ML20078B321

Nonproprietary Rev 1 to Brunswick Steam Electric Plant,Unit 1 Feedwater Nozzle Fracture Mechanics Analysis

ML20078B321
Person / Time
Site:	Brunswick
Issue date:	05/31/1991
From:	Stevens G GENERAL ELECTRIC CO.
To:
Shared Package
ML20078B301	List: BSEP-95-0012, Forwards Nonproprietary Version of Brunswick Steam Electric Plant,Unit 1 Feedwater Nozzle Fracture Mechanics Analysis, NEDC-30634,Rev 1.Requests That Copies of Document Originally Submitted by Util Be Destroyed ML20078B321
References
NEDC-30634, NEDC-30634-R01, NEDC-30634-R1, NUDOCS 9501250349
Download: ML20078B321 (42)

Text

_ __

O GENuclearEnergy l

175CemerAuenn Sn.kse.CA sst2s l

NEDC-30634 Revision 1 DRF B11-00491 Class III May 1991 i GE NUCLEAR ENERGY l

BRUNSWICK STEAM ELECTRIC PLANT, UNIT 1  !

FEEDWATER N0ZZLE FRACTURE MECHANICS ANALYSIS G. L. Stevens 7

Approved:

o~ ~ .w.

~ fr S. Ranganath, Manager i Materials Monitoring and Structural j Analysis Services 1 l

1 1

4 9501250349 950116 PDR ADOCK 05000325 P PDR

I 4 NEDC-30634 REY. 1 4

k j

i *

{~~[ IMPORTANT NOTICE REGARDING 4

CONTENTS OF THIS REPORT PLEASE READ CAREFULLY i

This report was prepared by the General Electric Company (GE) solely for

! the use of Carolina Power and Light Company. The information contained in ,

this report is believed by GE to be an accurate and true representation of the i

fac'ts known, obtained or provided to GE at the time this report was prepared.

l I The only undertakings of the General Electric Company respecting informa-tion in this document are contained in Carolina Power and Light Company Work 9

Authorization No. ZS70020028 and nothing contained in this document shall be construed as changing said contract. The use of this information except as defined by said contract, or for any purpose other than that for which it is intended, is not authorized; and with respect to any such unauthorized use, ceither GE nor any of the contributors to this document makes any representa-tion or warranty (express or implied) as to the completeness, accuracy or use-1 fulness of 'the information contained in this document or that such use of such information may not infrinse Privately owned rights; nor do they assume any q responsibility for liability or damage of any kind which may result from such

] use of such informacion.

4 1

1 2,

i.

i

ii

O f f

i NEDC-30634 REY. 1 i

i

CONTENTS i i

I

\

Page {

a ABSTRACT y-1

{' 1. INTRODUCTION 1-1 1

! 1.1 Background 1-1  !

) 1.2 Objective 1-1  !

1.3 Technical Approach 1-2

1.3.1 Feedwater Flow Cycling 1-2 i t

1.3.2 Thermal Cycling 1-2 ,

2.

SUMMARY

AND CONCLUSIONS 2-1  !

I 3. THERMAL CYCLE DEFINITION 3-1 f

4 4. THERMAL ANALYSIS 4-1 i

l I 5. THERMAL AND PRESSURE STRESSES 5-1 '

}

6. CRACK GROWTH ANALYSIS l

6-1 j 6.1 Stress Intensity Factor Calculations 6-1 ,

6.2 Crack Growth Data 6-5  ;

l 6.3- Crack Growth Evaluation 6-5

7. RESULTS AND CONCLUSIONS 7-1 I 8. REFERENCES

8-1 1

APPENDIX - THERMAL BOUNDARY CONDITIONS A-1 4

i-1 a

l 1 l 1 l' 1

a i ,

i b

b iii

- l

'NEDC-30634 REV. l'

{

t l

TABLES- I Table . Title Page, l'

.. ., 5-1 Surface Stresses to Choose Maximus Combined Stresses 5 ,

5-2 Limiting Stress Profile (Cross Section 4-4) 5  !

ILLUSTRATIONS:

Figure Title g 3-1 ' Temperature Cycling for 7/1/87 Scras 3-4 ,

3-2 Temperature Cycling for Planned Shutdown of 1/23/88 3-5  !

3-3 Temperature Cycling for 5/21/88 Maintenance Shutdown 3 3-4 Temperature Cycling for 7/13/88 Forced Shutdown 3-7 i

~

3-5

! Temperature Cycling for 11/10/88 Refueling Shutdown '3-8 l l

4-1 Nozzle Configuration  !

4-2 ,

l 4-2 Thermal Boundary Conditions i

4-3 5-1 Feedwater Nozzle-Brunswick 1 Location of Maximus Surface Stress (Steady State) 5-4 4 5-2  !

Feedwater Nozzle-Brunswick 1 Location of Maximus Surface '

Stress (Transient t = 4 minutes) 5-5 6-1 Boundary Integral Equation / Influence Function Magnification Factors for BWR Feedwater Nozzle 0-2 6-2 Stress Intensity Factor versus Crack Depth (Thermal Stresses, 4 minutes) 6-3 6-3 Stress Intensity Factor versus Crack Depth (Pressure Stresses) 6-4 6-4 Reference Fatigue Crack Growth Curves for Carbon and Low Alloy Ferritic Steels 6-6 7-1 Crack Depth versus Number of Years 7-2 iv

l l- i i

l NEDC-30634 REV. 1 i (

ABSTRACT The current revision of this report is based on actual feedwater cycling

, data collected for Brunswick 1 since the original (July 1984) revision of this report.

This report provides a plant-specific fracture mechanics assessment of the Brunswick 1 feedwater nozzles to show compliance with NUREG-0619 as ,

amended by NRC Generic Letter 81-11, dated February 20, 1981. The evaluation was based upon (1) the plant operating history supplied by Carolina Power and -

Light Company (CP&L), (2) low feedwater flow characteristics determined from actual plant feedwater cycling measurements, and (3) Moss Landing test data.

