._.. _ _ , . ._. _ _

,

*A1 *

;

;

J ,

,

d

4

i'

:

~

MARK I CONTAINMENT PROGRAMAUGMENTED CLASS 2/3 FATIGUE EVALUATION

METHOD AND RESULTS FOR TYPICAL TORUS ATTACHED'

,

AND SRV PIPING SYSTEMS

November 1982

MPR-751

,

Prepared by

MPR AssociatesWashington, D.C.

!

t

I

0212090155 821130PDR TOPRP EMVGENEC PDR

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

F. {.

t

ACKNOWLEDGMENT

This report was prepared by MPR Associates with assistance providedThethrough the joint efforts of the Mark I Architect / Engineers.specific organizations which contributed to the preparation of thisreport,and/or provided fatigue evaluation results are as follows:

Bechtel Power CorporationBechtel Power Corporation San Francisco, CaliforniaGaithersburg, Maryland

United Engineers and ConstructorsEDS Nuclear Philadelphia, PennsylvaniaSan Francisco, California

Southern Company ServicesNUTECH Birmingham, AlabamaSan Jose, California

TVATeledyne Knoxville, TennesseeWaltham, Massachusetts

-i-

.- _=

?,"

.

.

DISCLAIMER OF RESPONSIBILITY

Neither the General Electric Company nor any of the contributors.

to this document makes any warranty or representation (express or

of the information contained in this document or that the use ofimplied) with respect to the accuracy, completeness, or usefulnesssuch information may not infringe privately owned rights; nor dothey assume any responsibility for liability or damage of any kindwhich may result from the use of any of the information contained

-

in this document.

:|

.

t

|

|I||

:i

- 11 -

[- - _ - _ -

,c".,

.

'

4

TABLE OF CONTENTS

1.0 INTRODUCTION AND SUMMARY>

2.0 DISCUSSION OF EVALUATION METHOD'

3.0 RESULTS AND CONCLUSION

4.0 REFERENCES

APPENDICES

Augmented Class 2/3 Fatigue Evaluation Method, Tables andA.Nomenclature

Effect of Thermal Gradient Stresses on the FatigueB.Life of SRV Discharge Piping

.

,

|!

:

e

h

?

- 111 -

-- . - ..

V. .

.

.

.

1.0 INTRODUCTION AND SUtHARY

The fatigue approach being followed in the Mark I Long-Term Containment

Program for torus piping attached and SRV discharge piping is to followappl # cable Class 2 piping rules of Section III of the ASME Code(Rr(erence 1). The Code requirements address cyclic thermal and anchor

In August 1981 the NRC raised a concern regarding theinotion stresses.The Mark Icyclic stress due to the Mark I cyclic mechanical loads.

Owners Group and GE met with the NRC to discuss this concern in

September 1981 and proposed that a task group be assigned to evaluatewhether the Mark I Containment Program loadings and acceptance criteria

already contained sufficient margin for fatigue effects and to define aThecourse of action for a generic Mark I response to the NRC concern.

approach was to develop a method for piping fatigue evaluation and toapply the method on piping systems representative of the most limitingin Mark I plants. It was agreed that the fatigue approach should be

developed along the lines of the Class 2/3 piping design methods sincethese methods were already being applied for Mark I Containment Program

plant-unique torus piping analyses.

The Mark I fatigue task group was comprised of several of the Mark IA/E's and coordinated by the General Electric Company. The objectives

of the group were as follows:

Determine the loading cycles and loading cycle combinations appli-0

cable to the Mark I Containment Program load definitions.

1-1

i

1:- . .

.

,

Develop a methodology for performing an " augmented" Class 2/3O

fatigue evaluation on torus piping to account for cyclical mechani-

cal loads.

Provide the methodology to the Mark I Architect / Engineers for theirO

use in evaluating representative limiting piping systems.

Prepare a generic evaluation of the torus attached and SRV dis-O

charge piping with regard to mechanical fatigue.

The differing analytical approaches used for Mark I piping analysisrequired close coordin-tion between the A/E's in developing the fatiguemethods to ensure tha". he fatigue method could be applied with analysisresults available from each A/E's plant-unique piping analyses.

The final augmented Class 2/3 fatigue evaluation method described inthis report reflects the input and comments from all Mark I A/E's and

guidance provided by the General Electric Company.

Tables 1-1 and 1-2 indicate the scope of the fatigue evaluationsAs can be seen, essentially

performed with the final fatigue procedure.A total of 36 piping systems

all of the Mark I plants were addressed.SRV discharge lines comprised 30% of

were included in the evaluation.the systems.

The results of the evalt.ation of the piping for fatigue due to mechani-cal loadings in addition to thermal and anchor motion show that fatigueusage for Mark I torus piping is generally low, with fatigue usage typi-cally well below 0.3. None of the lines has a fatigue usage greater

than 0.5. It should be noted that the_

)

1-2

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

't . .

.

,

stress results for the most limiting piping systems and locations wereselected for each plant. Thus, these results are representative of themost limiting locations for fatigue usage and the remainder of thesetorus piping systems would have even lower fatigue usage.

.

1-3

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

?- .

.

.

TABLE l-1

SUMMARY OF PLANTS AND TORUS ATTACHED PIPING SYSTEMSINCLUDED IN MARK 1 CONTAINMENT PROGRAMAUGMENTED CLASS 2/3 FATIGUE EVALUATION

NUMBER OF TORUS

PLANT TYPE ATTACHED PIPINGSYSTEMS ANALYZED

BWR-4 2Hatch 1 and 2

BWR-4 2Peach Bottom 2 and 3

4BWR-4Cooper

3BWR-2Oyster Creek

BWR-3 3Pilgrim

BWR-3 1Millstone 1

BWR-4 1Vermont Yankee

BWR-4 2Brunswick 1 and 2

1BWR-3Quad Cities 1 and 2and Dresden 2 and 3

1BWR-4Duane Arnold

BWR-4 3Browns Ferry 1, 2and 3

2BWR-2Nine Mile Point 1

25TOTALS

|

t

k._ . _ .

