

• • t N

W z

e g

c4 t

x x.

s

_ N A

e N

g u

i N&

iA N

h 0

A A \\

A

\\

\\

U M

g E

Tu N t,N m

r-o

,o

'D o

x f)

N M

N i

s A

o.,

n n,

omes coooo onous omosto onoirs cooco RECIPROCAL TEMPERATURE,1/DEG. K

• e ocw rw o..

co us tem Am 20

$ enset 69 m.... se te. urs r= om Figure 2-2 Comparison of Temperature Effects Results with Other Laboratory and Field Data Rev.1 2-6 October 1998 o:\\3694non. doc;1b:10/09/98

. _ =.

cracx47 drw i

1E 09 330*C 325'C

• ,,.. = * *,,,. =,,..

,V)

Q

,o***..

s 3.

. y UJ 1E.10

.g...

H

.79 4

g 1,

.,s'........su.

j o

i...

a::

O

.L.

. j..

V. t g. <

o f;.

.g..

. f...

go 4,...

t r

t i

IE 12 O

20 40 60 80 100 K - MPa SQRT(m)

Figure 2-3 Summary of Available Westinghouse Laboratory Data for Alloy 600 Head Penetrations at 325'C Rev.1 27 October 1998 c:0694non. doc 1b:1&O9/98

1 Comparison of Field & Laboratory Data 10 1

-...:.=_.,..._:.-.-------

. _... =.-.:.: =_:=. _.. :.... ____..._....._....__=.._._.:--....

... _._=._..

_ _ = _

• • ~.._.__e

-e

. _ e.o__

e e

y 7

,I e

a 7

oo. s em s. 0.1 -

-q

- _ ~ _ _

,g oy-

• ~

~

.s e

m oo 0 01

= _ _ _. - - -

-- e FIELD r- _ _ _

o Westinghouse Lab ~~-

0.001 0

10 20 30 40 50 60 70 80 Kl(Mpa m")

Figure 2-4 Comparison of French Field Data and Westinghouse Laboratory Data M results reduced to 290*C using Q = 130 KJ/ mole) [6]

Rev.1 2-8 October 1998 oM694non. doc:1b:1&D0/98

I 3.0 TECHl0 CAL DESCRIPTION OF PROBABILISTIC AND ECONOMIC DECISION MODELS The following two sections of this report describe the models and software for calculating the probability of failure with time and performing the economic decision analysis. The input to these models and the calculated results are described in Section 4 of this report.

~ 3.1 PROBABILISTIC MODEL To calculate the probability of failure of the Alloy 600 vessel head penetration as a function of operating time t, Pr(t s t,), structural reliability models were used with Monte-Carlo simulation methods. This sedion describes these structural reliability models and their basis for the primary failure mode of crack initiation and growth due to primary water stress corrosion cracking (PWSCC). The models used for the evaluation of ths V.C. Summer vessel head penetration nozzles are based upon the economic decision tools developed previously for the Westinghouse Owners Group (WOG). The capabilities of this software have already been verified in the following ways:

1.

Calculated stresses compare well with measured stresses (see Figure 3-1),

2.

A wide range (both high and low valuer,) of calculated probabilities are consistent with plant observations as discussed below.

The model predictions have been used to justify the scope for the second inspection performed at D. C. Cook Unit 2, when the cracked penetration was successfully repaired. The model accepts measured microstructure (replication) and has capability to ignore its effects, if desired.

Recent improvements have also been made by Westinghouse to the software models in order to maximize their utility for individual plant predictions. Among the changes were:

1.

Improved the reiationship of initiation time to material microstructural effects and yield strength to more closely match the observations from the recent inspection at North Anna Unit 1, 2.

Added statistically based Bayesean updating of probabilities due to initial inspection results (e.g. the lack of any indications at any given plant),

a 3.

Updated the uncertainty on crack growth rate after initiation to reflect that observed in the recent Westinghouse test data and the recent in-reactor measurement data to be published by EdF (see Figure 3-2) and 4.

All models have been independently reviewed by APTECH Engineering (Begley and Woodman), including an improled model for the effect of monotonic yield strength on time to initiation.

Rev.1 3-1 October 1998 c:\\3694non. doc:1b:10/09/98

m____

The most important parameter for estimating the failure probability is the time to failure, t,in hours. It is defined as follows:

t, = t, + (a, - a ) / da/dt (3-1) where:

time to initiation in hours, t,

=

i failure crack depth in inches, a,

=

crack depth at initiation in inches and a,

=

da/dt crack growth rate in inch / hour.

=

In equation (3-1), both the crack depths at failure and initiation may be specified as a fraction of the penetration wall thickness, w. The failure depth a, depends upon the failure mode being calculated. Since the failure mode of concem is cracks in the penetration that are deeper than the structural limit of 75% of the penetration wall thickness w, it would be specified as:

l a, = 0.75 w (3-2) l The time to PWSCC crack initiation, t,in hours, is defined by a model that includes the following terms and their uncertainties:

a.

a log-normal distribution on an initiation coefficient, which was based upon the data of Hall and others [8] for forged Alloy 600 pressurizer nozzles, with only the uncertainty based upon the data of Gold and others [9],

b.

a grain boundary coverage factor, which is based upon the data of Norring and others

[10),

c.

the residual and operating stress lev 11 derived from the detailed elastic-plastic finite-element analysis from the WOG study of Ball and of" s [11) as shown in Figure 3-1.

Its normally distributed uncertainty was derived fro

- ' ariation in ovality from Duran and others [12] (see Figure 3-3), which is a trigono,s.-

., action of the penetration diameter and setup angle (local angle between the heaa and longitudinal axis of penetration).

d.

an initiation activation energy, which a also normally distributed,

)

I e.

the penetration material temperature, which is uniformly distributed based upon the calculated variation of the nominal head operating temperature, and 3

I f.

the hours at temperature per operating cycle (year), which is normally distributed.

Rev.1 3-2 October 1998 c:0694non. doc:1b:10/09/98 I

g Either replication data can be used or a roodel can be used for grain boundary carbide coverage. The model [7] is a statistical cotrelation of measured values with the following materials certification parameters:

- Carbon content,

- Nickel content,

- Manganese content,

- Ultimate tensile strength and

- Yield strength.

The uncertainty on this model, which is as shown in Figure 3-4, applies equally well to both the predicted and measured values.