The evaluation considered an initial crack depth of 0.25 inch as specified in NUREG-0619. The results show that stress cycling from actual temperature and t flow profiles results in the growth of an initial 0.25-inch crack to less than 1 inch during the remaining life of the plant. Using the 1989 ASME Section XI fatigue crack growth curves, the analysis shows that the postulated 0.25-in.

crack becomes 0.56 inch deep after the 40-year plant design life.

i j These results demonstrate full compliance of the low flow feedwater 4

controller with the requirements specified in NUREG-0619.

r l

l l

l v l i

l

NEDC-30634 REV. 1

1. INTRODUCTION This report provides a plant-specific feedwater nozzle fracture mechanics assessment based on the Brunswick Steam Electric Plant, Unit 1 (hereaf ter . ...

called Brunswick 1) plant operating history and actual feedwater cycling data.

This is in response to the Nuclear Regulatory Commission (NRC) requirements regarding feedwater nozzle crack growth. These requirements are contained in i

the NRC Generic Letter 81-11 which states that a fracture mechanics evaluation

, must predict an end-of-design-life crack size of 1 inch or less.

1.1 BACKGROUND

The General Electric Company (CE) feedwater nozzle final report (Refer-ence 1) recommended design and operational changes to minimize both the proba-bility of crack initiation and rate of crack growth in feedwater nozzles. The low flow feedwater controller discussed in Reference 1 would not significantly reduce the probability of crack initiation, but would reduce crack growth.

The NRC (NUREG-0619) accepted the GE recommendation (Reference 1) and required that operating reactors install a low flow feedwater controller with the char-acteristics described in Reference 1 and reroute the Reactor Water Cleanup System (RWCS) flow to all of the feedwater lines. The low flow controller required above must meet strict requirements specified in Subsection 3.4.4.3 of Reference 1. The NRC later clarified its position in Generic Letter 81-11, stating that plant-specific analyses may be performed to justify not imple-menting such modifications.

With respect to low flow feedwater controller installation assessment, feedwater nozzle crack growth rate analysis is required for Brunswick 1. '

1.2 OBJECTIVE Because of the absence of Brunswick 1 low flow feedwater controller data, temperature measurement hardware was installed subsequent to the July 1984

, revision of this report. The data collected by that hardware for the last 1-1

e s NEDC-30634 REV. 1 fuel cycle was utilized as a basis for defining the actual Brunswick i feed-water cycling characteristics used in the current analysis. The data col-  !

lected was considered typical for the entire design life of the reactor. i This report provides a plant-specific fracture mechanics assessment of the Brunswick 1 feedwater nozzles to show compliance with the requirements of NUREG-0619 as amended by Generic Letter 81-11, dated February 20, 1981. The purpose of this analysis is to determine whether stress cycling from actual cootroller temperature and flow fluctuations will result in a final crack depth of 1 inch or less during a 40 year plant design life. The evaluation considers an initial crack depth of 0.25 inch as specified in NUREG-0619.

! 1.3 TECHNICAL APPROACH This analysis evaluates the growth of a 0.25-inch crack over a projected 40 year plant design life.

1.3.1 Feedwater Flow Cycling The feedwater flow cycling was determined from actual feedwater tempera-ture and flow data obtained from CP&L (Reference 2).

1.3.2 Thermal cycling The thermal cycling of the fluid at the.feedwater nozzle was determined

! from the actual feedwater data. The number of startup/ shutdown and scram events for Brunswick 1 was linearly projected based on actual plant operating history during the first 12 years (1976-1988) of plant operation (References 2, 3 and 4).

The thermal boundary conditions used in this analysis differed from the Reference 1 document in that thermal sleeve annulus temperatures and annulus heat transfer coefficients, derived from Moss Landing test data (Reference 9),

were used to calculate the thermal stresses in the feedwater nozzle.

1-2

. . -~ . - - . . -- . . . -- - - ..

2 - .

d b

I

NEDC-30634 REY. 1-  !

l 4 i i

4 To evaluate the crack growth, thermal and pressure stress analyses were I conducted using the finite element computer code ANSYS (Reference 6). The locations of the peak thermal stress, peak ~ pressure stress and peak combined i thermal and pressure stress were determined and the crack growth-was calcu -

lated using a crack growth computer code. The crack growth relationship used represents the 1989 ASME Section XI Code Curves. The best-fit correlation to )

actual PWR and BWR data used in the original revision of this report was not t used, since the R-ratios (K,g,/K,,,) were typically high for the actual cycling data; that correlation is not valid for high R-ratios.

i i

?

8 f

L 1

t 1-3 i

_m. - . _ _ _ _ . _ __._. _ . _ _ . _ . . _ _. _ _

e e

[

l.

l

• NEDC-30634 REV. 1

2.

SUMMARY

AND CONCLUSIONS i

Application of the ASME Code,Section XI crack growth rate relationship '

l resulted in crack growth less than the acceptance criterion of 1 inch for a l

i 40 year plant life. The analysis resulted in a 0.56-inch crack depth after i

the 40 year plant design life. '

l r

This analysis is based on actual low flow feedwater controller character-istics obtained since the original revision of this report. The plant i operating history is based on the initial 12 years of operation extrapolated to 40 years. Because of " learning curve" effects which are typically - i experienced by operating reactors during their initial years of operation, the  !

extrapolation is most likely conservative.

l  :

j With regards to RWCU reroute, a plant-specific analysis was performed for

• I

! Brunswick 1 which demonstrates that RWCU-reroute leads to only a small improve-l , ment on thermal cycling and fatigue usage of the feedwater nozzle region '

(Reference 14). Based on that analysis, it was concluded that monitoring thermal sleeve seal is more important than RWCU reroute. ,

Therefore, based on the results presented herein, it is concluded that  !

the Brunswick 1 low flow controller f ully meets the requirements _ of NUREG-0619.

l Periodic examination, required by NUREG-0619, of the feedwater nozzle will >

I provide additional justification of these analytical results.

l l

l l

i 1

1 l

l l l

l 2-1 i

\

le  :.