Y. ..

.

4

TABLE l-2

SUMMARY OF PLANTS AND SRV DISCHARGE PIPING SYSTEMSINCLUDED IN MARK I CONTAINMENT PROGRAMAUGMENTED CLASS 2/3 FATIGUE EVALUATION

PLANT TYPE NUMBER OF SRVDISCHARGE PIPINGSYSTEMS ANALYZED

BWR-4 1Hatch 1 and 2

Peach Bottom 2 and 3 BWR-4 1

BWR-4 1Cooper

BWR-4 3Fitzpatrick

Millstone 1 BWR-3 1

Brunswick 1 and 2 BWR-4 1

BWR-4 1Fermi 2

Monticello BWR-3 1

Browns Ferry 1, 2 BWR-4 1

and 3

11TOTALS

1

e

k

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

F .., , . . _.--

.

.

2.0 DISCUSSION OF EVALUATION METH0_D_

This section contains a description of the evaluation proceduredeveloped for the Mark I torus piping fatigue evaluation: (1) the basis

for the method, (ii) the key assumptions made and (iii) the principalconservatisms in the method. Each of these subjects is covered in aseparate section below. The abbreviations and nomenclature used in thissection are defined in Section 2.0 of Appendix A.

2.1 Basis for Evaluation Method

In developing the basis for the Mark I piping fatigue evaluation twoapproaches were considered: (i) use of the ASME design fatigue rules forClass 1 piping; or (ii) use of the ASME design rules for Class 2/3'

'

piping augmented to include both mechanical and thermal cyclic stresses{ in the applicable evaluation equation. The latter approach was followed

since the Mark I analysis guidelines already specify that Class 2/3piping design should be used for the plant-unique evaluation and thusthe results of plant-unique evaluations which were already available

could be used. In this way the considerable effort of reanalysis of'

piping with the Class 1 rules could be avoided..

There are two ASME Class 2/3 piping design equations which account for'

repeated loadings: Equations 10 and 11 of Paragraph NC-3652.3(Reference 1). As has been shown in Reference 4, there is reasonablygood agreement between the Class 2/3 method and the Class 1 (previouslyjB31.7) method, especially below 20,000 cycles where all of the Mark I

| Reference 4 also describes the basis for the Class 2/3:

loadings fall.Mark 1

! design equations which were developed by Marki (Reference 5)..

2-1,

n- . _ . __

!- ..

.

.

The stressdeveloped his design equations from piping fatigue data.range reduction f actors, which are included in the Class 2/3 designequations, are basically a " stepped" approximation of Markl's equations.

The augmented Class 2/3 method uses the original Markl equations (onefor carbon steel and one for stainless steel). A comparison with thecurrent ASME Class 1 design fatigue curve and the Class 2/3 " stepped"

As can be seen, there isapproximation is shown in Figure 2-1.reasonable agreement between the Marki equation used in the augmentedClass 2/3 procedure and the Class 1 design fatigue curve and the Markl

equation is conservative when compared to the Class 1 curve below 10,000

cycles.

Tables 2-1 through 2-3 cover the details of the evaluation methodincluding the loadings considered, the loading durations and the load

These loadings are consistent with the loadings definedcombinations.in the Mark I Load Definition Report (Reference 3) and the load

As shown incombinations specified in the PVAAG (Reference 6).Tables 2-2 and 2-3, two evaluations were performed for each piping

(i) design break accident (DBA) plus normal operatingsystem:conditions (NOC); and (ii) intermediate or small break accident

Each evaluation includes both safe shutdown and(IBA/SBA) plus N0C.operating basis earthquakes. The IBA/SBA evaluation was performed so asto envelope the loading cycles of both the IBA and SBA events.

The stress results for each loading typically correspond to the maximumvalue which will not, in general, occur for each cycle of the loading.Many time history analyses of these loadings have been performed onMark I piping systems which show that most of the response cycles areless than the maximum value. To account for this variation, factorswere computed to convert the calculated fatigue cycles into effective

2-2

I

- , - ,_ -_ .

,.

full stress cycles. The method used to compute these factors involved.

analysis of the time history responses for each of the loadings on anumber of typical Mark I torus piping systems.

,

-

,

1

.k

I

i

Ii

!

2-3'

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

c ,.

.

2.2 Assumptions

This section lists the principal assumptions which were made indeveloping the augmented piping fatigue analysis methodology.

A typical loading history for a piping system attached to aO

Mark I containment includes:

Periodic SRV actuations for the life of the plant (NOC)-

with the total number of actuations determined for thespecific plant. One combined thermal and anchor motionload is assumed to act during each actuation.

Five operating basis earthquakes (Reference 2).-

One accident condition - either DBA or IBA/SBA which-

includes; (i) one combined thermal and anchor motion-

loading, (ii) operating basis earthquake (0BE) and safeshutdown earthquake (SSE) earthquake stresses, and (111)

periodic SRV actuations during IBA/SBA with the totalnumber of actuations determined for the specific plant.

Other thermal cycles due to normal operating conditions (for0

example, system testing) are considered negligible.

Stresses due to thermal gradients in the wall of SRV dischargeOSee

piping do not significantly contribute to fatigue usage.Appendix B for a discussion of the basis for this assumption.

2-4

;

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

,.- ,

,

f. ~. ..

.

.

Loads are grouped according to the combinations listed inO

Reference 6 considering the number of cycles for each

loading. Equation 11 from ASME Section III, Class 2,Paragraph NC-3652.3 is the basis for calculating fatiguestress ranges.

To evaluate the allowable number of fatigue cycles the MarklO

equations are used (Reference 4):

N = (245/iS)5 -- carbon steel

N = (281/iS)5 -- stainless steel

where:

N = number of cycles

iS = intensified stress range in ksi

The fatigue usage due to SRV discharge actuations taking place0

prior to the Mark I long-Term Containment Program is accountedfor by including an estimated number of discharge actuationscorresponding to the full 40-year plant lifetime.

