ML20053B461

Fracture Mechanics Analysis of Oyster Creek Nuclear Generating Station Emergency Cooling Sys

ML20053B461
Person / Time
Site:	Oyster Creek
Issue date:	04/20/1982
From:	Arzoumanidis S, Cotter K FRACTURE PROOF DESIGN CORP.
To:
Shared Package
ML20053B455	List: ML20053B454 ML20053B461
References
TASK-03-05.B, TASK-3-5.B, TASK-RR NUDOCS 8205280425
Download: ML20053B461 (68)

Text

-.

fIRAL REEDE rL b

FRACTURE MECHANICS ANALYSIS OF THE OYSTER CREEK NUCLEAR GENERATING STATION I

EMERGENCY COOLANT SYSTEM I

I Submitted to GPU NUCLEAR CORPORATION Parsippany, NJ I

Prepared by Keyren H. Cotter and Seraf Im G. Arzoumanldis FRACTURE PROOF DESIGN CORPORATION 27 Maryland Plaza St. Louis, MD 63108 I

I A,,,, 20, iee2 820328o se

INDEX I

2 0.1 INDEX.........................

I 3

0.2 EXECUTIVE

SUMMARY

4

1.0 INTRODUCTION

4 1.1 Backgrov.4d.......................

1.2 USNRC Alternative Saf ety Assessment Criteria...... 5 2.0 CRACK STABILITY CRITERIA................ 77 2.1 LEFM Considerations..................

8 2.2 EPFM Considerations..................

I 3.0 J ESTIMATION AND LEAK RATE ANALYSIS 11 11 3.1 J-Integral Estimation 12 3.1.2 Longitudinal Cracks I

12 3.2 Leak Rate Ana l ys i s...................

4.0 ASSUMPTIONS & FORMULATION, FULLY PLASTIC ANALYSIS 15 15 4.1 Structural Response and T" I

4 ' ' c-a ' ' "ce e".

' e 19 4.1.2 Plastic Hinge Behavior....

4.2 Cracked Secti on Parameters...............

19 4.2.1 Plastic Limit Moment.................

19 I

20 4.2.2 J-Integral......................

20 4.3 Stab i l ity Ana l ys i s...................

21 3

4.4 The J-T Diagram 4.5 Extrapol ati on of J

-T,9 Curve 23

g 5.0 SYSTEM DESCRIPTION.

9 25 25 5.1 Piping System 25 I

5.1.1 Return Line.....................

25 5.1.2 Supply Line.....................

28 5.1.3 Surrounding Structure 28 l

5.2 Piping Code Stress Analysis 6.0 RESULTS & D I S CUS S ION.................. 32 32 6.1 L eak Det ectab l i i ty...................

32 6.1.1 Circumferential Flaws t g 32 5

6.i.2 tongitudinai riaws..................

33 6.2 Level D Loads.....................

33 6.2.1 Long i tu d i na l F l aw s..................

33 6.2.2 Circumferential Flaws................

36 6.3 Fully Plastic Analysis.................

36 6.3.t. Applied Load....................

6.3.2 Crack Stabil ity Cal cul ations............. 36 36 1

6.3.2.1 Return Line A...................

36 6.3.2.2 Return Line B...................

36 6.3.2.3 Steam Line A....................

I 37 6.3.2.4 Steam Line B....................

37 6.4 Discussion.......................

50 7. 0 REF EREN CE S.......................

A-1 APPENDIX A, JTPIPE.....................

APPENDIX B, MATERI AL PROPERTY DATA.............

B-1 APPENDIX C, USNRC CRITERIA................. C-1 I

2

EXECUTikE

SUMMARY

In 1979 Jersey Central Power and Light Company (JCP&L) performed an evaluation of mergency condenser HELBS outside containment for the Oyster Creek Nuclear Generating Station and concluded that a pipe break on the 75' elevation of the reactor building could result in damage to the mergency condenser Isolation valves and controls. This normally mandates the Installation of a system to protect against the ef fect of a pipe break.