. .- i l

l NEDC-30634 REV.1 i i

3. THERMAL CYCLE DEFINITION I

l The feedwater nozzle thermal cycle. definitions are represented by. Figures 3-1 through 3-5. These figures represent the minimum and maximum temperature

{

points for one scram event and several startup/ shutdown events obtained from l CP&L for Brunswick 1. The feedwater loop which has the RWCS mixing (Loop B) had the most severe cycling and was therefore used throughout this analysis. i

{ The following events appropriate for this analysis were identified- by CP&L personnel from the last available fuel cycle:

Figure Date Event Description 3-1 7/1/87 Scram 3-2 1/23/88 Planned Shutdown 3-3 5/21/88 Maintenance Shutdown a

3-4 7/13/88 Forced Shutdown 3-5 11/10/88 Refueling Shutdown These events are depicted in Figu;es 3-1 through 3-5, and were digitized into computer form from microfilmed strip chart recordings of feedwater temperature. The temperature measurements were taken by hardware installed subsequent to the original issue of this report, and are symbolic of the cycling occurring at the feedwater nozzle. The events are shown exactly as digitized from the microfilm recordings; as a consequence, actual progress of -

each event is from right to left. '

A total projection of 163 startup/ shutdown events and 323 scram events over the 40 year plant life waa made for Brunswick 1 based on operating data obtained from the first 12 years of operation (References 2, 3 and 4). This [

projection was determined using the methodology of Reference 5 as follows:

i f i 6

0 3-1 l  !

! [

I

e .

s NEDC-30634 REV. 1 Number of Number of Year Time Period Startups/ Shutdowns Scrams

• l 1 11/76 - 11/77 2 29 2 11/77 - 11/78 1 21 3 11/18'Ill'/79' 5 17 j 4 11/79 - 11/80 7 10 5 11/80 - 11/81 3 10 6 11/81 - 11/82 5 15 -

7 11/82 - 11/83 4 7

! 8 11/83 - 11/84 7 9 9 11/84 - 11/85 3 4 10 11/85 - 11/86 0 8 11 11/86 - 11/87 6 1 ,

12 11/87 - 11/88 5 3 12.5 11/88 - 5/89 2 1

• Note: Turbine trips identified in Reference 4 were counted as scram events.

Although conservative, this method of counting is not considered to be a significant contributor to the final crack growth results.

l Based on the methods of Reference 5, the following equation is used to determine the projected number of events for the 40 year design life:

n49 n4 + n12.5[36/(12.5-4)]

where n " Projected number of events at year 40 40 l n 4 = number of events at year 4 n12.5 = number of events between year 4 and year 12.5 l Thus, for startup/ shutdown events, the following is obtained:

l i

n 40 - 15 + 35(36/8.5) - 163 i

3-2 l

l l

o .

NEDC-30634 REV. 1 l and for scram events:

l l ,

n4o = 77 + 58(36/8.5) = 323

For the purposes of this analysis, a single cycle is defined when the nozzle fluid temperature, initially at some value T ,Ochanges to some other value Ty and then returns to T0*

i For the purposes of defining complete startup/ shutdown cycles, the start-ups corresponding to each of the four shutdown events (Figures 3-2 through .

3-5) were assumed to be mirror images (i.e., identical) to the shutdown events. Therefore, four (4) complete startup/ shutdown cycles and one (1) complete scram cycle define the thermal cyclic duty experienced by Brunswick 1. '

These events were assumed to be typical for the entire 40 year design life of  ;

the reactor. i

+

l i

l s

l 3-3 1

1 1

, , , . - .r.

nRgb E? e 6

1 1

r i

n r ,

i {,I l i

m a

r c

S 7

8

, i ,, ,

r

/

u 1

/

7 r

, o f

' i g _

n

) i s

r l

c u y o C H

( e _

o r u

t Tim a r

e p

m e _

T _

1 _

3 e

r u

g i

i F

-; - - 0 0 0 0 0 0 0 0 0 0 0 0 5 4 3 2 1 a

.je i .

,A

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

% M 500 400 -

Ei

" R e 300 -

6 Y

t o g 2

y ll M

80 ll h k '

l  ?

5 200 - '

, " e H ,

c, i f 100a 'u n

O , , , , , , , , , , , , ,

Time (Hours)

Figure 3-2. Temperature Cycling for Planned Shutdown of 1/23/88

. t i

t r

500 i i

4 t

4 i

• g_

l E

N e

s_, 300 - 8 y . 6 '

• 11 R

as I

t I-

m. .

F.*

100 -

4 i

0 , , , , , , , , ,

O 45 Trne (Hours)

Figure 3-3.

Temperature Cycling for 5/21/88 Maintenance Shutdown 1

< r

. - . . . . . . - _ . - - _ _ _ _ _ _ _ - _ _ _ - - . .._ --- -._.. . ....,. --. - ,, ,-,...n. - - . - . - - -,n-e-..----,.,--+_-,---, , ., ., . - - - - - - - -- - , , , . - - . . , . . ~ - _ _ _ _ - - - - -

l

?uy n<, g 0

5

. l' ' 1

, n .

w .

o .

d t

u h

S a " ,

d

' 'q '

F e

c r

o i

8

,' ,, ' l

, 8

/

3 1

/

7

" t r

e o

, ) f

_ s r g u n

. o i H

( l e c y

n a C T e -

r u _

t _

a r -

e p

m

_ l e

jF i i s

T 4

_ s "

i ,

i

's i

3 s I ii e

ie ,

r u

• g i

i j

s F

_ l , Q i i'

" i 0 0 0 0 0 O 0 0 0 0 0 5 4 3 2 1 C*,, .ga5h w4

_ __ . . __ . . - _ . ~

~ ~ '. l 500 il a un 400 - * ' i t i '

l l lis ij n l' I i i

C \0 ' '

L 300 - i u ll iN l i l

E g

e l l i u i a

f Y

=

$ i

's i '

e si n 0

. g " u "

i nj 4-5 200 - ,

• n M F "i f *

, , i i .