Each earthquake load contains ten (10) significant responseO

cycles (Reference 2).

Frechug and IBA condensation oscillation (IBACO) loads areO

assumed to act at the highest load frequency of 9.5 Hertz,thus giving a reasonable estimate of the number of response

cycles per second for this loading.

F 2-5

.

Y< . .

.

.

Conservatisms in Evaluation Method2.3

This section summarizes the principal conservatisms in the fatigue

evaluation method.

Stresses from hydrodynamic loads were computed on the basis of0-

peak loads which bound full-scale test results.

Absolute summation was used to combine all dynamic responses.O

All effective full stress cycles were assumed to be in phaseO

when two events are combined; further, events were assumed tocombine peak on peak for the duration of the combined event.

Deadweight (DW) is not a fatigue load and would not normallyO

be included in the fatigue stress summation; however, it isincluded in all combinations since it is a required loadingfor ASME Class 2 piping, Equation 11 (Reference 1) and it is

conservative to do so.

Stresses for operating and safe shutdown basis earthquakes areO

combined with most limiting DBA/IBA/SBA stresses for the

appropriate number of full stress cycles.

Thermal stresses for accident events are combined with 'heO

appropriate mechanical stresses although thermal loadings aresingle cycle loads whose contribution to fatigue would notnormally be considered.

<

2-6

hi

J

:e ,,

- . .

..

.

Stresses for the safe shutdown earthquake (SSE) are consideredO

in the fatigue analysis. Typically fatigue analyses do notIt is noted thatrequire consideration of SSE stresses.

fatigue requirements in the PUAAG (Reference 6) specify thatonly operating basis earthquake need be considered in the

~ fatigue analysis of Mark I torus shells.

I

(

2-7,

;-

i i .._

-

.

TABLE 2-1

MARK I LOCA EVENT AND LOADING DURATIONS(Reference 3)

__

.

- LOCA EVENT DURATION (seconds)

LOCA TYPE DBAC0 PRECHG/IBACO CHUG

30 30 30DBA

IBA (with steam-driven feed 300 200pumps)

--

IBA (with motor-driven feed 1100 200pumps)

--

_

m, -

900900--

SBA

NOTE:

with IBA and SBA combined as one LOCA event, the following boundingSince the augmented Class 2/3 fatigue method was applied generically1.

LOCA loading times were used in determining the fatigue loadingcycles:

1100 secondsPRECHG (same as IBACO)

--

900 seconds--

CHUG

The most limiting IBA event duration (i.e., motor-driven feed pumps)was used to determine the bounding IBA/SBA LOCA event durations.

,

U

j

e- -

F.. .

.

TABLE 2-2

FATIGUE LOADING COMBINATIONS AND CYCLESFOR NOC + DBA CONDITIONS

"~CYCLES

COMBINATION _(Note 4)

1DBACO + EQ(0) + MCOBA + PRDBA

9DBAC0 + EQ(0)

890OBACO

10CHUG + EQ(S)

(Note 1)CHUG

102PRECHG

40SRVNOC + EQ(0) + MCNOC + PRNOC (Note 3)

(Note 2)SRVNOC + MCNOC + PRNOC (Note 3)

(Note 2)SRVNOC

'

_D_B A,BNOC_

MCDBA + PRDBASRVNOC

DBACO PRECHG CHUGMCNOC + PRNOC i

EQ(S)EQ(0)

NOTES _:

Number of cycles depends on dominant piping response frequency.1.

Number of cycles depends on plant evaluation of number of normal SRVactuations and reactuations over the life of the plant (See

I 2.

Appendix A).i

MCNOC and PRNOC are the thermal-expansion and pressure stress rangesand are applicable to SRV discharge piping only under normal3.

operating conditions (NOC).!

Pool swell excluded from fatigue analysis per guidelines containedin the PUAAG (Reference 6). For nomenclature, see Appendix A.4.

.

b

.0 . ~ , ~ '

'!*

,

r

. TABLE 2-3,

FATIGUE LOADING COM3INATIONS AND CYCLES >

_FOR NOC + IBA/SBA CONDITIONS

COMBINATION (Note 5) 'h CYCLES

4 1CHUG + SRVIBA + EQ(S) + MCIBA + PRIBA

9,

CHUG + SRVIBA + EQ(S) ,

*

CHUG + SRVIBA [ (Note 1)

(Note 2)CHUG

j

10,450 -' '

IBACO (same as PRECHG)

SRVNOC + EQ(0) + MCNOC'+ PRNOC (Note ok50 ,f 7 ,

-

SRVNOC + MCNOC + PRNOC-(Note 4)(Note 3)

'oiw,

(Nate 3)SRVNOC

IBA/SB/1_NOC

MCIBA + PRIBA

SRVIBASRVNOC

MCN0C + PRNOC IBAC0 CHUG' -

EQ(0) EQ(S)[

'

NOTES: 1

Number of cy(cles depends on plant evaluation of the number of'SRV4

1. ,actuations See Appendix A).

'

Number of cycles depends on frequency of piping (See Appendix A).2. IhNumber of cycles depends on plant evaluation of number,'of normal SRV 'i3. actuations and reactuations over the life of the plant- (See ,

Appendix A). i

MCNOC and PRNOC are the thermal expansion and pressure stress ranges4and are applicable to SRV discharge piping only under normal '

| operating conditions (N0C). ,

5. For nomenclature, see Appendix A.

.

.

't.

{<

i | \

J- s

--.. .

.

Je

.

NUMBER OF CYCLES5 106

3 104 102 1010 I10 '

.

_ . . _. .

. , i || . -....

..;.. ..- .g.

..'.

. ; ._ .

. g ,, ,. .-.__4...s . .. . {. -.

' .... . . . - , , .-i . | . .. , ,

103 :4.._,.. . g;j._3.__......i...|.-| ..-.-|.3 , .

. . - , . . -.. . .-.. . ....

.- .