Once the crack has initiated, it is assumed to have a depth of a, and its growth rate, da/dt in inches per hour, is calculated by the Peter Scott model, which matcliefs the latest Westinghouse and EdF data and the previous data given in the WOG report on the industry Alloy 600 PWSCC growth rate testing results [13), and discussed in Section 2. The key parameters in the model are:

a.

a log-normally distributed crack growth rate coefficient (see Figure 3-2),

b.

the stress intensity factor conservatively calculated assuming a constant stress through the penetration wall for an axial flaw at the inside surface with a length 6 times its depth using a simplification of the Raju and Newman equations for pressure vessel evaluation

[14], and c.

an activation energy for PWSCC crack growth, which is also normally distributed.

To calculate the effects of an in-service inspection (ISI) for the economic decision analysis of Section 3.2, the structural reliability ISI model uses a simple but conservative assumption that the probability of detection is directly proportional to the ratio of the depth of the crack to the wall thickness (e.g. 50% detection probability for a crack depth of 50% of the wall. No credit is given for previous inspections so that the effect of the first inspection can be calculated for each year of operation.

The probability of failure of the Alloy 600 vessel head penetration as a function of operating time t, Pr(t s t,), is calculated directly for each set of input values using Monte-Carlo simulation.

To apply the simulation method for vessel head penetration nozzle (VHPN) failure, the existing Westinghouse PROF (probability of failure) Software System (object library) was combined with the PWSCC structural reliability models described previously. The Westinghouse PROF library provided standard input and output, including plotting, and probabilistic analysis capabilities (e.g. random number generajon, importance sampling). The result was program VHPNPROF for calculation of head penetration failure probability with time.

The Westinghouse PROF Software Library has been verified by hand calculation for simple models and altamative methods for more complex models. Recently the application of this I

Rev.1 3-3 October 1998 c:\\3694non. doc:1b:10/09/98

same Westinghouse PROF methodology to the WOG sponsored pilot program for piping risk based inspection has been extensively reviewed and verified by the ASME Research Task Force on RBI Guidelines [15) and other independent NRC contractors. Table 3-1 provides a summary of the wide range of parameters that were considered in this comprehensive benchmarking study that compared the Westinghouse PROF calculated probabilities with those from the pc-PRAISE program [16). As shown in Figure 3-5, the comparison of calculated probabilities after 40 years of operation is excellent for both small and large leaks and full breaks, including those reduced due to taking credit for leak detection.

To verify the proper operation of the VHPNPROF Program in predicting the probability of getting a given crack depth due to PWSCC, calculated results were compared for four plants

~

where sufficient head penetration information and inspection results were available. The four plants are identified in Table 3-2 along with the values of the key input parameters and calculated failure probabilities. For comparison, the latest available inspection results are also provided. Table 3-2 shows acceptable agreement between the observed plant and VHPNPROF calculated failure trends due to PWSCC.

The input and output parameters for the VHPNPROF program runs for the 65 V.C Summer head penetration nozzles are provided in Appendix A and discussed in Section 4.1.

Rev.1 3-4 October 1998 oM694non. doc:1b:10/09/98

TABLE 3-1 Pt.' AMETERS USED FOR THE PC PRAISE BENCHMARKING STUDY Type of Parameter Low Value High Value Pipe Material Ferritic Stainless Steel

+

Pipe Geometry 6.625" O.D.

29.0" O.D.

0.562" Wall 2.5" Wall Failure Modes Small Leak, Full Break, Through-Wall Crack Unstable Fracture Last Pass Weld Inspection No X-Ray Radiographic Pressure Loading 1000 psi 2235 psi Low-Cycle 25 ksi Range 50 ksi Range Loading 10 cycles / year 20 cycles / year High-Cycle

• 1 ksi Range 20 ksi Range Loading 0.1 cycles / min.

1.0 cycles /sec.

Design Limiting Stress 15 ksi 30 ksi Disabling Leak Rate 50 gpm 500 gpm Detectable Leak Rate None 3 gpm

• Note: Mechanical Vibration (Iow stress range and high frequency) for small pipe, Thermal Fatigue (high stress range and low frequency) for large pipe.

l Rev.1 3-5 October 1998 o:0694non. doc:1b:10/09/98

TABLE 3-2 COMPARISON OF VHPNPROF CALCULATED PROBABILITIES WITH PLANT OBSERVATIONS Parameters Almaraz 1 D. C. Cook 2 Ringhals 2 North Anna 1 Hours of Operation 85,400 87,000 108,400 91,000 Setup Angle (')

42.6 50.5 38.6 Ternperature ('F) 604.3 598.5 605.6 600.0 Yield Strength (ksi) 37.5 58.0 51.2 51.2 Percent GBCC 57.0 44.3 3.0 2.0 Flaw Depth / Wall 0.10 0.43 0.25 0.10 Initiation Probability 1.1 %

41.4 %

37.6 %

15.3 %

Failure Probability" 1.1 %

38.1 %

34.6 %

15.3 %

Penetrations 0

1 3

0 With Indications (2 with scratches)

• Calculations performed at an equivalent setup angle for the 2nd highest stress location that could be inspected.

" Defined here as the probability of reaching the specified flaw depth for the limiting penetration.

t l

i Rev.1 3-6 October 1998 oA3694non. doc:1b:10/09/98

s e

.8" O

8 O

O O

b 3

d t

a 8

a E

a O

8

[

7 8

g8 i

o r

~

6 1

U3 a

u a

m 5

R 8

v 5

2

• s E

1 o

8 5

n O

3 b>

w do g

a e

a n

o e

O

~

g OO s-L 82

=

o 1

o-1 b

d s

,n

_o e

d E

e 8

8 8

R 8

8 8

8 od E

-g

( !s>t) sse.21s dooH Mood episuI M

,E E o i

10 i i i

i

i

.i.:

j i

i*

I I l I

i

?

+

l...,i,i..

i i

I !jl

.;.I j

! I i.

1 i :

I i i i t :

1 i

.I l e l ~

T ;

l p...

j

~.

i :

i i..

e J,

,,,i p 1 i,-

i 1

on l

l p 4Y 8i i !

-J

o

!o j

el!

r i

s E

00' #'1 i P-

$ *E D

3 0.1

(

l

~~

l' i. l I.

g

/

e-1 j

),

/

o4'k.i e

s

/

I i

i.

e i s

i l

fi l l/

I I

I f

Io';

I.ll ll

' /'

9 j:

!lj

!!.'j,!ll 0.01 I l j t e FIELD i ;

d

oWestinghouse Lab t

8 l

,I i

t t

1-l-..

g l

j t lii i

• I.

I 1

t i i-l*!ll!

I i i

?

li i

l

'i I l!l l

! l i

l 0.001 0

10 20 30 40 50 60 70 80 MI(Mpa m")

l Figure 3-2 Comparison of Recent Alloy 600 Data with the Ocack Growth Rate Model Rev.1 3-8 October 1998 oi3694non. doc:1b:10/09/98

=

5 B

o l ci l

s 6

I4 l

W

+i E

\\ -

y

\\

l l

5

'\\

1 I

o

(

h v

e

\\

l o

2 E

\\:

\\

g a:

\\

?