[ 11 g i

100 -

0 , , , , , , , , , , , , , , ,

35 Time (Hours)

Figure 3-5.

Temperature Cycling for 11/10/88 Refueling Shutdown

. . . - - . .- _ ~

j *

• i NEDC-30634 REV. 1 i

4. THERMAL ANALYSIS The finite element computer code ANSYS (Reference 6) was used to develop i an axisymmetric model which simulated the Brunswick 1 feedwater nozz1t.. The i

~

~ 'so' i parametric heat conduction element (STIF $5) was used.

The model was '

developed using the nozzle configuration shown in Figure 4-1 (Referer.ces 7 and .

8). The same model with an isoparametric stress element was subsequently.used for the stress analysis. Further. discussion of the model configuration is included in Section 5.

The heat transfer coefficients and temperature boundary conditions were l

derived from Reference 9. The. method of derivation is explained in the Appen-dix to this report. The use of annular temperatures and heat transfer coefficients removed the necessity of specifically modeling the thermal sleeve in the finite element analysis. The feedwater nozzle thermal sit: eve design is a single sleeve welded to both the feedwater nozzle safe end and to the feedwater sparger. These heat transfer coefficients with the appropriate I temperature boundary conditions are shown superimposed upon a derswing of the -

finite element model in Figure 4-2.

The initiation of feedwater flow was modeled by varying the temperatures in Zones 2 and 3 from 550*F down to the temperatures indicated in Figure 4-2, '

over a 3-see interval. The temperatures were maintained et this level until j steady-state conditions were reached. The 3-see ramp was used rather than a step change, since it is conservative and still assures numerical stability in the computer solution.

The finite element model was regenerated and executed for the current analysis to assure identical results to those obtained predously in the original analysis.

4-1

> r i s *  !

e .

dgwe <

S

. S l

E A N NI DKC

,a* '

• l sMOLH AI yAM aNCT rI nD

-

. =

5 7

5 9 6

2 N_ M b _

n -

d 3

8 r o

I i .

6 t a

r

_ 8 J u 3 g

. i

o. S f 4 i E 4

H n 9 C o

_ 1 N C s I N e

_ L I l

E z q R z E

A o L-S N N _

Z.

Z O I

- O S .

N. N 1

. E -

. M 3 3 4 6 4 I

7 D e _

L r L u

- A g _

= 5 E

i

- 4 T F _

- O _

8 8

N _

1 C C +

+

N MI 2

1 2 -

+ -

j o.

s y >J -

j . .

I NEDC-30634 REV. 1 l

l I

i I

26.3 HEAT HE AT TRANSFER 800NDARY TRANSF ER COEFFICIENT F LulD TEMPERATURE ZONE (8tw/hr h2 or3 (op) 23.0 --- d 1

' 1 2

864 790 550 212.5-464.5 3 790 464.5-550 4 o.2 120

_\,

.._ h's N

13.s -

\

l <

h

( y (2) l\ 4 fi h 4 /

G l 2

l ,., _

.____  :- == b i

\

17 - 1 1

p

\ p Y l

i Y

, /

/

1f N

s )

3 ,,,

1 1

l _ga I t I i f I I 4.0 10.2 t s.s 22.s 29.o 35.3 41.5 INCHES l

l Figure 4-2. Thermal Boundary Conditions 1

4-3 i

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

I NEDC-30634 REV.'l  !

1

5. THERMAL AND PRESSURE STRESSES l i

The results of the thermal analysis were applied to the previously men- j tioned finite element stress model to determine the thermal stresses. ' Iso-  !

parametric stress elements.(STIF 42) were used in the stress analysis. The  !

nozzle was modeled by an arisymmetric finite element mesh with the vessel being represented as a spherical shell. This is a common approximation used i in the stress analysis of a three-dimensional nozzle configuration in a cylin-  ;

-drical shell. This was adequate for thermal stresses, but pressure stresses  !

t required a scaling factor based on a three-dimensional analysis.- The lengths f of the nozzle safe end and pressure vessel section were each modeled to at l least 2.5 /Et, where r is the radius and t is the thickness of the nozzle. >

This was done to assure that end effects did not influence the~ stresses in the  ;

nozzle corner. P I

l Thermal stresses were evaluated during several time intervals over the  ;

course of the transient by analyzing node pair temperature differences at var- i ious locations.in the nozzle blend radius. Only the steady-state stresses and'  ;

the stresses occurring at 4 minutes were used in the subsequent crack growth  !

analysis, since they resulted in the most limiting stress profile in the noz-zie blend radius region. The highest thermal stress occurs on the inside sur-f ace of the nozzle blend radius as shown in Table 5-1. The thermal stresses which developed from a AT of 450*F are linearly scaled to the AT described in the thermal cycle definitions (Section 3). The scaled thermal stresses are subsequently used in the crack growth analysis.

The maximum thermal stresses occurred in the stainless steel cladding area of the nozzle blend radius (ends at Element 337; Figure 5-1).

Pressure stresses for the case.of a 1000 psi vessel pressure were also calculated; however, these stresses required application of a scaling factor.

This was necessary because of the limitation of modeling a three-dimensional l

component with a two-dimensional axisymmetric model. Because the three-dimensional characteristics near the nozzle corner were not modeled, the peak  !

5-1

.j NEDC-30634 REY.'1'-

Table 5-1 SURFACE STRESSES TO CHOOSE MAXIMUM COMBINED STRESSES (Steady State)

Thermal Hoop Pressure Hoop Pressure Ratioed Combined }'

Element (psi) (psi) to 1.5004 (psi) (psi) i 169 835 -27196 40805 41640 f 185- 4910 28391 42597 47517 l 193 9178 29065 43609 52787  !

209 13412 29519 44290 57702.  !

217 17395 29725* 44600* 61994-  ;

233 20809 29661 44503 65312 241 23795 29295. 43954. 67749  !