F..- . . . _ ., .| CLASS:,._..g....... ;. , ,,

,

CED J j

FOR ;A106. GrB . MATERIAL..j REDU,

'' ' ~ .- ~ ~ ~ ~ ~ ' -!'- ~ ~ -' ~ ' ' - - ~~ '~~

1. DESION FATIGUE CURVE.

, .. .' I _ .;._.._..i . . _ _ _._ ._. - . . .L, . y - . _ .

-

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

~ ~T-___

M_ _ -

i BY K :i*i i! .: ' t '' )' ~'.i .) ! g . -- ''':N t - -i- t '. !' '

{ . CLASS 1 DESIGN FATIGU CU VE __ ; . ..s ..- - It' ._ _

%

VALUES ._ N I" '" '

ji ! I 'I "I ' 1 I ~ "''"''I'" '..... " " " '~~ --~ ~~ ~ ~ '

'I i"- - '- -! |"" ' ~

CLASS 2/| 3| <}||- 1

-'

,rOF''

S ' ' , .~~

' '5~~ + 1

..

EQUATION (ll)j LIM::TS' '

!-[" -~ ~ ~ ,

|,,

% 'T : ;|

| .|- | 6a , ! M;. ;'

.

~'l'!

- :- +j' ; '. - .(KSI) ! % .,

f 6.,

' I ,! ,N, g . s - .. ..

. .} , . .,i*I- -| .. j . , . , , , . .

. .f . .i,

|gi -. e ! .'2 e- ...

10 - .,.-7... ._

j- ..

4 .,.s.- . - . . . .

.i,, . p, j ,. --g

~@ . .: .. 7

'-|

-.--- 4-- -- -- r ,, .,,.,

-_.g ..4-~.~& .... . -- -

*% ~ .__ . p . ., . 7.,

. . . . %&- -.. m. ..~..i"*+'..W'' . -. .

-- . ., y. . -- ..

t :. i4 *.

L1 .: . . .

$. . i..- . . . ,,1 . .. } . , .. ,.

& | ,) ..

.., ) | . . J .. .i . r l. . - .4 a e te

'*

1I II1II- *l''' .I~ h ._~ .f,;

.l. .| | 'T '' , .

_ . i .. ; .. L' l '

_j. , . , . __.. .._ _ _ .% ''IiN q. ;.%'

- _ . _ _ ,. _..

]. . . , .i . |. .4 .;- %'

M --,

g .; '' ..

*.

. ' --r.,, .

.. _ ...

, ' , , .;- .___. ___ . - ,

'.

.. .. . y ,.. g-%

..._ ..; . ' . _.~ I'j.. . . . .

.. - ;;;j ; '; . . . .. .. ...;;;

. ..s

g....|| |- g. g .. g .) }"i i *;* g -

i.

,

* ;t -[ ,

| |, 4'gl ;p;: ;,,1, .I. ,

..*

II

..[.3([ ;,.

10- ,

,

o o o oo o o o. e .

o o o oo .

o N m o _

ar o -r.-4 m w

re- Augmented Class 2/3 Fatigue Curve

.

Marki FATIGUE CURVE(Reference 4)

COMPARISON OF AUGMENTED CLASS 2/3 FATIGUE.

LIMITS WITH CLASS 1 DESIGN FATIGUE CURVE

FIGURE 2-1

. .

-- ~ _ . '7'~~

r., 1

I.

3.0 RESULTS AND CONCLUSIONS

This section contains the results of the fatigue evaluationsperformed on over 30 torus piping systems. These systems wereselected by each A/E as representative of the most highly stressedtorus piping systems in their respective plants. Thirty percent ofthese were SRV discharge lines and the remainder were linesattached to the torus with sizes ranging from 2-inch to 24-inch.

TheAll torus piping systems had a fatigue usage less than 0.5.f atigue evaluation results, which are tabulated in Table 3-1, are

i summarized as follows:4

; SRV Discharge Piping:O

.

Percent less than 0.3 fatigue usage -- 72.7%Percent less than 0.5 fatigue usage -- 100%

-

0 Other Torus Attached Piping:

.

Percent less than 0.3 fatigue usage -- 92.0%Percent less than 0.5 fatigue usage -- 100.0%

i

A very conservative methodology has been developed for fatigueanalysis of Mark 1 Class 2 piping. The fact that the calculated

!fatigue usage factors are low coupled with the very conservative

|approach used to develop the fatigue analysis methodology shows

Thus thisthat fatigue is not a concern for attached piping.] report answers the concern expressed by the NRC regarding the4

effect of cyclic mechanical loads on fatigue. Accordingly, there!

is no need for a complete evaluation of torus piping fatigue on a,

plant-unique basis.

3-1I

f

4

[.~

.

.

.

TABLE 3-1

SUMMARY OF RESULTS OF AUGMENTED CLASS 2/3 PIPINGFATIGUE EVALUATIONS FOR REPRESENTATIVE

MARK 1 PIPING SYSTEMS _

.

SYSTEM FATIGUE USAGE__

ARCHITECT ENGINEER(Size) N0C+DBA NOC+1BA/SBA

Utility

Plant__

A. Bechtel-Raithersburg1. HPCI Pump Suction 0.000 0.002

Georgia Power(16-inch)Hatch-2

2. Core Spray Pump 0.001 0.002

(10-inch)

3. SRV Discharge 0.077 0.096

(10-inch)

B. Bechtel-San Francisco

Philadelphia Electric 1. Core Spray System - C 001 0.022

Peach Bottom-3 P-14-5(16 and 14-inch)

2. Inerting System 0.013 0.004

P-9-2(20-inch)

3. SRV P-1-5 0.046 0.202

(12-inch)

,

i

}- - - - -

]]'. *

*

_

.

TABLE 3-1 (Cont'd)

SUMMARY OF RESULTS OF AUGMENTED CLASS 2/3 PIPINGFATIGUE EVALUATIONS _0R REPRESENTATIVEF_

MARK I PIPING SYSTEMS.