8 e

=

e2 i

3 m

g a

.t

• c

\\.

i g

3

\\l g

j

\\

u h

2

=

6 e

o 8

=+

3 W it:

2 i

1

\\

,o g

es e

g

• T 3

o s

._ (

\\

f

\\

s

\\

t a

n n

E

\\

l oo

+

\\

s L

u.

=

\\A\\

[

oe c

5 8

3 8

8 m

e 6

6 e

e e

d y

a g

(goug AytsAo amerng

.l ts E O 4

1 l

Plot of All 40 Data (C^2.6,Mn,UTS,YS,(C/Ni)^0.7)

-l l

l l

a +

i ie 4

i a i l.

il 95

/

3

/

p f

75 p.4-g '-

J j

./

/

/

~

/

~

/

Q 55

/

/

~~

k

/

/

w i

~

m

/

~

e m

/

O

~

j l

/

35

/

~

~

/

/

y l

/

/

/

i 15

/

S,/

.. */

/

5 l

l/,

Iii i li

, e i

ii l

i i i i

<5 15 35 55 75 95 PREDICTED Figure 3-4 Curve Fit of Alloy 600 Grain Boundary Carbide Coverag,e Results (7]

Rev.1 3-10 October 1998 c:0694non. doc:1b:10/09/98

0.1 a...P..a*......

I i

.001

~

0 r

'DO n

=

l j 1E05

.O t

r

~ - ~

~

f-

..O o

q'p........-

u i

a.

g.

1E07 r

~

g x

r en 88

.... - g l

=

1E-09 r

- O i

9

[l ....,..=*

"b 1E-11 1E-11 1E-09 1E-07 1E-05 0.001 0.1 l

PRAISE Probability Perfect SmaN Large fun FR Leak Lock Break o

a i,

L Figure 3-5 Comparison of Calculated Piping Probabuttles i

Rev.1 3-11 October 1998 o:0694non. doc:1b:1GV9/99

i 3.2 ECONOMIC DECISION ANALYSIS MODELS The basis for the economic decision analysis model is the Influence Diagram for Plant Life Extension (PLEX) shown in Figure 3-6. The relationships shown by the dashed lines are not included since VHPN cracking due to PWSCC is not a safety issue. The component mitigative strategy in this case is the first inspection of the outer three rows of vessel head penetration nozzles and repair of those with detectable cracks. The probability of failure, which is a crack depth'75% of wall, and probability of inspection detection (1-PND) for each year of operation and group of penetrations come from the output files for the VHPNPROF analysis runs. The effectiveness of this mitigative strategy on future failure costs can also be calculated directly using this same information instead of being estimated as is done in other decision analysis models.

The output files for the V. C. Summer vessel head penetration nozzle decision analyses are incl.td in Appendix B. The first page of the output file summarizes the input, which is

<escribed in Cection 4.2. The next two pages are the results of model calculations, which can be described as follows for each column heading on each page.

CYCLE:

Number of operating cycle (year) when values of the parameters below are l

calculated.

MAX-PROB: This is the maximum failure probability calculated by VHPNPROF for all the penetration nozzles.

i PROB-ONE: This is supplementary information about the probability that at least one of the head penetration nozzles will fail.

l t

AVG-PROB: This is the average failure probability, which is the expected number of failures that is used lo calculate the failure cost divided by the number of head penetration nozzles.

NPVFC-50:

The Net Present Value of the median (50% probability) failure cost, which is the product of the expected number of failures and the median cost per penetration nozzle failure.

NPVFC-05:

5% Lower confidence bound on the NPV of the failure costs.

NPVFC-95:

95% Upper confidence bound on the NPV of the failure costs.

CYlSI:

Number of operating cycle (year) after which the first in-Service inspection (ISI) would be performed.

NPV-CISl:

This is the NPV of the median inspection cost, which is the number of nozzles in the outer three rows times the average inspection cost per nozzle. Because of the time value of money, the later the inspection is performed, the lower its NPV.

i NPV-CREP: This is the NPV of the median repair cods, which is the average repair cost per nozzle times the fraction of inspected nozzles with cracks large enough to lead to i

Rev.1 3-12 October 1998 c:0694non. doc:1b:10/09/98

failure and to be detected during ISI. The value of this fraction is calculated directly from the VHPNPROF output for the groups of nozzles being inspected.

NPV-CBEN: This is the NPV of tne median cost benefit of doing the inspection. The benefit is the elimination of the future failure costs for those nozzles that have been repaired. There is no reduction in failure probability and the associated expected failure cost contribution until a partially cracked nozzle is repaired.

NPVTC-50:

This is the median NPV of the total cost integrated over a 60-year plant life. It is the sum of the NPV of the failure cost for all nozzles at 60 years and the inspection and repair costs less the NPV of the cost benefit of the repairs. The best economic decision would be to perform the first inspection when the NPV of this cost is a minimum.

NPVTC-05:

5% Lower confidence bound on the NPV of the total cost.

NDUTC-95:

95% Upper confidence bound on the NPV of the total cost.

The input to these models and the output values calculated by the decision analysis program VHPNECON are described in Section 4.2 for the V. C. Summer vessel head penetration nozzles.

i i

1 J

e 1

Rev.1 3-13 October 1998 o:\\3694non. doc:1b:100948

{qw'

~e' v

=:

PiliX f

Essent of I I l

Fadise

.EX

/

Activety o

Impact.

l se.CDF Forced jf a

\\

conor w

nepense.-

A' ToealCost (NPV RR)