257 26598- 28664 43007 -69605-I 265 29168 27663 41506' '70674' 281

.f 31721' .26293 39450 71171* {

289 33700 24462 36703

< 70403  !

305 35642 22808 34221 '

69863  ;

313 37359 21232 31856 5 69215 329 39081* 19821 29739 68820 ,

337 38874 18380 27577 66451 353 27304 16725 25094 52398  ;

369 25130 14798 22203 47333

• Maximum stress i l i i

i l

5-2 1

[

t t

l NEDC-30634 REV. 1 t

l Table 5-1 j SURFACE STRESSES TO CHOOSE MAXIMUM COMBINED STRESSES (Continuid) <

l (4 min after beginning of transient) l l

l Thermal Hoop Pressure Hoop Pressure Ratioed Combined Element (psi) (psi) to 1.5004 (psi) _( psi)  ;

169 -2015 27196 40805 38790 l

?

185 2423 28391 42597 45021 i l 193 7066 29065- 43609 50675 l

[

209 12018 29519 44290 56308 217 16853 29725* 44600*- 61452 233 21289 29661 44503 65792 241 25421 29295- 43954 69375  ;

257 29476 28664 43007 72483 265 33343 27663 41506 74849  !

281 37813 l 26293 39450 77263 i 289 41970 24462 36703 78673 305 45883 22808 34221- 80104 i

313 49460 21232 31856 81316 r

j 329 53070 19821 29739 82809* t 337 54279* 18380 27577 81856 353 46090 16725 .25094 71184 369 45903 14798 22203 68106 1

l l

• Maximum stress 4

t i i

5-3

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

' . . i NEDC-30634 REV. 1 l

1 l

)

r i

i 4  !

- ~  !

i d Il IiIl\\\l i

1 3

1

(

/

/% x 1

I f\\

N '

} 1 l

/ \ \ \\

K i \ \\ \ \

/ w, i \ \\ \ \\

'58

/

• i i \\ \ \\ -

I L \ \\ \ \\ \- l

/ iU I

X\ A Y\ O k I \ \ \ \ \ \\ \ \ 186 188 y88 i N.\\\

L .., \ \ \\\ \- 21, 447 434 421 400 306 3W 337 313 2W 205 34I M 23 329 305 - 2 MAXIMUM PRESSURE b 1 281 STRESS: 44600 pel a

MAXIMUM r MAXIMUM THERMAL COM81NED STRESS: 39001 psi STRESS: 71171 psi i

Figure 5-1.

Feedwater Nozzle-Brunswick 1 Location of Maximum Surface Stress (Steady State)

I 5-4 j

4 m

l

• * \

l l.

1 l

NEDC-30634 REV. 1 I

I I

i l

l l

i 3

4 1

/ xN \

I i i

)I\

l \\\

/ -

x l

\ \\\\

/ x l \\\\ \ \

-i \\\\ \- \

'55

^

i l /

/ N i \\\\ \ \

l \\\\ \ \ -

\ _1\ Y\ /\ /\ -

$a l l

/ \ \\\\\ \ \ \ \ 185 N -111 \ 1 '. T x \s -

j j g1 1 1 ' ' ' '

gg3 200 l l N. I .'. I . 217

! g 447 434 421 400 395 3W 337 \ 313 2W 295 257 382 33 329 305 g AXIMUM PRESSURE

(, STRESS:44800 psi MAXIMUM 4 MAXIMUM '

' THERMAL ~ COMBINED STRE$$: 54279 poi STRESS: 82009 psi i

l,

?

i I

! Figure 5-2.

l Feedwater Nozzle-Brunswick 1 Location of Maximum I Surface Stress (Transient t = 4 minutes) l i 1

5-5 l

l l

1 NEDC-30634 RIV. 1  !

stresses at the nozzle corner were not accounted for accurately. Therefore, a generic three-dimensional model developed by Gilman and Rashid (Reference 10) was used to scale the pressure stress. The scaling factor for the pressure; stress is given by the ratio of the peak pressure stress on the inside surface reported by Gilman and Ra'shid to the peak pressure stress on the inside sur-face from the finite element model used in this report. The peak pressure stress of the finite element model was 29,725 psi, while the peak pressure stress reported by Gilman and Rashid is 44,600 psi.. This resulted in a scal-ing factor of 1.5004. The scaled peak pressure stress on the inside surface  !

l is shown in Figure 5-1.

a l

i The combined thermal and scaled peak pressure stresses were examined to determine the area where the combined peak stress on the inside surface occurs, as shown in Table 5-1.

The stresses at the cross section associated I

with the limiting stress profile-(see Table 5-2, and cross section 4-4 on l Figure 5-2) were used to calculate crack growth.

i 1

i 5-6

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

l. ,

NEDC-30634 REV. 1 i

i Table 5-2 '

LIMITING STRESS PROFILE (CROSS SECTION 4-4)

Distance from Pressure Hoop Thermal doop Inside Surface - - Ratioed to 4 minutes (in.) 1.5004 (psi) (psi) l 0.0 27577 54279 0.074 27028 52036 0.229 26325 35698 0.397 25241 32438 0.596 24161 29058 0.846 23049 24536 1.147 21849 20309 1.501 20587 15354 1.886 19843 12675* r 2.336 18252 10057 3.004 16570 7723 3.632 15084 5688 4.251 13705 3878 j 4.882 12360 2227 5.501 10975 669 .

I r

I l

• Steady-state thermal stresses were used from point **" to the outside surface j of the nozzle because the steady-state thermal stresses result in greater crack growth.

l l

u l

5-7 1.