However, the USNRC recently developed Criteria which permits plant operators to use alternative methods to obviate the need to consider pipe rupture events and the associated pipe-whip protection requirments. Based on this USNRC Criteria, GPUN performed an analysis of the 75' elevation piping and concluded that the Criteria was completely satisfled.

it is thus recommended that no pipe-whip protection be Installed at the 75' olevation provided periodic (i.e.,every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />) visual inspections of the floor are made.

I I

I 3

1.0 INTRODUCTION

I 1.1 Background.

I In 1979 Jersey Central Power and Light Company (JCP&L) perfortred an evaluation of mergency condenser HELBS outside containment for the Oyster Creek Nuclear Generating Station and concluded that a pipe break on the 75' elevation of the reactor building could result in damage to the emergency condenser Isolation valves and controls.

JCP&L reported this finding to the NRC regional office in October 1979.

Subsequent inspection conducted on site by the NRC staff (SEP Branch)

Identified the same potential damage.

The results of the Inspection were transmitted to JCP&L by the NRC in a letter dated July 10, 1980.

The letter requested modifications to provide adequate protection from the effects of postulated HELBS in the mergency condenser systm.

Engineering design studies were undertaken by JCP&L to determine practical and of fective methods of providing protection against the adverse consequences of a postulated break in the emergency condenser piping on the 75 foot level of the reactor building outside primary containment.

These studies included the evaluation of the feasibility and offectiveness of:

(1) adding pipe-whip restraints ar d Jet Impingment barriers to protect the isolation valves outside containment; (2) installing a third Isolation valve in the steam line of the emergency condenser piping just inside containment; and, (3) closing one of the t

l normally open steam line isolation valves catside containment and providing a signal to open upon demand. The results of these evaluations showed that none of l

these modifications could reasonably be accomplished on a retrofit basis in a manner which would ef fectively resolve all of the potential areas of concern and 1

i also not impose other significant limitations on access for inspection and I

maintenance.

The main results of the evaluations which lead to this conclusion l

4

trere summarized in JCP&L letter to the Commission dated September 30, 1981.

The letter also stated that an analysis to demonstrate that the emergency condenser piping will leak before a significant break could occur was to be performed and the results were to be transmitted to the NRC.

The USNRC recently developed criteria which permits plant operators to use alternative methods to obviate the need to consider pipe rupture events and the pipe-whip protection requirements.

The USNRC Alternative Criterla(2Q) is presented in Appendix C.

GPUN adopted the USNRC Criteria, and it was used for this analysis.

I 1.2 USNRC Alternative Saf ety Assessment Criteria.

I In order to be exempt from the requirement to InstalI pipe-whip restraints, it must be demonstrated that the particular piping in question exhibits:

A) Detectability Recuirements. A detectable leak, for both longitudinal and circumferential th rough-crack lengths of A t (A times the wall thickness) j j

under normal operating loads, where A >2 and A is to be determined, plus j

j B.1)

Integrltv Requirements.

Level D.

Stabil Ity of both longitudinal and circumferential cracks that have a length equal to A t+2t under Level D loads j

must be demonstrated; lower bound material properties are to be used, plus B.2) Integritv Reautrements. Extreme Conditions.

Stabil Ity of a circumferential crack that is the greater of A t+2t or 90 degrees circumferential length j

l under fully plastic bending loads; hanger effects are to be neglected;

(

snubbers are to be assumed as inef fective unless specially justified; lower 1

i bound material properties are to be used.

!I C) Sub-critical Crack Growth.

Consideration shalI be given to the types of sub-critical cracks that might exist in the piping system.

5

D) Aunmented ISI. An optional cpproach involving special Insp2ction procedurcs may be used If other corrective measures are not practical.

The solution to the USNRC Criteria A) requirement involves an analysis which I

utilizes linear elastic fracture mechanics methodology and computation of leak l

rates. The normal (or Level A) operating stresses are used to compute the crack length that would result in a detectable leak rate. These calculations are reasonably straight forward and require the gathering of the stress analysis results.