FATIGUE USAGE _SYSTEMARCHITECT ENGINEER _ (Size) NOC+DBA NOC+IBA/SBA

Utility

Plant_

C. EDS-Nuclear _1. Core Spray Suction - 0.113 0.149

Nebraska Public PowerX227A (16-inch)Cooper

2. Core Spray Suction- 0.009 0.012

X227B (16-inch)

3. RCIC Turbine 0.279 0.280

Exhaust (8-inch)

4. HPCI Turbine 0.020 0.058

Exhaust (24-inch

5. a. SRV Discharge 0.116 0.246

(10-inch)

b. SRV Discharge .005 .006

(10-inch)

i

4

' . ,ry ,

.

.

TABLE 3-1 (Cont'd)

SUMMARY OF RESULTS OF AUGENTED CLASS 2/3 PIPING _FATIGUE EVALUATIONS FOR REPRESENTATIVE

MARK 1 PIPING SYSTEMS

FATIGUE USAGE _SYSTEMARCHITECT ENGINEER (Size) NOC+DBA NOC+IBA/SBA

Utility

Plant

D. MPR Associates1. Vacuum Relief-Type 2 0.434 0.258

GPU Nuclear(24 and 18-inch)Oyster Creek

2. Demineralizer 0.087 0.067

Relief (20-inch)

3. Core Spray Suction0.084(12, 16 and 20-inch) 0.131

.

:. . !.

.

.

TABLE 3-1 (Cont'd)

SUMMARY OF RESULTS OF AUGENTED CLASS 2/3 PIPINGFATIGUE EVALUATIONS FOR REPRESENTATIVE

MARK 1 PIPING SYSTEMS _

FATIGUE USAGESYSTEM

NOC+IBA/SBAARCHITECT ENGINEER (Size) NOC+DBAUtility

Plant

E. Teledyne1. SRV Discharge-C 0.189 0.117

PASNY (10-inch)Fitzpatrick

2. SRV Discharge-B 0.022 0.027

(10-inch)

3. SRV Discharge-A 0.252 0.303

(10-inch)

1. HPCI Turbine 0.000 0.000Boston Edison Exhaust (24-inch)Pilgrim

i

0.000 0.0002. RCIC Pump (6-inch)Suction

3. Core Spray Suction 0.000 0.000

(18-inch)

Northeast Utilities 1. Heat Exchanger 0.053 0.003

x 210A (8-inch)Millstone 1

2. SRV MS-8F (10-inch) 0.0270.034

i

l

i

4.

!:..,

.

.

.

TABLE 3-1 (Cont'd)

SS 2/3 PIPING _

_ SUMMARY OF RESULTS OF AUGMENTED CLAFATIGUE EVALUATIONS FOR REPRESENTATIVEMARK I PIPING SYSTEMS

FATIGUE USAGESYSTEM

ARCHITECT ENGINEER _ (Size) NOC+DBA NOC+IBA/SBAUtility

Plant. _ _

1. Core Spr y - Part 6 0.002 0.002Yankee AtomicVermont Yankee (10-inch

1. Containment Spray .012 .000Niagara Mohawk X8-326(3-inch)Nine Mile Point

.001 .0362. Pump IIIX5-337(12-inch)

F. United Engineers0.086 0.340

1. RCIC TurbineCarolina Power Exhaust (8-inch)Brunswick

0.001 0.0032. RCIC Barom.

Condenser (2-inch)

0.486 0.4753. SRV Discharge(12-inch)

!!

O

.'

. ' . ' . '~

.

.

TABLE 3-1 (Cont'd)

SUMMARY OF RESULTS OF AUGMENTED CLASS 2/3 PIPING __

FATIGUE EVALUATIONS FOR REPRESENTATIVEMARK I PIPING SYSTEMS

.

ARCHITECT ENGINEERSYSTEM FATIGUE USAGE

(Size) NOC+DBA NOC+IBA/SBAUtility

Plant

,

G. NUTECH

Commonwealth Edison HPCI Turbine Exhaust 0.047 0.056'

Quad Cities Unit 1 (24-inch)

Iowa Electric Core Spray Suction 0.059 0.039

Duane Arnold (12-inch)

SRV Discharge Piping 0.099 0.056,

Detroit EdisonFermi II (12-Inch)

SRV Discharge Piping 0.307 0.197Northern StatesMonticello (10-inch)

H. TVA

RCIC Turbine Exhaust .010 .021Browns Ferry

(2-inch)

RCIC Turbine Exhaust 0.003 0.096

(12-inch)

ECCS Suction Line .007 .023

(16-inch)

(a) SRV - Line H Elbow 0.003.232

(10-inch)

(b) SRV - Line H .047 .318

Ramshead (10-inch)

U

). ___.

'i-

-_ ,

.

.

4.0 REFERENCES

ASME Boiler and Pressure Vessel Code, Section III,19771.-

Edition with Addenda through Summer 1977.

NUREG-75/087, Standard Review Plan for the Review of Safety2.Analysis Reports for Nuclear Power Plants _, May 1980.

l

NEDO-21888, Revision 2, Containment Program Load Definition.

3.'

Report, General Electric, November 1981.

!

ORNL-TM-3520, " Comparisons of Test Data with Code Methods for|

4.Fatigue Evaluation"; E. C. Rodabaugh and S. E. Moore, Oak

|Ridge National Laboratory, November 1971.

i

|Markl, A.R.C., " Fatigue Tests of Piping Components,"S.

| Transactions ASME, Volume 74, pp. 287-303, 1952.i

6. NED0-24583-1, Mark I containment Program Structural Acceptance'

Criteria Plant-Unique Application Analysis Guide, General

Electric, October 1979.

,

1

e

e

i

i

i

4-1

<

i .,

.

.

APPENDIX A

AUGMENTED CLASS 2/3 FATIGUE EVALUATIONMETHOD, TABLES AND NOMENCLATURE

i

,

i

c. -

.

.

STEPS'FOR' AUGMENTED ASME CLASS 2/3 PIPING FATIGUE ANALYSIS~

1.0

In performing the augmented ASME Class 2/3 piping fatigue

analysis first enter information on each piping system in the

blanks shown in Table A-1. Then proceed with Steps A-G as

fol' lows :

Calculate stress resultants for each DBA, IBA/SBASTEP A:and NOC occasional, thermal and anchor motion

loading condition. See Tables A-2 and A-3 for

individual loads and Section 2.0 for nomenclature.