.-- - . . .-. - - .- ---. . - .. . . ~ .. - - - . - . . - . . -

NEDC-30634 REY. 1 L

i

'l I '

6. ' CRACK GROWTH ANALYSIS  :

1 1

6.1 STRESS INTENSITY FACTOR CALCULATIONS i

" Itress intensity factors were determined.using solutions for standard stress distributions (e.g., constant, linear, quadratic, and cubic variations)'  !

and applying the superposition principle. The stress intensity solution for +

these unit load cases was expressed in terms of the crack length and'appropri-  ;

ate magnification factors for the' specific crack geometry (Figure 6-1). The.

stress intensity for an arbitrary stress distribution was then obtained by fitting a third-order polynomial of the form:  ;

o=A 0

AX+A y 2

+A[3 l

8

. and applying the principle of superposition. Once the curve fit parameters l

{ A' O l' A ,2 and A3 were known, the stress intensity factor was determined as i

, a function of crack depth using the equations in Figure 6-1.

i Magnification factors for several common two-dimensional geometries are >

available in References 11 and 12. For the feedwater nozzle, a set of three-dimensional magnification factors is presented in Reference 1.' As-illustrated in Figure 6-1, the nozzle corner factors (0.706, 0.537, 0.448, and 0.393) were l obtained by averaging the magnification factors developed for circular surface  ;

crack geometries in half and quarter spaces. This expression (labeled' FUN 11)

]

was used to calculate stress intensity factors in the fracture mechanics eval-uation which follows.

I i

l The pressure and thermal stress distributions were fit to third-order polynomials using a standard least squares procedure. Overall accuracy of the polynomial representations is considered more than adequa*,e.

Substituting these polynomial coefficients (A O ' '1, A2 ""O 3) into the TUN 11 stress intensity factor expression of Figure 6-1 leads .to the stress intensity factor versus crack depth data shown iu Figures 6-2 and 6-3. (These stress intensity factors apply to cross section 4-4.) j

!. 6-1 i

l 1

,m. , - . .r. - - - - , .,-1

i NEDC-30634 REV. 1 t 1

i l

' ZX ,

l 2

4 5

i

_ -g-FUN 9 - SEMI-CtRCULAR CRACK IN HALF SPACE .

2 3 K, = $ (0.688 A, + 0.522 Ga/r) Ag + 0.434 (a /2) A2 + OJ7714a /3sl A 31 i

1 "

i X

i

{ FUN 10 - OUARTER CIRCULAR CRACK IN OVARTER-SPACE 2 3 Kg = S10.723 A, + 0.551 ca/s) A, + 0 462 (a /2) A2 + 0.40014a /3s) A3 ]

I ,

I i

I I i

! I l 4

1 l

j i

\ /x 3

a s

N.

FUN 11 - SIMULATED 3-0 NOZZLE CORNER CR ACK 2

Kg = Ml0.706 A, + 0.537 Qa/w) A g + 0.448 (a /21 A2 + 0.39314aI /3e) A3 ]

Figure 6-1. Boundary Integral Equation / Influence Function Magnification Factors for BWR Feedwater Nozzle 6-2

e .

50 40 -

~i AT = 450*F (550 - 1008F1 j 4 MINUTES AFTER TRAN$ TENT E 30 -

B 4

B 0

CROS54ECTION 44 H

E

[

o f

w

• 5 H  %

n 20 -

N E

• t;

• 10 -

0 I I I I I I I I I I I I I I O 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 28 3o CRACK DEPTH DISTANCE ALONG SECTION fin.1 Figure 6-2.

Stress Intensity Factor versus Crack Depth (Thermal Stresses. 4 minutes)

b 50 F

40 -

s n

5 a

d

~

Z.

O $

D PRESSURE = 1000 pel

< u o

• o A E O g CROSS SECTION 44 g 2o -

t; J

10 -

0 I I I I I I I I I t g g g ,

0 0.2 0.4 0.6 0.8 3,2 t.0 3.4 1.6 1.8 2.0 2.2 2.4 2.6 28 3.0 CR ACK DEPTH DISTANCE At.ONG SECTION (in.)

Figure 6-3. Stress Intensity Factor versus Crack Depth (Pressure Stresses) 4

l i

NEDC-30634 REV. 1 l

l 6.2 CRACK GROWTH DATA l

7 Figure 6-4 represents the fatigue crack growth data for low alloy steel from Section XI of the ASME Code (Reference 13). The R-ratio (K,1,/K,,,)

l l

dependence of this data is built-in by representing three cases: (1) R-ratio l 1ess than 0.25, (2) R-ratio between 0.25 and 0.65, and (3) R ratio greater than 0.65. These data were used to determine the growth of an assumed l

0.25-inch initial depth crack.

l The best-fit compilation of fatigue crack growth data used in the origi-nal revision of this report was not used in the current analysis. That rela-tionship is not valid for high values of R ratio. Much of the data obtained for Brunswick 1 yielded high R-ratios (>0.9). As a result, unrealistic i

crack growth would result from the use of this relationship.

6.3 CRACK GROWTH EVALUATION The thermal cycle definitions are represented by Figures 3-1 through 3-5 for startup/ shutdown and scram / return to full power events. A projected total of 163 startur .autdown events and 323 scram events was made for Brunswick 1 over the 40 year design life of the plant as described in Section 3.

The analysis conservatively assumed that the initial crack depth of 0.25 inch included the cladding thickness. Since the thermal stresses are higher in the stainless steel cladding region, the corresponding stress intensity factor would also be greater, thereby resulting in a more rapid crack growth propagation. 1 I

l The procedure for cal'eulating the crack propagation is as follows: For each cyc1c, the msIimum and minimum stress and the number of occurrences were calculated. Frcm this, the stress intensity factor range and corresponding R-ratio were calculated for the cycle being analyzed. Using this and the selected crack growth relationship, the incremental crack growth was 6-5

4 $

NEDC-30634 REV. 1 10C0

/

/

700 -

• LINE AR INTE RPOLATION 15 RECOMMENOEO TO ACCOUNT &

[ /

FOR RATIO OEPENOENCE OF [

WO - WATER ENVIRONMENT CURVES.