The USNRC Criteria B.1) calculations require the postulation of a crack having a length of A t+2t (t = the walI thickness of the pipe.) The "+2t" amount j

is included to permit a margin for sub-critical crack growth due to fatigue or stress-corrosion.

The crack is assumed to be oriented longitudinally or circumferentially and extends through the wal1 of the pipe, it must be shown in USNRC Criteria B.2) that, using lower bound material properties, the piping wilI not exhibit Instability under upper-bound loading.

The upper bound loading assumes that the section containing the crack is fully plastic and the crack is oriented circumferentially. The upper bound loads are assumed arbitrarily to be equivalent to the bending moment required to Induce a fully plastic section at the crack location. This upper bound is one means of accounting for extremely low probability events such as water hammer and snubber failure under seismic loading.

Note that such loads are not expected to occur, but are felt more realistic than the load used for typical pipe-rupture analysis; namely a load that occurs instantaneously and is equal to the ultimate strength of the uncracked section of the pipe.

I I

I 6

I 2.0 CRACK STABILITY CRITERIA I

In order to analyze the stability of cracks in nuclear piping using fracture mechanics methodology, it is necessary that the material properties, distribution I

of crack sizes and shapes, applied loads (or stresses) and the crack stabil ity criteria be specified. As all but the crack stability criteria have been defined aforehand, only this aspect will be discussed.

Because Code loads are determined, as a matter of course, using linear-elastic structural analysis methods, it is logical to first examine the applicability of LEFM methods.

2.1 Linear-Elastic Fracture Mec'anics Considerations.

Recall that the crack driving force using LEFM is defined in terms of the And, for stabil Ity, K co' puted f cr the applied stress stress-Intensity factor, K.

m and crack size must be less than the fracture toughness, K f

he piping lc, material.

Stabil ity can also be defined in terms of the J-Integral; which, for 2

LEFM, can be computed from J=(K /E). The parameter J can be considered an lc equivalent toughness and thus for J<J lc, stability is insured.

Because J is used throughout this report, consideration of LEFM methods is presented in terms of J.

Exact J solutions are typically not available for the postulated crack geometries and loading.

Thus, estimates of J are developed for the crack geometries of Interest using the accepted practice of basing J estimates on plastic zone corrected stress-Intensity f actor solutions (i.e., K(a+Ry)).

(Note that estimates of J based upon K solutions, that i s,

LEFM, result in unconservative estimates of J as the limit moment of the cracked section is approached.)

I I

7

2.2 Ela-tic-Plastic Fracture Mechanics Considerations.

Before ensidering the application of EPFM to a typical piping probim, it is worth reviewing some theory.

In the application of LEFM to brittle f racture, crack Instability is assumed to be incipient when K>K Physically, this is ic.

interpreted as instability that accompanies the onset of crack extension.

But, in ductile fracture, it is known that crack instability does not generally accompany the onset of crack extension. Rather, the K (or J) at instability can be welI above the K r J lc) p int.

It is important, from design and safety ic considerations, to be able,to take advantage of the higher J values (or loads) that co-exist with the stable crack extension but, until the recent development of the tearing modulus concept, it was not possible.

Analysts had been faced with the probi m of using a J value for Instability predictions unless k

representative R-curves could be developed which were typical of the significant material dimensions actually used in structures of Interest.

In order to expand on the use of the J-integral, it is necessary now to delve into computations in the elastic-plestic and fully plestic states.

Note that the foregoing paragraphs presented J values based or. K solutions that are not valid under elastic-plastic or fully plastic conditions.

I Solutions to probi m s in EPFM Involve expressing the Intensity of the crack-tip deformation field by an appropriate elastic-plastic fracture parameter.

Based on the fracture parameter, the behavior or growth of cracks can be expressed f unctiona l ly.

It follows that the use of a parameter like J Infers that crack growth is controlled or determined by the value of the parameter.