Conscupsenessi Damage Casas Desmoonot l

Coss ef

,S Ouease

~ _ -

g, l

~ ~.,

~

NOTE

~~~~~...

l-----l Encluded from decesson/mvesemene snodel 1

- safety /anpactcosas

- conseepsensialdamage cosas i

Figure 3-6 Component DecisionAnvestment Modelinfluence Diagram t

f l

Rev.1 3-14 October 1998 oM694non. doc:1b:10/09/98 i

I

4.0 RESULTS OF PROBABILISTIC AND ECONOMIC DECISION MODELS i

4.1 INPUT AND RESULTS OF PROBABILISTIC ANALYSIS The V. C. Summer reactor vessel and closure head were manufactured for Westinghouse by the Chicago Bridge and Iron Company. The closure head contains 65 head penetrations

, fabricated from Alloy 600 tube which are welded to a stainless steel flange. This assembly is then welded to the low alloy steel closure head utilizing a J-groove weld. An outside view of the closure head which shows the penetration numbers is provided in the following sketch. These penetrations are utilized for a number of purposes. These purposes are for Control Rod Drive Mechanisms (CRDM), capped latch housings (CLH), part length mechanisms (P/L),

thermocouple column locations (TCC), reactor vessel level instrumentation system connection (RVLIS), and spare penetrations.

O t

l

@l D'@

30 __

__g__..g- _ g.

_ g _. g.___ g _______ __ 4

,0

=

@.G l@

~

l l

A review of the fabrication records for V.C. Summer indicates that the closure head penetrations were fabricated from two different heats of Alloy 600 material. Both of these heats of material were supplied by Babcock and Wilcox and are designated as M6369 and M6370.

Rev.1 4-1 October 1998 o:0694non. doc:1b 10/09/98

)

i Table 4-1 provides a summary of each head penetration and its use and associated heat of material.

Table 4 2 provides the input values to the probabilistic analysis and Table 4-3 provides the j

results of the analysis in terms of the probability of failure (%) after 10,20,30,40,50, and 60 years of operation (note that penetrations 1 through 25 are bounded by penetrations 26 through

,33 since they utilize the same heats of material and their set-up angle is less than that of penet' rations 26 through 33).

t i

t The detailed input and calculated results for the V.C. Summer vessel head penetration nozzle i

probabilistic analysis are given in the VHPNPROF output print files in Appendix A. The first I

page of each file is a descriotion of the input for each analysis, including the standard uncertainties that were used for the probabilistic analysis. The second page of the output file lists the calculated probabilities.

j The first column is the cycle number; the second is the probability of failure during the cycle; the third is the accumulated probability at the end of the cycle. The foudh and fifth columns are the l

same types of probability as the second and third columns respectively but for an in-service inspection (ISI) each cycle. This is of course an unrealistic assumption, but provides useful informction for the economic analysis.

Figure 41 shows the increase of the best-estimate crack depth with time for the penetrations l

with the highest failure probability in some of the outer rows. The shortest mean time to failure l

(depth of 75% of the wall thickness or approximately 0.5 inch) of [

] years is for group 1 l

(penetrations 58 to 65) or case 1 in Appendix A. For the scond row in (penetrations 49 to 52),

i the residual stresses are lower so that the time to crack initiation is longer and the crack growth rate is smaller. In this case, the mean time to failure increases to almost [

] years.

Likewise, for the third and fourth rows in the mean times to failuru are approximetMy [

T*and

[

] years, respectively. Because of the effects of all the uncertainties that are considered in the probabilistic analysia, the uncertainty band on the time to failure is quite wide. Even with a mean time of failure of[

T* years for the case 1 penetrations (58 to SS), there is about at

[ ]'*% probability of failure by year 60 (see Table 4-3). However, as the mean time to failure increases for the inner rows, then the probability of failure at a given time, say 60 years, decreases. For the case 8 penetrations (34 to 38), there is only a [ ]'*% probability of failure by year 60 because the mean time to failure increased to [

] years as shown in Figure 4-1.

To calculate the combined offects for all the vessel head penetration nozzle (VHPA) failures (creck depths of 75% of the wall), a second program (VHPNECON) was run. The results of these calculations are given in the VHPNECON output file, which is shown in the first page of Appendix B. The column headings used in the output file ar.d their meaning are described below.

. CYCLE:

Number of operating cycle (year) when values of the parameters below are calculated. Each cycle has 7446 hours0.0862 days <br />2.068 hours <br />0.0123 weeks <br />0.00283 months <br /> at temperature. For these calculations each cycle was assumed to be one year.

Rev.1 4-2 October 1998 oM694non. doc.1b:10/09/98

MAX-PROB: This is the maximum failure probability calculated by VHPNPROF for tt e penetration nozzle most lil< sly to fail.

PROB-ONE: This is the probabliity that at least one of the head penetration nozzles will U. It is calculated as follows:

P = 1-D, (1 - p,)"

(4-1) where p, = failure probability for the ith group nl = number of penetrations in the ith group N = number of groups p

AVG-PROB: This is the average failure probability, which is the expected number of failures divided by the r. amber of head penetration nozzles.

E(NUMFS):

This is the expected value of the number of failures in all the penetrations, it is calculated as follows:

E(N,) = 4 ni p, (4.

Table 4-5 provides the results of the analysis for the probability of at least one penetration failure in the head.