STEP B: For SRV piping, determine the discharge pressure

(PSRV) and the stresses due to thermal expansionand anchor motions (MCNOC) and discharge conditions

(SRVIBA).

Determine number of SRV actuations that would beSTEP C:

expected to occur:

(1) During normal and transient operation over thelife of the plant - nSRVNOC

(2) During an IBA or SBA - nSRVIBA'

Determine maximum characteristic frequency of theSTEP D:

piping system for dynamic loadings (fmax)*

. _.

9. -

.

.

STEP E: Determine the location in the piping system withthe most limiting intensified stress conditions.When piping systems have both stainless and carbonsteel runs, limiting stresses in both runs shouldbe considered in determining the most limiting

location for fatigue.

STEP F: Perform the NOC + DBA fatigue evaluation by

completing the information in Table A-2.

STEP G: Perform the NOC + IBA/SBA fatigue evaluation by

completing the information in Table A-3.

!1

|

|

A.2

)

f. -'

-.

.

.

2.0 NOMENCLATURE _

Flexural stresses due to chugging loading forCHUG

DBA/IBA/SBA as defined in Section 4.5 of the=

LDR (Reference 3). Include stresses due tounderwater drag and fluid structure inter-action for submerged piping segments (ksi).

~ Nominal outside diameter of piping in inches=D at location where fatigue evaluation is per-

formed.

Stress in ksi due to deadweight of piping=DW system including fluid where appropriate.

Corresponds to M /Z in Equation 11 ofASection NC-3652.3 of Section III, ASME Code(Reference 1).

Flexural stresses due to condensation loadingDBACO

as defined in Section 4.4 of the LDR (Refer-=

Include stresses due to underwaterence 3).drag and fluid structure interaction forsubmerged piping segments (ksi).

i

Flexural stresses due to operating basisEQ(0)

=

earthquake (ksi).i

Flexural stresses due to safe shutdown earth-'

EQ(S), =

quake (ksi).

Highest characteristic or participating Usefrequency of the piping system in Hertz.f

='

max

= 30 Hz unless a lower value can be'fj$$(ified from analysis of the piping system.-

*

Stress intensification factor for ASME Class 2piping analysis (Section NC-3673.2(b)).i =

Flexural stresses due to IBA/SBA condensationIBACO oscillation loading or DBA prechug loading as=

f (PRECHG) defined in Section 4.4 and 4.5 of the LDRI

(Reference 3) (ksi).i

!.,

k

A.3

b _. __

i. .

,

.

e

.

Flexural thermal stresses due to thermalMCDBA,expansion of piping during the most limiting

=

condition of DBA or IBA (whichever isMCIBA applicable) plus the corresponding flexuralstresses due to anchor motion (ksi).

The

MCIBA stresses should also include theflexural stresses due to anchor point motion

-

resulting from SRV actuation during IBA orSBA.

Flexural thermal stresses due to thermalMCNOCexpansion of SRV discharge piping and anchor

=

motions during SRV actuation (ksi).

Number of significant response cycles forA value=

nCYC loadings included by SRV actuation.of 15 is used which is arbitrary since it isused in the calculation of RSRV; however,15is a reasonable estimate of the number ofsignificant response cycles for SRV thrustloadings.

Effective full stress cycles for load combina-=nk tion k.

Allowable cycles for total combination stressTE calculated as follows (Reference 5):

N =g

S

f r carbon steel, STE(ksi) TE)(245/SN =

K

(281/Sfor stainless steel, STE

(ksi) TE)5N =K

Number of SRV actuations and reactuations thatnSRVIBA would occur during an IBA or SBA accident

=

(whichever is greater).

Number of SRV actuations that would occurSRVNOC under normal operating conditions over the

=n

remaining life of the plant.

Pressure range inside the piping due to themost limiting pressure condition of DBA, IBAP =

Aor SBA, whichever is applicable (ksi).

Pressure range in SRV discharge piping due toSRV SRV actuation. This load is only applicableP =

to SRV piping evaluation (ksi).

A.4'

}

7 -1 .

.

.

Stress range in ksi due to maximum internalpressure occurring during a DBA event (ksi).PRDBA =

PRDBA = PA x D/4 tn

Stress range in ksi due to maximum internalPRIBA pressure occurring during an IBA or SBA event

=

- (ksi ) .

PRIBA = PA x D/4 tn

Stress range in ksi due to internal pressurePRNOC resulting from SRV actuation (ksi).

=

PRNOC = PSRV x D/4 tn

Equivalent maximum stress cycle factor for theRCHUG -

=CHUG loading. For recommended value, seeTable A-4.

Equivalent maximum stress cycle factor forRDBAC0 DBAC0 loading. For recommended value, see

=

Table A-4.

Equivalent maximum stress cycle factor for theThe recommendedRIBAC0 IBAC0 (or PRECHUG) loading.

=

value is 1.0 since the load is essentially asingle frequency harmonic loading.

Equivalent maximum stress cycle factor for theSRV loadings including thrust bubble drag andRSRV

=

For recommended values,torus SRV excitation.see Table A-4.

Intensified stress ranges ir ksi calculatedS, =

using the resultant moment and sectioni

modulus as defined in Section III, ASME CodeS , etc.

2 (Reference 1).

Flexural stresses due to actuation and reactu-SRVIBA ation of SRV system during an IBA or SBA=

Includes stresses due to thrust andaccident.SRV bubble drag where appropriate (ksi).

Flexural stresses due to actuation of SRVSRVNOC system during normal operating conditions.=

Include stresses due to thrust and SRV bubbledrag where appropriate (ksi) unless includedin SRVTQF below.

A.5

,

f. .

.

;

.

Flexural stresses due to bubble drag resultingfrom actuation of SRVs if not included withSRVTQF

=

SRVNOC stresses.