This logic Ieads to the term "J-controlled g rowth. " Typically, the quantifying of the fracture parameter is accomplished by computing the value of the path Independent J-Integral, developed by Rico (B), either by use of direct integration around the crack-tip or by use of any one of a number of acceptable estimation schemes.

Relative to any J computation, it is Interesting to note that the phrase 8

I clastic-plcstic f racture m::chanics inf ers that problans involving p!csticity can be analyzed for any typo of loading.

But Rice (f) proved the path Independence of a material which exhibits J only for the ideallzed case of no crack growth and l

"non-linear elastic" behavior.

Unfortunately, real materials do not behave exactly as non-linear elastic materials and the problans of Interest involve crack growth.

However, the violation of this idealized behavior is not suf ficient to invalidate the path independence of the J-Integral if certain restrictions are a need for these restictions, Hutchinson and Paris (2) set-forth met.

Based on strict theoretically based guidelines for J-controlled crack growth.

Extensions beyond those limits are possibt'e under the conditions discussed in Ref erence (A).

I Although the value of J is Indicative of the intensity of the crack-tip I

def ormation field, it is not sufficient for resolution of the question of stability. To resolve this, Paris, et al.(1) def ined a non-dimensional parameter, called the tearing modulus, which assumed the validity of J-controlled growth.

It is applicable to material property data and applied loads alike, and is expressed T=E

.dl (2-1)

CT* da where E is the elastic modulus, a is crack length, (T is a flow stress, and J is g

the J-Integral.

J contro!!ed growth requires that the crack extension, da, occurs under the equllIbrium condition I

J (2-2)

J,pp 9,

=

I which applies whether or not stability of the crack extension is present.

In this expression, J

is the value of J on the material J-resistance curve, and the g,

is the computed value of the J-Integral for a given load and crack length.

J,pp For a crack under the preceeding equilibrium conditions stability is determined from T,pp T,9 (stable)

(2-3)

I 9

(2-4)

T T

(unstablo) g, where T,,9 Is determined from the material J-R curve and T is dependent app upon the crack geometry and loading existing in the actual structure.

This stability criteria has been experimentally verified for several specimen types.

Paris, et al.

(.1), were the f irst to demonstrate applicabil ity through experiments using A471 steel 3 point bend bars in a test system of variable compliance.

The variable compliance feature was used as a means of controlling the T,pp.

Similarly, Zahoor and Kanninen (10) tested circumferentially cracked 4-inch diameter TP304 stainless steel pipes in 4-point bending, and Gudas and Joyce (_11) evaluated several materials of varying degrees of toughness in 4-point bending.

.I i I I

I I

I

.I 10

s 3.0 J ESTIMATION AND LEAK RATE ANALYSIS gpp I

3.1 J-! ntegral Est!=ation.

I The J-integral, J,p p, was calculated using the relation J,pp

, (1-v )

(3-1)

=

I where K, is the opening modo plastic zone corrected stress-Intensity factor, E is the elastic modulus and v is Poisson's ratio.

3.1.1 circumferential cracks. For circumferential cracks, the K, consisted of contributions for three types of loads:

axial load, bending moment and membrane stress due to pressure. The K,

due to pressure loading, K,,

was obtained by utilizing the solutions from Ref erence (.15), giving I

K, 0,'/TTRI F,

(3-2)

=

where O,' is the membrane stress (extal).

F, is the shell correction f actor that depends upon the length of the crack and the geometrical dimensions of the shell.

I The K due to the applied axial tension load is g

0~ /TTiiG F (3-3)

K

=

9 9

9 where F depend upon the same parameters as F,.

The function F can be derived 9

9 from the recent work of Erdogan and Delale(16).

FPDC has developed its own approximate, but conservative, expression for F, which was used in this study.

O,' is the axial stress due to the tension (pressure) load P U~

= P/(2TTRt)

(3-4) 9 11

Similcr to the tension loading casa, FPDC had previously dsvoloped en estimate of K for the externally applied bending load; and the K due to this loading is K

(b/URO F (3-5) b b

where F Is a correction factor for a circumferential crack in a shell subjected b

to a bending load.