Figure 4-2 shows the failure probability with time for eedt of the penetrations (58 to 65) in the highest group (1 or case 1 in Table 4-2 and Appendix & This figure also shows the increama in the average failure probability with time for all 65 per;etration nozzles for the V.C. Summer vessel hea:'. This average probability is 1/65th of the expected number of failures used in the economic decision analysis of Section 4.2. For reference, [

]'*% is the calculated failure (75% wall depth) probability in the worst penetration in D.C. Cook Unit 2 when a crack depth of 43% of tN wall thickness was found after 87,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of operation. The corresponding average rLi.ute probability is [

]"% and the probability of at least one failure is [

]'*%for all 78 penetration nozzles in D.C. Cook 2.

4.2 INPUT AND RESULTS OF ECONOMIC DECISION ANALYSIS The output files for the economic decision analysis on when to perform the first inspection of the outer three rows of vessel head penetration nozzles in V. C. Summer is listed in Appendix B.

The first page of the output is a summary of the input to the VHPNECON Program.

The reference year for the net present value calculation was set to cycle (year) 14 based upon the total hours of operation at temperature to date and an average 7,446 hours0.00516 days <br />0.124 hours <br />7.374339e-4 weeks <br />1.69703e-4 months <br /> per cycle used in ti. : VHPNPROF analyses. The interest rate of 5% is based upon an assumed discount rate of 9% less an assumed 4% escalation rate.

The range of costs for failure inspection and repair were calculatud using the same method to combine uncertainties as was used for the simpla WOG cost model. The cost calculations for Rev.1 4-3 October 1998 oA3694non. doc:1b:10/09/98

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

-. - ~

the V. C. Summer decision analysis are summarized in Table 4-4. The cost of inspection would include eddy-current inspection of all the sleeved and unsleeved penetrations in the outer three rows and a ultrasonic inspection of one flaw in one penetration. The repair costs are based upon excavation of one flaw in one penetrat:on.

The failurs costs arc based upon excavation of one deep flaw and weld overlay repair for one

, penetration only. Also included are the additional industry /NRC interaction costs and ALARA penalty costs from the simple cost model developed for WOG. Not included in the failure costs were the follow-on inspection costs for the repaired nozzle. Replacement power costs for extension of critical path time or unexpected shutdown due to leakage of a nozzle were not included in the subtotal of the failure costs in Table 4-4. Tnis cost penalty at an assumed

[

]' per day significantly increases the total failure cost in Table 4-4 as well as the cost avoidance benefit of the penetration nozzle inspection and repairs.

i Figure 4-3 shows the 5,50 (median value) and 95% confidence bounds on the NPV of the minimum total costs of failure through 60 years including the NPV of the inspection and repair costs at the time (cycle) for the first inspection. The minimum failure costs do not include the I

high downtime replacement power penalty costs. As can be seen, the minimum NPV cost would occur for no inspection at all (cycle 59). Because of the low failure cost for the low failure probabilities of the V. C. Summer vessel head penetration nozzles, the expected benefit of repairing the detected cracked penetrations never offsets the inspection and repair costs.

However, the benefits of the first inspection and repair of detected cracks are increased significantly when the total failure cost includes the replar:ement power costs for an unplanned repair of failed penetration nozzle. Figure 4-4 shows the 5,50 (median value) and 95%

confidence bounds on the NPV of the maximum total costs of failure througii 60 years, where the maximum total failure costs include the replacement power penalty costs. For this maximum cost case, the minimum NPV cost would occur for inspection at end of cycle [

]**

(calendar year [

]'*).

i i

i i

I i

Rev.1 4-4 October 1998 c:G694non. doc:1b:10/09/98

TABLE 41 HEAD PENETRATION USES AND ALLOY 600 HEAT NUMBER Row Penetration No.

Use Thermal Sleeve Heat Number 0

1 P/L YES M6369 1

2 CRDM YES M6369 3

CRDM YES M6363 4

CRDM YES M6370 S

CRDM YES M6369 2

6 CRDM YES M6369 7

CRDM YES M6369 8

CRDM YES M6370 9

CRDM YES M6369 3

10 CRDM YES M6370 11 CRDM YES M6370 12 CRDM YES M6370 13 CRDM YES M6370 4

14 SPARE NO M6370 15 SPARE NO M6369 16 RVLIS NO M6369 17 SPARE NO M6369 5

18 P/L YES M6370 19 P/L YES M6370 20 P/L YE8 M6369 21 Pit.

YES M6370 6

22 CRDM YES M6369 23 CRDM YES M6369 24 CRDM YES M6369 25 CRDM YES M6369 7

26 CRDM YES M6370 27 CRDM YES M6370 28 CRDM YES M6369 29 CRDM YES M6369 30 CRDM YES M6369 31 CRDM YES M6370 32 CRDM YES M6369 33 CRDM YES M6369 Rev.1 4-5 October 1998 o:W94non. doc:1b:10/09S8

l TABLE 4-1 (Continued)

Row Penetration No.

Use Thermal Sleeve Heat Number 8

34 CRDM YES M6370 35 CRDM YES M6370 36 CRDM YES M6370

~

37 CRDM YES M6370 38 CRDM YES M6370 39 CRDM YES M6369 40 CRDM YES M6369 41 CRDM YES M6369 9

42 CRDM YES M6370 43 CRDM YES M6370 44 CRDM YES M6369 45 CRDM YES M6369 10 46 CLH YES M6369 47 TCC NO M6369 48 CLH YES M6369 49 TCC NO M6370 50 Cl.H YES M6370 51 TCC NO M6370 52 CLH YES M6370 53 TCC NO M6369 11 54 CRDM YES M6369 55 CRDM YES M6369 56 CRDM YES M6369 57 CRDM YES M6369 12 58 CRDM YES M6369 59 CRDM YES M6369 60 CRDM YES M6369 61 CRDM YES M6369 62 CRDM YES M6369 I

63 CRDM YES M6369 64 CRDM YES M6369 65 CRDM YES M6369 I

Rev.1 4-6 October 1998 o:0694non. doc:1b:10/09/98 l

l TABLE 4-2 INPUT VALUES FOR PROBABILISTIC ANALYSIS Case Pen.No.

Temp.

SA Y.S. (ksi)

GBC (%)

1 58 thru 65 557.3 'F 46.1 40.531 12.3 2

54 thru 57 43.1 40.531

-12.3 3

49 thru 52 41.6 42.034

-2.1 4

46,47,48 & 53 41.6 40.531 12.3 5

44 & 45 40.1 40.531

-12.3 l

6 42 & 43 40.1 42.034

-2.1 7

39 thru 41 35.5 40.531

-12.3 8

34 thru 38 35.5 42.034

-2.1 9

28,29, 30,32 & 33 30.6 40.531

-12.3 j10 26,27 & 31 30.6 42.034 2.1 TABLE 4-3 RESULTS OF PROBABILISTIC ANALYSIS (PROBABILITY OF FAILURE %)