FOR 0.25 < R < 0.65 FOR E[

/

I SH AL LOW 5 LOPE es/dN * (1.01 s 10~'l 02 A K8$ U'

&2 [

j O2* 3.75 A + 0.06 R* K,,,M,,,

[&d 200 -

/

1 /

E e

~-

$UBSURFACE FLAWS &

100 - (AIR ENVIRONMENTI !E 1 i

da/dN

• 10.0267 a 10-3) A Kg 3.726 n

e

/ e:

s' e k ~ OfTERMINE THE AK AT WHICH

• 1

$ THE LAW CHANGES 6Y CALCULATION E j OF THE INTERSECTION OF THE TWO

[ 9 5 #

z CURVES l

i SURFACE FLAWS '

b (WATER RE ACTOR f y E NV6';ONME NTI i 4 20 -

APPLICABLE FOR b R < 0.25 l

f

• 0.25 < R < 0.65 f

R > 0.65 to -

p,g fg man f

8

/

5 - l 'LINE AR INTERPOLATION 18 l

(

,.?.

f "(' RECOMMENOEO TO ACCOUNT FOR R RATIO OEPENOENCE OF 2 f* WATER ENVIRONMENT CURVES.

l FOR 0.25 < R < 0.65 FOR STEEP k l $ SLOPE: '

i O 9 y f da/dN * (1.02 a 104) Og A K5M I f #

Og a 263R - 5.725 f

- 1 R = K,;,/K,,,

I l i I i 1  :

i 1 I 1 2 5 7 10 20 70 50 100

$TRES$ INTENSITY FACTOR R ANGE g(AK kalM

, Figure 6-4. Reference Fatigue Crack Growth Curves for Carbon and l Low Alloy Ferritic Steels 6-6 i

l l

)

NEDC-30634 REV. 1 i I

(

calculated for that event. The crack size was updated and the procedure  ;

repeated. This continued for every cycle until the entire life was analyzed.  !

The pattern of events was assumed to be:  !

41 sets of: '4 startup/ shutdown events followed by  !

8 scram / return to full power events '

(Figure 3-1). l

-i Note that 41 x 4 = 164 startups -and 41 x 8 =' 328 scrams. The effects of

• h modeling five extra scran events and one extra startup/ shutdown event, although-

• conservative, is considered small.  ?

The four startup/ shutdown events were further broken'down as follows:

1 1 1/23/88 Planned Shutdown event (Figure 3-2)

-f (including the mirror-image startup)  !

i 1 5/21/88 Maintenance Shutdown event (Figure 3-3)

(including the mirror-image startup) '

i 1 7/23/88 Forced Shutdown event (Figure 3-4)

(including the mirror-image startup) 1 11/10/88 Refueling Shutdown event (Figure 3-5)

(including the mirror-image startup) i i

l Total = 4 l

One crack propagation calculation was made corresponding to the limiting stress profile shown in Table 5-2.

6-7

=~ w- -

v n -m - <-m waw,e e -en

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

. r NEDC-30634 REV. 1

7. RESULTS AND CONCLUSIONS l Because of the recent acquisition of plant-unique data, the feedwater I i

nozzle thermal cycle definitions defined in Section 3 were assumed to be l representative of all startup/ shutdown and scram events. The number of events over the plant life as projected using plant-specific data (Referencee 2,. 3-  !

and 4) was utilized. A plant-specific finite element stress analysis was  !

performed for the feedwater nozzle. i The fracture mechanics analysis was based on the thermal stresses obtained from the finite element analysis, the thermal cycle definitions derived from actual plant feedwater data, and the historical frequency of-the i number of startup/ shutdown and scram / return to service events. I

{'

The results of the fracture mechanics analysis are given in Figure 7-1 E for the limiting location (cross section 4-4) as a function of the number of years since initial plant startup. 'f l

l Using the 1989 ASME Section XI fatigue crack growth curves, the analysis (

shows that the postulated 0.25-in. crack becomes 0.56 inch deep after the  !

40 year plant design life.  !

I r

The operating history used in the current analysis was based on an l j

extrapolation of the initial twelve years of plant operation. Because of

" learning curve" effects which are typically experienced by operating reactors {

during their initial years of operation, this extrapolation is most likely conservative (as recognized by the large number of startup/ shutdown events).

Based on this extrapolation and the utilization of temperature magnitudes considered typical for Brunswick 1 operation, this re-evaluation demonstrates  !

l full compliance with the requirements of NUREG-0619. Periodic examinations, f

as required by NUREG-0619, can provide additional justification of these analytical results.

7-1

1.2 1.1 -

i

1. 0 - Allowable Crack Depth = 1.0 inches----- - - -

I '

O.9 - '

e  !

g 0. 8 - . is e t ^ e 4 & Crack Depth at End of g

0.7 - o 40-year Rant Life = 0.56'

• g 8 0.6 -

. R, O 3 sa y .t; o.5 -

1 g g

~ a e y

O 0.4 - 8 y Q.

O.3 - f

• i "

O.2 - i i

O.1 - i O , . .

O 10 20 30 40 50 Number of Years Figure 7-1. Crack Depth versus Number of Years '

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

, a e l

NEDC-30634 REV. 1 4

8. REFERENCES J

1. NEDE-21821-A, " Boiling Water Reactor Feedwater Nozzle /Sparger Final Report," General Electric Company, August 1979.

2. Letter EGE-146, File BG0029A-AA-A500, J. M. Brown (CP&L) to G. L. Stevens (GE), ' Brunswick Steam Electric Plant, Unit Nos.1 & 2 NUREG-0619 Analyses, PCN C00029A Information Transmittal," November 17, 1989.

3. NEDO-22196, " Reactor Pressure Vessel Thermal Cycle Fatigue Evaluation for Brunswick Steam Electric Plant Units 1 and 2," General Electric Company, March 1983.

4. Letter EGE-164, File BG0029A-AA-A500, A. M. Lucas (CP&L) to G. L. Stevens (GE), " Carolina Power & Light Company, Brunswick Steam Electric Plant Units 1 & 2, Feedwater Nozzle NUREG-0619 Crack Growth Analysis Update,"

December 10, 1990.