O'b Is he maximum bending stress due to the external moment, M,

2 O'b M/(TIR t)

(3-6)

=

The total K, due to these three types of loading Is I

(3-7)

K, K,

=

+ K9+Kb Equations (3-7) and (3-1), when combined together, give the functional form for can then be found by differentiating the equation The form for T,pp J,pp.

with respect to crack length, <;lving for J,pp U-8)

T, p

,pp g

=

3.1.2 t_ono t tud i na l Cracks.

The computatation of crack stabil ity for longitudinal flaws is based on plastic zone corrected stress-Intensity factor solutions.

For a longitudinal through crack in a pipe K = S / Tc F(X)

(3-9) h where S is the hoop stress, c is half the crack length, X=c//RT and the shelI h

correction term F(A)=(1.+1.25X ).5 2

I 3.2 Lenk Rat.g Ana l vs I s.

I 12

The estimate of the leak rato for various cracks tras based upon the mathods given in Reference (H).

In general, the leak rate depends upon the appIIed stress and crack length. Thus, the calculation of leak rate necessitates the development of a fluid flow model for fluid leaking through a crack.

It also requires consideration of the thermodynamics of the flow and the surface roughness of the crack. The simplest leak rate analysis (H) assumes that the opened area of the crack can be modeled as a rectangular slit of constant height.

This idealization was used herein.

I The typical presentation o,f the calculations is by a curve of leak rate vs.

crack length, as shown in Figure 3.1.

It is noted tnat the appiled stress is a parameter. The higher the stress, the larger is the crack opening and the larger the corresponding leak rate. The locations at the 75f t. elevation are expected to experience stresses of 5-6 ksi under Level A loading.

The estimated leak rate behavior for these stress levels are shown in Figure 3.1.

I I

I I

I 13

I 1000 l

o - 28.1 ksi i!I I

100 o = 9.11 ksi

I o = /,;,o kN 6~= 5 o Ksf i'4

~

e l

I 10 I

.e*

a 1

l l

l a/2c = 0.1 I

I I

0.1 10 20 30 40 Tnrough Crack Length, 2c, inches I

Figure 3.1 Eff ect of the Applied Stress on the Leak Rate Through a Fatigue Crack (.11)

,I I

i4

I 4.0 ASSUMPTIONS & FORMULATION, FULLY PLASTIC ANALYSIS I

Tada, et al. (.12) were the fIrst to apply the tearing modulus stability criteria to actual structural problans. They applied it to a piping system for l

the purpose of evaluating the stability of a circumferential crack in a BWR l

recirculation loop.

In Section 6, the anergency condenser piping systems will be analyzed based on an extension of their analysis methods; that is, USNRC Criteria B.2.

In this Section, the method of analysis and the crack stability criteria are discussed. The analysis method presented below follows Tada's approach (12),

but takes Into account the behaylor of structures having more complicated boundary J,,,

is included conditions.

Additionally, the dependence of T,,,

on through the use of a J-T stability diagram.

Details of the cracked section are shown in Figures 4.1 and 4.2.

Only one th rough-th e-th ickness crack (TC) is assumed to exist and that crack is oriented circumferentially. Use of a TC assumption has been justified, per Criteria B.2, by Zahoor(18) for fully plastic bending.

He considered circumferential vs. radial Instability for part-through cracks (PTC) and concluded that the PTC becomes a TC.

See Figure 4.3.

Under the postulated loading, the following conditions are assumed:

a) The cross section containing the circumferential, through-the-th ickness crack (Figure 4.2), is fully yielded.

I b) The material local to the cracked section (or hinge) ' exhibits rigid perfectly plastic behavior.

mtd T,pp.

4.1 Structural Resnonse The behavior of the pipe is idealized as sections which behave elastically, separated by a plastic hinge. To compute T,pp, there are two system parameters which must be evalueted. The first is the compliance of the elastic section and the second is the rotation of the plastic hinge at the assumed crack section under the prescribed loading.