Case Pen.No. I 10 Years 20 Years 30 Years 40 Years 50 Years 60 Yes 1

58 thru 65

-0 0.2 0.9 2.6 4.8 8.1 2

54 thru 57

-0

<0.1 0.4 1.2 2.7 4.5 3

49 thru 52

-0

<0.1 0.3 1.0 2.3 3.8 4

46,47,48 &

-0

<0.1 0.3 0.9 1.9 3.3 53 5

44 & 45

-0

<0.1 0.2 0.7 1.4 2.6 6

42 & 43

-0

<0.1 0.2 0.7 1.6 2.8 7

39 thru 41

-0

-0

<0.1 0.2 0.6 1.0 8

34 thru 38

-0

-0

<0.1 0.3 0.6 1.1 9

28,29,30,

~0

-0

-0

<0.1 0.2 0.3 32 & 33 10 26,27 & 31

-0

-0

<0.1

<0.1 0.2 0.4 l

l l

Rev.1 4-7 October 1998 oN3694non. doc:1b:10/09/98

TABLE 4-4 COST CALCULATIONS FOR V. C. SUMMER VHPN ECONOMIC ANALYSIS Inspection of Nonies in Outer Three Rows

($K)

High Median Variance

_ _d W Cost Range Utility Cost Range Total Cost Range Total Cost Range / Nozzle Repair of 1 Nonle in Outer Three Rows

($K) i Low High Median Variance

_ _ d W Cost Range Utility Cost Range PCI Cost Range Total Cost Range Failure of 1 Nonle Anywhere

($K)

Low High Median Variance

- _d

)

W Cost Range Utility Cost Range PCI Cost Range NRC/ Industry lateraction Costs ALARA Penalty Subtotal Cost Range Down Time Penalty Total Cost Range w/DTP Rev.1 4-8 October 1998 c:(1694non.dx:1b:10/09/98

i TABLE 4-5 PROBABILITY (%) OF A FLAW WITH DEPTH = 0.75T IN AT LEAST ONE PENETRATION I

l 10 Years 20 Years 30 Years 40 Years 50 Years 60 Years (74500 hrs.)

(149,000 hrs.)

(223,500 hrs.)

(298,000 hrs.)

(372,500 hrs.)

(447,000 hrs.)

-0 2.1 12.0 33.5 57.0 76.9 I

l l

l i

)

Rev.1 4-9 October 1998 oM694non. doc:1b:10/09/98 l

l l

P i

i a,b l

t i

i i

I t

1 r

I i

I 1

a r

i s

t I

i r

i i

l i

i Figure 4-1 Mean Crack Growth for V.C. Summer VHP Nozzles j

i t

Rev.1 4-10 October 1998 i

o:\\3694non doc:1b:10/m i

i

o "o tu b

o fD %

., 3 cr pmesumy 3

3 3

a 14 10*E* :

1.12E+0si i

tisE m ii i l'il" 1.27E 4sia,

134E+08 >i

, ag.o i >

gc 20 - 14eE4si,

i 3

a 1SSE+06q, i N

164E+081 y

mn 9

E 171E+06 q p

e g

1.79E+06q g

2 A.

2$

1.80E+0S q i s

a m

7 194E464i >

t 2.01E464,,

O i

2 09E+0$4 m, i

g t

\\

}

\\

p 2 i.E.oS o 1

t (J) i e

30 - 2.24E+0s <

l 2.31E45.

l c

9 2.3eE+06.

2.4eE+06 1,.,

2.53E+06 i,<>

r 35 - 2 siE*0s.i 2 saEe.i l

l l

l

,4 t

r f

2 76E+05. i,<>

, o; 2s3E+05.e A

2 9tE45. q l

k,,

o 40 - 2 96E*05 cy tD a

.p r

Q 1

i l

.. _._.~. -.

. _ - ~ _

a,b 4

Figure 4-3 Expected Total Costs for No Downtime Penalty l

Rev.1 4,32 October 1998 o:Uti94non. doc-1b:100MMI m

m.

m.

r

-~

~~-'e-

+%-

i- -

i I

t E

a,b 4

L t

i t

t t

i I

i f

?

t t

i l

t r

I i

Figure 4-4 Expected Total Costs with Downtime Penalty l

t ti Rev.1 4-13 October 1998 o:0694non. doc:1b:10RMW98 r

r

l 5.0

SUMMARY

/ CONCLUSIONS A detailed evaluation of the reactor vessel closure head penetrations has been completed for the Virgil Summer plant. One of the two degradation mechanisms covered by Generic Letter 97-01 has been addressed: Primary water stress corrosion cracking (PWSCC).

. An in, depth probabilistic assessment has been completed for all of the reactor vessel closure head penetrations, using state-of-the-art methods which have been independently reviewed.

These methods have also been ver;fied by comparison with actual inspection results, as shown in Table 3-2, and discussed in Section 3.

The results of the assessment show that the mean time to failure for the worst penetration is

[

J years, indicating that the V.C. Summer plant is not at risk for this issue. The probability of a flaw initiating and reaching 75% of the wall thickness in 40 years was calculated to be

[

] percent. For 60 years, the probability increases to [

] percent.

The probabilistic results combined with the economic decision analysis model, and the conclusion reached was that the optimum time (minimum cost) for a head penetration i

inspection at V.C. Summer would be at [

] calendar years of service, as shown in Figure 4-4.

i e

Rev.1 5-1 October 1998 c:0694non. doc;1b 10/09/98

_ =. _ _

l

6.0 REFERENCES

l

[1]

Scott, P. M., "An Analysis of Primary Water Stress Corrosion Cracking in PWR Steam Generators," in Proceedings, Specialists Mceting on Operating Experience With Steam Generators, Brussels Belgium, September 1991, pages 5,6.

l

.[2]

Mc liree, A. R., Rebak, R. B., Smialowska, S., " Relationship of Stress Intensity to Crack Growth Rate of Alloy 600 in Primary Water " Proceedings Intemational Symposium Fontevraud 11, Volume 1, p. 258-267, September 10-14,1990.

j

[3]

Cassacre, T., Gelpi, A., " Measurements of Crack Propagation Rates on Alloy 600 Tuber ' VR Primary Water," in Proceeding of the 5th Intemational Symposium on Enyhnmental Degradation of Materials in Nuclear Power Systems-Water Reactors,"

August 25-29,1991, Monterey, Califemia.

[4]

Personal Communication, Brian Woodman, Combustion Engineering, October 1993.

[5]

Hunt, S L. and Gorman, J., " Crack Predictions and Acceptance Criteria for Alloy 600 1

Head Penetrations"in Proceedings of the 1992 EPRI Workshop on PWSCC of Alloy 600 in PWRs, December 1-3,1992, Orlando Fl (published in 11s93).

[6]

Personal communication - C. Amzallag to W. Bamford, Feb. 26,1997.

[7]

G. V. Rao and T. R. Leax, Microstructural Correlations with Material Certification Data in Several Commercial Heats of Alloy 600 Reactor Vessel Head Penetration Materials -

WCAP-13876, Rev.1, June 1997.

[8]

" Evaluation of Leaking Alloy 600 Nozzles and Remaining Liia Prediction for Similar

. Nozzles in PWR Primary System Application," Hall, Magee, Woodman and Melton, in Service Expen*ence and Reliability improvement, ASME PVP-Vol. 288,1994 j

[9]

"The Status of Laboratory Evaluations in 400*C Steam of the Stress Corrosion of Alloy 600 Steam Generator Tubing," Gold, Fletcher and Jacko in Proceedings of 2nd Inturnational Topical Meeting on Nuclear Power Plant Thermal Hydraulics and Operations,1986

[10]

"intergranular Stress Corrosion Cracking in Steam Generator Tubing, Testing of Alloy 690 and Alloy 600 Tubes," Norring, Engstrom and Norberg, in ThirdIntemational Symposium on Environmental Degradation of Materials in Nuclear Power Systems -