Total intensified stress range in ksi forcombination based on Equation 11 of NC-3652.3STE

=

of Section III, ASME Code (Reference 1).~ Calculated as follows:

STE " S1+S2 + ...... + (0.75 i x DW)Nominal thickness of piping in inches at

t=

location where fatigue evaluation is per-n

formed.

i

A.6

~

l:- .

.

.

TABLE A-1

MARK I PROGRAM ATTACHED PIPINGAUGMENTED ASME CLASS 2/3 FATIGUE EVALUATION

I. G'ENERAL INFORMATION

A. PLANT (UNIT):

B. UTILITY:'

C. ARCH / ENGR:

D. PREPARED BY: DATE:

II. PIPING SYSTEM INFORMATION

A. SYSTEM IDENTIFICATION:

B. NOMINAL PIPE SIZE:

C. MATERIAL:

III. PIPING ANALYSIS INFORMATION

A. ANALYSIS METHOD:

B. INTENSIFICATIONFACTOR(1):__

C. LOCATION OF MAXIMUM STRESS:(Describebrieflyor attach figure)

D. DEADWEIGHT STRESS (DW):

E. NOMINAL THICKNESS (t ):n

F. OUTSIDE DIAMETER (D):

. - - . _-

! ,,j

.

E3,;u

L AjS ,C UnA =g

S'L E

GE A% SE U8

LS1

T I L _A A _F T

OC Tl

) _w ND O

Ix T1g5 AT7 U

+ 2L-) * = - As = = = - r = A Vt * - r = Et r ; Yl

L E; ;

N L EO B UN AGI

1 TTAFPFI ,

Y A|_ L_ N B

OA _ - V DB RGD _ sN, Rs _ - IP RPp OI D

P ES .DAtU 6|

* J q3 LS6S f CI 0

- _CE ( XS -S A Q EYG

_OS B E LEN SARE D

_RE C x . Nu

WANS S 1 -_PMT 3

h !E Nx 2A U I|B 1 LGD

|LI DET E

3 * WA1/

)) C C SFrS O O2 )

OtQ E

i iN N0

(Q C C eLMCS

(Q M M T OO ES O OR PA E E

x x N PFsL 2 xC 5 x x

1 11

/E 1 a 2 I|M 2 2

S ||

A I.|

/+ G|* C C F

l| c C O O Q U

/ H |,O O O d G r O N N T

CC C U U C N V v V S,/C A A H H E V R R RA B B C C R R aS S SB D D P S c/1 x x x x x xD

x ix x 1 ts x 1 2 /1 1 1 2 a2 !|1 1 2 2 2 g G1 /2 |

-2 2 B H

D / CC EM / R|J

P

/0E , 1V 2 8 /|aS 5 0 6 "I Ei 3 0 n "

MCnifi|F Y

|I /1n 1 , a 1LO

/ CA

|

CF /E = . . BD

. . . /' ,o,0

= .4- . !|

) C ooC C O c)C0 gY O N - ^Y .C N . V1 C ( V a p - ~- (

n.R $ |-S ^ -

_I " -Z H iH , -x x, x C - _a ,, OS m f

n. - _E I

L - _C x sxY x

x __=n CC- |f| O | i v

L O c C x p -A C u 0 A s v|

J_ ,A s B 8 aU 1 yB e 3 ST D B 8 R - _C " SA _ |1| xr ! * - _x x e n ,_x p o - _) , C) i

C C)C E E s t Y - _

e a c C *E S 1 lt Cu y " - _ ^S(

n.S(7

c *y c a - _ .

*0 0 C A3 0 ( - _

11, - - 5

, ,'

g5 $ r I "

33~ 1

. 3L5.

>

-. .

.

-~

_

_lgi

I |

_ _

_ .__

3, _ _ ,.

-gg, N . -,

_g g. _

/ |_.

. g/._'t

,G ge -_ _A t|

_

S' -

I

_, ,,

L'

,U -

,U ,G E! L

,_T I

-A ALF WC

O ._ _L

, NL

A _ O_ I

Tj

A) |deD Um L

AE 4T 5

7 - 3V _+ . = . - E3

.

. . - r r A= ES . _L .

, 7 r EUY( J| = L LG, N IO B TA V

B RG A A5 I SN TF3 R I

P RPOI AFP

B1.| , Y S_ L1|N /A - O

V AB

[1_-I RG

h SN BC-t

IP I4 - RPS m -1

iOIFP

1. | , 4_ !.S ;]

E )- _S 1| ) Ot

E S Q - _M_S )

S(Q E - _R (gyP,T

S E E x - _ ,3 ua 1 - _A 3 r ;.g,B 2

1S 1

/ 2 2 .AIBS 1. |

j

. CI 5 C Ob 1j

A A A O N3/ B B B N C

I ! C M2 1

V V V MS R R R xS y S S S m

1A S s m x

W1L 1 I| -1 |C 1 22 . I|2 kE F| . j CM . l| | ~C Q Q.S O C - O $ T

A G C O VN /V V R GG U A NG GR S A UU H B V

U U H C I S B HR TH H C S m CC C x E r

m x 1 v$ x a 1 2 R

j3 m u 1 _

1 1 i 2 >2 S

1 1 2 2 2 | ,Si

1

| _ |

2 ;|2 2 Q

E|

i|

Y,

\7 a "5 " "

\S 3 4* " Og I "E , n _ CL "g

C_ I |_

\A,

BI -

\Y_ 7 '

.IC_ g .

_ .

N_ 0-

'

, 0 5 x ). 5

Ij_0 _ i

C

O(Q1

_. C_ O _ _0 + O ) N

N " V _ _ E1 ) V R- n R S - _ |

[[( ) S | fl Z " . _ _- I|N {-

i

( I|g 1 C _ __ -

_ y H m 5 GN _ _,

Sa.g 9 V

R _ _E nL f I]a S . - _ cI.[CY |(|

| n

g eC lj O| V

ax x vC sL y A 8 a|| j

A S* B Vg G g R _ _U g u I.|T R 8 S

C u ri 8 _ __

A c8 x e n l]I.| po _ _r) |l | i x

) j C _ _en a C st Ypo _ c C _ _ *) E e s.C C il u *i _ y "st

ea S 0 yc ._ "c - - * .5 E ( ct _

0 C A . *1 l u (ct ( ) _ _

1 1yc - - '2 1CA 3;

_ 0 'c ,- | ", _ 1(

~ , g

$, _

.

j , , _

-_

j/$...