15

I I

I I

Z I

I f-(

g 3

a I

I is o

e

(

x -ey 3

fr

=

l

~N3 l

m/

c

'I i

e I

lI lI I

I lI ie

m e

e e

aus e

e e

e e

e e

e e

e e

e e

m gCRACRED s

l k

(

/

FULLY p l

Figure 4.2 Fully Plastic Bending of a Pipe

ll

g II l

I 1.0

, ' s X=a/t [a

's a

4 g

?

gs i

i g\\

ll 0.6 4

\\

X=0.1 gg iI 0

s l

X O.8

\\

!g

\\

\\

i X = 0.5 g

\\

\\

0.2 s

N lg AXtAL LOADING g

g

---

PURE BEND LOADING g

g i

\\

\\

'g I

I I

OO O.2 0.4 0.6 0.8 1.0 I

8w 1

'I j

Figure 4.3 Stability of Part-through Crack Under Fully Plastic Bending iI 18

I I

4.1.1 Comollance.

By using fInito olonent mothods along with the essumption that a plastic hingo is developed at the cracked section of the pipo, the rotational compliance of the clastic section about the hinge location is determined using the JTPIPE program (14) described in Appendix A.

Note that the elastic compliance does not depend on the crack size because the crack section has boon idealized to behave as rigid perfectly plastic; thus, only the uncracked section of the pipo behaves ela'stically.

4.1.2 Pl astic Hinge Behavior. The rotational response at the plastic hinge simulating the cracked section requires computing the finite discontinuity in rotation taking place at the cracked section, 6

(See Figure 4.2).

The cr solution is developed by satisfying compatibility at the hinge.

This discontinuous rotational angle is due to the localized def ormation at the fully plastic cracked section.

4.2 Cracked Section Parameters.

I 4.2.1 Plastic Limit Mamert. The plastic limit moment of the cracked section, I

(M can be defInod in terms of simple parameters.

For a th in pipe, (t/R)<<1, c p, Tada, et al.

(12) have shown that the (M )p can be expressed as c

2 1/2 sin (0))

(4-1)

(M 'p R t(cos(0/2) c o

where 0;istheflowstress; and R, t, and 0 are, respectively, the mean radius and thickness of the pipe and the angle defined by the through wall crack.

(See Figure 4.1).

I I

19

It is noted that 4(T R t is the plcstic lImit mom:nt of cn uncracked pipe g

section, M.

p For convenience, the nondimensional plestic 1Imit moment of the cracked section can be obtained by normalization with M.p I

(M }p cos(0/2) - 1/2 sin 0 (4-2)

=

c 4.2.2 J-Integral.

For the fully yleided cracked section and the rigid-perfectly plastic material behavior assumed above, the J-Integral can be expressed as f of lows (4-3)

F) O 0," R J

=

cr, where I

F

= sin (0/2) + cos 0 (4-4) y I

and O is the rotational angle caused by the plastic hinge at the cracked cr section.

4.3 stabil Itv Ana lvs t s.

a procedure The approach used to determine stability of a crack is based on similar to that developed in (.12).

It is assumed that for a fixed displacement loading, the sum of the displacement changes at the cracked section, which can be separated into the elastic part and the plastic part, should be equal to zero.

Carrying through with the mathematics, we find I

J,pp /(O'g*

E R)

(4-5)

T,pp=

F (0)

L,ff/R

+

F (0) j 2

where 2F /TI F (0)

=

j j

20

(cos(0/2) - 2 sin 0)/2F)

F (0) o 2

L,,f El/[K],

[K] = min. stiffness at hinge

=

end F Is given by Equation (4-4).

y Note that Equation (4-5) depends upon the geometric configura1 Ion as well as the boundary conditions of the piping system.

4.4.The L-I Diagram.