Water Reactors - Proceedings, The Metallurgical Society,1988

[11]

WCAP-13525, Rev.1, RV Closure Head Penetration Alloy 600 PWSCC (Phase 2), Ball et al., December 1992 (Class 2)

[12]

WCAP-13493, Reactor Vessel Closure Head Penetration Key Parameters Comparison, l

Duran, Kim and Pezze, September 1992 (Class 2)

Rev.1 6-1 October 1998 oM694non. doc:1b:10/09/98

[13)

WCAP 13929, Rev. 2, Crack Growth and Microstructural Characterization of Alloy 600 Head Penetration Materials, Bamford, Foster and Rao, November 1996 (Class 2C)

[14]

Newman, J.C. Jr. And Raju, l.S. " Stress intensity Factors for Internal Surface Cracks in Cylindrical Pressure Vessels" Transactions ASME, Journal of Pressure Vesse!

Technology, Volume 102,1980, pp. 342-346.

\\15}

Risk-Based Inspection - Development of Guidelines, Volume 1, General Document, ASME Research Task Force on Risk-Based Inspection Guidelines Report CRTD-Vol.

20-1 (or NUREG/GR-005, Vol.1), American Society of Mechanical Engineers,1991

[16}

NUREGlCR-5864, Theoreticaland User's Manualforpc-PRAISE, A Probabilistic Fracture Mechanics Computer Code for Piping Reliability Analysis, Harris and Dednia, July 1992 l

Rev.1 6-2 October 1998 o:0694non. doc:1b:10/09/98

Appendix A Output Files From VHPNPROF for Proba.bilistic Failure Analysis of the V.C. Summer Vessel Head Penetration Nozzles t

Rev.1 A-1 October 1998 o:0694non. doc:1b:10/09/98

WESTINOHOUSE VESSEL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON ESBU-NSD 65 Nozzles at Virgil C. Summer Plant on 05-31-97 06/06/97 CYCLE MAX-PROB PROB-ONE AVG-PROB E(NUMPS) c.b 14 15 16 17 18 19 20 21 22 j

l 23 24 25 26 27

{

28 l

29 I

30 l

31 l

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 J

I 58 J

59 60 A2 l

l 1

, Output Print File VHPNPROF.P01 Opened at 16:44 on 05-12-1997 Limit Depth Fraction of Wall 0.750 Monotonic Yield Strength (Ksi) 40.5 Penetration Setup Angle (degrees) 46.1 Penetration Tenperature (F) 557.3 Center Penetration Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-12.3 Months in Operating Cycle 12.0 LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating cycles per Year 1.000 e

STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 1: RV Head Penetration CGE 58-65 c.b 4

0 A-3

,~n.

I 0.b i

6 I

I e

m 14memmum 4

I e

i i

1 J

A-4

Output Print File VHPNPROF.P02 Opened at 16:51 on 05-12-1997 Limit Depth Fraction of Wall 0.750 Monotonic Yield Strength (Kai) 40.5 Penetration Setup Angle (degrees) 43.1 Penetration Temperature (F) 557.3 Center Penetration Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-12.3 Months in Operating Cycle 12.0 LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating cycles per Year 1.000 e

STRUCTURAL RELIADILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 2: RV Head Penetration CGE 54-57

~

0.b e

9 A5

h 1-

--a a

J

-4,

---+

eam.

O.b G

l

)

ammmuus N

O i

\\

A-6

Output Print File VHPNPROF.P03 Opened at 17.00 on 05-12-1997 Lirnt Depth Fraction of Wall 0.750 Monotonic Yield Strength (Ksi) 42.0 Penetration Setup Angle (degrees) 41.6 t

Penetration Tertperature (F) 557.3 Center Penetration Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-2.1 Months in Operating Cycle 12.0 LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 e

STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 3: RV Head Penetration CGE 49-52 c.b 4

e e

A-7

C.b I

l i

i l

i D

4 0

ensame mammme h

e a

I A-8

Output Print File VHPNPROF.PO4 opened at 17:06 on 05-12-1997 Limit Depth Fraction of Wall 0.750 Monotonic Yield Strength (Ksi) 40.5 Penetration Setup Angle (degrees) 41.6 Penetration Temperature (F) 557.3 Center Penetration Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-12.3 Months in Operating Cycle 12.0 LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 o

STRUCWRAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 4: RV Head Penetration CGE 46; 47; 48; 53 c.b e

a A-9

0.b N

-meme e

A-10 i

I Output Print File VHPNPROF.POS Opened at 17:12 on 05-12-1997 Limit Depth Fraction of Wall C.750 Monotonic Yield Strength (Kst) 40.5 Penetration Setup Angle (degrees) 40.1 Penetration Temperature (F) 557.3 Center Penetration Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-12.3 Months in Operating Cycle 12.0

?

LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 5: RV Head Penetration CGE 44 & 45

~

o,b 4

s Mummus A-11 1

a 2

c.b B

emanuma N

I i

9 A 12

Output Print File VHPNPROF.P06 Opened at 17:17 on 05-12-1997 Limit Depth Fraction of Wall 0.750 Monotonic Yield Strength (Ksi) 42.0 Penetration Setup Angle (degrees) 40.1 Penetration Temperature (F) 557.3 Center Penetratian Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (t)

-2.1 Months in Operating cycle 12.0 LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 6: RV Head Penetration CGE 42 & 43

~

o,b t

A-13

ar-A Ob e

i f

f M

A-14 l

l

.-..- ~.

. ~.

~. _.

l l

i Output Print File VHPNPROF.P07 Opened at 17:22 on 05-12-199'1 Limit Depth Fraction of Wall 0.750 Monotonic Yield Strength (Ksi) 40.5 i

Penetration Setup k.3 e (degrees) 35.5 1

Penetratior Tamperature (F) 557.3 i

Center Penet7. tion Stress (Kai) 34.4 Grain Boundary Carbide Coverage (%)

-12.3 Months in Operating Cycle 12.0 LOG's of Years Between ISI 0.00 9

i Wali /raction for 50% Detection 0.500 Operating Cycles per Year 1.000 I

i a

STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA) l WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 7: RV Head Penetration CGE 39-41 I

l c.b l

l l

l t

l l

1 i

l i

I 6

I t

4

[

3 i

m A-15 4

4 l

4 l

1 1

O.b i

l D

0 emmunnum A $6

__._,...._m__.__.

__..._.._,m__

m.

i I

Output Print File VHPNPROF.P0B Opened at 17:28 on 05-12-1997 Limit Depth Fraction of WA11 0.750 Menotonic Yield Strength (Ksi) 42.0 Penetration Setup Angle (degrees) 35.5 Penetration Temperature (F) 557.3 Center Penetration Etress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-2.1 Months in Operating Cycle 12.0 LOG 10 of Years Between ISI C.00 l

Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 l

STRUCTURAL REI.IABILITY AND RISK ASSESSMMNT (SRRA) l WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD i

INPUT VARIABLES FOR CASE B: RV Head Penetration CGE 34-38 l

l c.b i

l l

1 l

1 l

l l

l 6

p 4

i 3

1, l

A-17 J

J 1

0.b I

e e

ammuumm ummumus 9

e I

i A-18

..~ -..

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

.. - ~ _...

l l

I i

l Output Print File VHPNP.P09 Opened at 17:33 on 05-12-1997 l

Limit Depth Fraction of Wall 0.750 l

Monotonic Yield Strength (Ksi) 40.6 I

Penetration Setup Angle (degrees) 30.6 l

Penetration Temperature (F) 557.3 Center Penetrution Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

--12.3 l

Months in Operating Cycle 12.0 l

LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 l

STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE 9: RV Head Penetration CGE 28;29;30:32;33

~

0.b i

I

(

i i

1 4

)

ammu l

A-19 4

4 e'

+

y-i

~

0.b 1

r 6

m N

o 1

i l

1 I

\\

i A-20

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

i l

I i

Output Print File VHPNPROF.P10 Opened at 17:37 on 05-12-1997 Limit Depth Fraction of Wall 0.750 Monotonic Yield Strength (Ksi) 42.0 j

Penetration Setup Angle (degrees) 30.6 Penetration Temperature (F) 557.3 Center Penetration Stress (Ksi) 34.4 Grain Boundary Carbide Coverage (%)

-2.1 Months in Operating Cycle 12.0 LOG 10 of Years Between ISI 0.00 Wall Fraction for 50% Detection 0.500 Operating Cycles per Year 1.000 STRUCTURAL RELIABILITY AND RISK ASSESSMENT (SRRA)

WESTINGHOUSE PROBABILITY OF FAILURE PROGRAM VHPNPROF ESBU-SMPD INPUT VARIABLES FOR CASE los RV Head Penetration CGE 26; 27; 31 0,b 1

l o

l 4

el A-21 d

o.b 9m a

M M

s I

\\

i i

A-22

.. _. - _ =.

I 4

l Appendix B N

Output Files From VHPNECON for Economic Decision Analysis of the V.C. Summer Vessel 4

i Head Penetration Nozzles j

4 4

1 4

I i

J

]

I i

i 1

i i

Rev.1 B-1 October 1998 QM694non. doc:1b:1009/98

...-. ~

-. ~

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

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

WESTINGHOUSE VESSEL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON ESBU-NSD 65 Nozzles at Virgil C. Summer Plant on 06-10-97 06/06/97 j

Ref. Year & Interest Rate (%) for NPV Calculations = 1.400E+01 5.000E+00

- d Min, and Max. Failure Cost per Penetration ($K)

=

. Min, and Max. Inspection Cost per Penetration (SK)

=

^

Min. and Max. Repair Cost per Penetration ($K)

=

Reading Probabilities for 8 Nozzles in Group 1 From Files VHPNPROF.001 Reading Probabilities for 4 Nozzles in Group 2 From File: VHPNPROF.002 Reading Probabilities for 4 Nozzles in Group 3 From File VHPNPROF.003 Reading Probabilities for 4 Nozzles in Group 4 From File: VHPNPROF.004 Reading Probabilities for 2 Nozzles in Group 5 From File: VHPNPROF.005 Reading Prot lities for 2 Nozzles in Group 6 From File VHPNPROF.006 Reading Probabilities for 3 Nozzles in Group 7 From File VHPNPROF.007 Reading Probabilities for 5 Nozzles in Group 8 From File: VHPNPROF.008 Reading Probabilities for 5 Nozzles in Group 9 From File: VHPNPROF.009 Reading Probabilities for 28 Nozzles in Group 10 From Files VHPNPROF.010 1

l t

l 1

1 B-2

I i

I t

i WESTINGHOUSE VESSEL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON l

ESBU-NSD 65 Nozzles at Virgil C. Summer Plant en 06-10-97 06/06/97 CYCLE MAX-PROD PROB-ONE AVG-PROB NPVFC-05 NFVFC-50 NPVFC-95

- Ab 14 15 16 17

.18 i

19 20 21 22 23 24

]

25 26 27 28 29 30 31 32 l

33 34 35 36 37 38 39 40 41 42

)

43 44 45 46 47 48 j

49 50 51 52 53 I

l o 54 l

55 56 l

57

^

58 59 60 l

4 l

v3 4

4

I WESTINGHOUSE VESSEL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON ESBU-NSD 65 Nozzles at Virgil C. Summer Plant on 06-10-97 06/06/97 CY.SI NPV-CISI NPV-CREP NPV-CBEN NPVTC-05 NPwTC-50 NPVTC-95 a

14 l

15 16 17

.18 l

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 B-4

l i

WESTINGHOUSE VES3EL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON l

ESBU-NSD 65 Nozzles at Virgil C. Summer Plant on 06-10-97 06/06/97 Ref. Year & Interest Rate (%) for NPV Calculations =

1.400E+01 5.000E+00 d

Min, and Max. Failure Cost per Penetration ($K)

=

Min. and Max. Inspection Cost per Penetration ($K)

=

Min. and Max. Repair Cost per Penetration ($K) j

=

l Reading Probabilities for B Nozzles in Group 1 From File: VHPNPROF.001 Ruading Probabilities for 4 1;ozzles in Group 2 From File: VHPNPROF.002 l

Reading Probabilities for 4 Nozzles in Group 3 From File: VHPNPROF.003 I

Reading Probabilities for 4 Nozzles in Group 4 From File: VHPNPROF.004 Reading Probabilities for 2 Nozzlee in Group 5 From File VHPNPROF.005 Reading Probabilities for 2 Nozzles in Group 6 From File VHPNPROF.006 i

Reading Probabilities for 3 Nozzles in Group 7 From File: VHPNPROF.007 I

Reading Probabilities for 5 Nozzles in Group 8 From File VHPNPROF.008 Reading Probabilities for 5 Nozzles in Group 9 From File: VHPNPROF.009 Iteading Probabilities for 28 No'.zles in Group 10 From File: VHPNPROF.010 i

l 1

e l

i l

l l

B-5 b

1 i

j

WESTINOHOUSE VESSEL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON ESBU-NSD 65 Nozzles at Virgil C. Summer Plant on 06-1C 97 06/06/91 CYCLE MAX-PROB PROB-ONE AVG-PROB NPVFC-05 NPVFC-50 NPVFC-95

- a.b 14 15 l

i 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 3B 39 i

40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 i

56 57 SB 59 60 l

B-6

l l

l l

1 WESTINGHOUSC VESSEL HEAD PEN. NOZZLE ECONOMIC DECISION ANALYSIS VHPNECON t

ESEU-NSD 65 Nozzles at Virgil C. Summer Plant on 06-10-97 06/06/97 CYISI NPV-CISI NPV-CREP NPV-CBEN NPVTC-05 NPVTC-50 NPVTC-95

- a.b l

14 15 l

16 I

17 18 19 20 21 22 I

23 24 l

25 26 27 28 29 30 31 32 3 *>

34 35 36 37 38 l

39 40 41 42 43 44 45 46

(

47 l

48 49 50 l

51 52 53 54 55 56 57 58

=

59 l

t i

B-7