_

.

e,G.

_

_.

F..

,

.

.

TABLE A-4

CYCLE REDUCTION FACTORS FOR MARK I DYNAMIC LOADINGSASME Code Section Ill, Subsection NC-3611.2)(Reference:

Type of Load Significant Load Cycle Reduction

(Abbreviation)Cycles (n) Factor (R)

1. Acceleration due to SRV 15 0.3

discharge (SRVNOC)

2. Thrust due to SRV discharge 15 0.3

(SRVNOC)

i 3. Bubble drag due to SRV 15 0.3

discharge (SRVNOC)i

(whT$Never.13/f or 0

4. Acceleration due to DBA 900(Note 1)

condensation oscillation is greater,;(DBACO) Note 2)-

5. Acceleration due to IBA 102 DBA 1.0

condensation oscillation10,450 IBA/SBA

(IBAC0 or PRECHG)

(whT$Never.12/f or 0

; 6. Acceleration due to 321(Note 1) DBA,

chugging (CHUG) 9600(Note 1) IBA/SBA is greater,;

Note 2)

.

I

NOTES:

I Based on a 30 Hertz maximum participating frequency (actual fmax' 1.

could be used),

;

f,,x = highest participating frequency of piping system.2.

Ie

i-

!

}

.

6 ,

-01

---

_

_

.

.I.

-_

--

5 .

-0 '

?'1

-

: -N

:::- O- I

T: Aa, a - U

L.S- A

VE4 ,

-01 ES

- UEL E GC - V

1

Y I

- R T -CU A AF )n

- C FO(R ER - EO3E R

- UF/ UBMU G 2 G

-N I

T S I

F3 A S_ ot F A

1 >o. L- C-

- D;; E- ; T

- j N

y,SE- M

- ~_

a G- c U

A2 ._

0 - .1------

5

---

0 s1 0

3 1* .

01

S

' *:n.

*

.

.

.

| | f!| [| | (i im

f .., .

.

.

APPENDIX B

!

.

,

k'

RESPONSES TO NRC QUESTIONSAND COMMENTS

STAFF PRESENTATION OF SEPTEMBER 10, 1982;

e

!4

9

|'!i::

:

I

i'1

i

,

t

!n

,

t

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

b'. -

.

e

.

This Appendix provides additional information and justificiaionfor several of the assumptions employed in developing the

This materialAugmented Class 2/3 fatigue evaluation methodology.is provided in response to comments and qtestions resulting from a

10, 1982. The NRCpresentation to the NRC staff on Septemberstaff comment or question is listed first followed by the corre-

sponding response.

1. Comment:

Provide documentation of the fatigue methodology for the NRC

staff.

Response:

Section 2.0 of the body of this report and Appendix A contain

the requested documentation.

2. Comment:

Identify which piping systems were evaluated for each plantconsidered and the fatigue usage results for each.

Response:

(

Table 3-1 in the body of this report contains the requestedi

|

finformation in tabular form.

!

3. Comment:

Document how prior fatigue usage has been considered in the

fatigue evaluation methodology.

||

|

|a

$ >

,

t. .

.

.

Response:

The number of SRV actuations used in the analyses was basedon the expected number of SRV actuations over a 40-year plant

Thus, the results reported in Table 3-1 of thelifetime.body of this report account for prior fatigue usage.

In many cases the support arrangements for the SRV discharge

piping have been upgraded to withstand all of the Mark IContainment Program loadings. The stress distributions oforiginal piping arrangements are not generally available andin any event would not be comparable to the stress

Thus, itdistributions in the upgraded piping arrangements.was concluded that a reasonable approach would be to

,

extrapolate the fatigue evaluation results for the upgradedpiping configuration for the full 40-year life of the plants.

,

4. Comment:

Provide a rationale to justify not including thermal gradientstresses in the fatigue evaluation methodology.

Response:

f ASME Class 2/3 piping design equations do not include'

stresses due to thermal gradient stresses since they areTo

generally not significant to the design of these systems.justify the assumption for the fatigue evaluation method-ology, calculations were performed of the fatigue usageresulting from the thermal gradient stresses. Typical valuesfor key parameters were taken as follows:

,

!:

B.2

i

{;.' ..

- .

.

Temperature of steam inside the SRV discharge piping in.

O

0the wetwell - 350 F.

Initial temperature of torus water adjacent to SRVO0

discharge piping in the wetwell - 70 F.

For these conditions and typical sizes of piping used inMark I plant SRV discharge piping, the peak thermal gradientstress ranges were calculated as follows for the two types ofmaterials used for this piping:

3,900 psi (compressive)O Carbon Steel -

O Stainless Steel - 15,500 psi (compressive)

These stresses are compressive and occur on the inside of theStresses which occur on the outside of the pipe

(pipe wall.

Theare tensile and are somewhat lower in magnitude.

{ calculated stresses for the stainless steel piping are largerdue to the lower thermal conductivity of stainless steel:

compared to carbon steel. The effect of the maximum stresseson fatigue usage was estimated based on a bounding number of

')SRV actuations of 3,100(500-100014 moretypical) occurringover a plant lifetime. For the two materials, the fatigue

| usage was calculated using the methods of Appendix A with the'

i result as follows:

I less than .001O Carbon Steel -*

0 Stainless Steel - less than .002

,i|! Since all fatigue usages due to stresses other than thermal

'

gradient stresses were calculated for all Mark I plants to beless than 0.5, the effect of the usage due to thermal

,

:gradient stresses on fatigue lifetime would be insignificant.

i B.3

1

|+

1

-