5. GE Services Information Letter (SIL) Number 318, "BWR Reactor Vessel Cyclic Duty Monitoring," December 1979.

6. G. J. DeSalvo and J. A. Swanson, "ANSYS Engineering Analysis System User's Manual," Swanson Analysis Systems, Inc., May 1, 1989, Revision 4.4.

7. GE Drawing 767E723, Rev.1, "Feedwater Nozzle Safe End," July 1975.

8. Chicago l eidge and Iron Drawing Number 31, Rev 17, "12-In. Diameter Feedwater Nozzles," October 1971.

9. NEDE-21659-1, C. M. Kwong and H. Choe, " Moss Landing Feedwater Nozzle /Sparger Test Data Files," February 1979.

10. J. D. Gilman and Y. R. Rashid, "Three-Dimensional Analysis of Reactor Pressure Vessel Nozzles," Proc.1st Int. Conf. on Structural Mechanics in Reactor Technology, Vol. 4 Part G, September 1971.

11.

ASTM-STP-590, C. B. Buchalet and W. H. Bamford, " Stress Intensity Factor Solutions for Continuous Surface Flaws in Reactor Pressure Vessels,"

Mechanics of Crack Growth, American Society for Testing and Materials, 1975.

12.

~

ASIM-STP-590, R. Labbens, A. Pellissier-Tanon and J. Heliot, " Practical Method for Calculating Stress Intensity Factors Through Weight Functions," Mechanics of Crack Growth, American Society for Testing and Materials, 1975.

13. ASME Boiler and Pressure Vessel Code,Section XI, 1989 Edition.

14. NSEO-75-882, " Effects of Reactor Water Cleanup Reroute on Feedwater Nozzle Fatigue Usage, Brunswick Steam Electric Plant Units 1 and 2,"

August 1982.

8-1

NEDC-30634 REV. 1 l

l APPENDIX '

i THERMAL BOUNDARY CONDITIONS A.1 HEAT TRANSFER COEFFICIENTS i i

l I I

The annular heat transfer coefficients were developed from the data of Reference A-1 as follows. The data base consisted of four tests which were

! run at low feedwater flow and zero leakage flow. The heat transfer coeffi-cients during the tests were determined from eight heat flux meters mounted circumferential1y around a section of the nozzle blend radius.

P The highest heat transfer coefficient measured was taken and corrected to account for the difference in nozzle blend radius between the test sparger and the Brunswick 1 sparger. The Nusselt number, Nu, is proportional to the Reynolds number to the nth power, where n is typically 0.8. The Reynolds num-ber is in turn directly proportional to the nozzle blend radius, R. There-f ore, the Nusselt number is proportional to the nozzle blend radius raised to l

l t

the 0.8 power. In equation form, i

Nu o R

• 1 The heat transfer coefficient, h, is given by l

l h = Nu (k/R) '

i where k is the thermal conductivity of the fluid. Thus haR -0.2 l

1 This proportionality is used to correct the heat transfer coefficient. In the l tests of Reference A-1, h = 840 Btu /hr-f t

'F and R = 2 inches. For the Brunswick 1 sparger, R = 2.69 inches. Therefore, h = 790 Btu /hr-ft 2_,p, t

l l

l 1

l l A-1 1

a O I

. a .

NEDC-30634 REV. 1 GE PROPRIETARY INFORMATION CLASS III A.2 BOUNDARY TEMPERATURE CCNDITION Boundary (or annulus) temperatures were taken from the aforementioned I four tests. The fourth test was not used because of the abnormally high per-centage of steam carryunder. The test data is expressed in terms of a normal-ized temperature which is equal to the difference of the annular fluid temper-ature and the feedwater temperature divided by the difference of the reactor  !

temperature and the feedwater temperature. Readings are available at several circumferential locations at four sections of the nozzle. At each section,  ;

the lowest readings for each test were averaged to produce the final result.

The annulus fluid temperatures, as determined from these Moss Landing tests, are given in Table A-1 as a function of position of the welded sleeve nozzle configuration. The expression for obtaining the annulus fluid ttmpera-ture is as follows: .

T=TFW + C1 (Ty -Tpg) where l

T = annulus fluid temperature (*F)

FW = feedwater temperature (*F)

T Ty = vessel temperature (*F) l Cy = coefficient from Table A-1 1

A.3 REFERENCE l

A.1 NEDE-21659-1, C. M. Kwong and H. Choe, " Moss Landing Feedwater Nozzle /Sparger Test Data Files," February 1979.

I I l I

A-2 i

l l

t , 4 NEDC-30634 REV. l' i

t s

! Table A-1  ;

i.

C1 COEFFICIENT j (GE Proprietary)  ;.

i C-  !

Location. 1

. ~

F2 0.225-  ;

A 0.25 ,

3 0.54 C

0.81  !

Use linear interpolation between locations as illustrated in Figure A-1.

i l y r

o l

I c

n t

It

• 7 A-3 1

, :I h; '

n r

• 9 F

gtoI $* m t .

D D 7 -

l l

+

^ " o

• a W

r = l

% N e

+ + =N.

t 2

2 I

4 s

s C C r

1 5 e _

n L 1

c V) d1

~

o 8 A

- n- ~

T .

. A . .

aA L .

U 4 _ IN lc ee .

t e

S _

7 6 z zen -

7 L 1 o r ,

L A Nef -

. W re .

K eR C t .

I H

a , -

T m 7

ws d t s es

_ 5 2

ee .

FT .

f g .

o o on i nd A

ia on -

R tL O a S _ t s N _

no e s .

E

w. S mM ru m

. E

. C n A t o F o s r n

u S

s N

E

- nf I(

L s -

. A o t .

T E

u L , 1 e -

t F o o A QO e

- s sF2 i

r u

g a F W

O L

F F

O N

O I

T C

E M

t O

~

s

i e** 3

, *g ,

j i

NEDC-30634 REY. 1 l r

l l

DISTRIBUTION l

l Name M/C K. F. Cornwell 732 - ---

S. Ranganath 747 W. Yee BRU A. D. Ketcham BRU G. L. Stevens (15) 747 GE-NE Library (2) 528 l

!