When using the tearing modulus stabliIty concept for the assessment of safety I

in a nuclear piping, it is convenient to plot the results on a J-T diagram. The J-T diagram compares the applied (or calculated) values of J and T with the material (Invariant characteri stic of a material) values. That is, a plot of T,9 to determine vs.

is compared with a plot of J vs.

J,pp T,pp if the T,pp value is less than the T,9 val ue f or the J,pp value specified. A sample J-T diagram is shown in Figure 4.4.

The applied curve shown was generated for a through-the-th ickness crack having 20 = 120 degrees. The procedure for generating an applied J-T curve is as follows.

For a given piping and the material, the radius R and O', are known.

For a specified 0,

F is calculated using Equation (4-4).

With all the above y

Information J can be computed, using Equation (4-3), if 6 is specified.

The rotation caused by the plastic hinge at the crack section, 0, depends upon the Interaction between the various segments of the piping and the boundary conditions imposed.

Because the limit moment of the cracked section is reached prior to lc, and it is reasonable to set J,pp equal to reaching J ductility is assured J lc*

I I

l 21

M M

M M

M M

M M

M t

l APPLIED M ATERI AL N

N

( UMsTABLE) i N.

I N x I

V I

\\

N l

M (STABLE)

N

'N IC Figure 4.4 Schematic of J-T Stability Diagram

1 are obtained via The T,pp values corresponding to the computed J,pp Equation (4-5).

In this Equation, the only quantity that is unknown is Lof f, the ef fective length of the piping.

Since actual piping systes are typically 3-dimensional structures, it is not always easy to compute the ef fective length I

(of an equivalent straight-pipe) by simple analysis methods.

Hence, a finite elmont analysis of the piping system is perf ormed and the elastic compliance is computed from which Leff is determined. The computed (applied)

values, J

p are then plotted, giving the applied curve as shown in Figure 4.4.

T,pp, and Note that the appplied J-T curv'e shown does not account for crack growth, that is, I

20 is assumed constant throughout. The error resulting from this approximation is small and allows conservative conclusions to be drawn from the analysis.

The material curve shown on Figure 4.4, was derived from a J-resistance curve for the material of Interest. The details of this can be found in Appendix B.

T,pp values are dependent upon the J For Note in Figure 4.4, that the p.

T,pp is less than the T, values, stable crack behavior is cases where the lower T,,9 value corresponding to higher J,pp assured.

On the other hand, a

values can cause unstable behavior.

I The procedure for generating the J-T diagram described above will be used for assessing the stability of all the lines discussed in Section 6 of this report.

I 4.5 Extraonnition nt.t.ha J,9-T Curve.

In Instances where the J-resistance curve may not be available for higher J values, the assessment of whether a system is stable or unstable based on a J-T diagram may require extrapolation of the material curve. One way of extrapolating the resistance curve is to assume that the matorial continues to tear with the same slope. Recent exportments on a large diameter pipe (.13)

Indicate that this assumption may be valid.

This will mean that in the extrapolated regime, the 23

I T,9 remains constant.

The extrapolation of the materici curvo on the J-T diagram (Figure 4.4) will then be a vertical line extending from the maximum J,,,

value point on the material curve.

An alternative to this extrapolation is to assume that there is no further increase in the J-resistance with crack growth. Such a behavior would imply that the T,,,

reduces to zero in the extrapolated regime. That is, on a J-T diagram such as shown in Figure 4.4, the extrapolated material curve would be a horizontal lIne.

These two extrapolations r'epresent the upper and lower bounds of resistance I

curve behavior for continued growth.

In reality, the T,,, val ue is expected to decrease gradually with increase in the J,,, value, leading to the possibility of a zero value of the T,9 at some higher J,,,value.

The validity of J-controlled growth is dependent upon the satisfaction of several requirements.

One of these is that be "large." For typical Type 304 stainless steel, valid J-resistance curves may have over 1 inch of crack growth and values from 10 to 20. Considering T,9 to decrease abruptly to zero, simply implies that the value, which is proportional to T,

also decreases to zero.

This would invalidate the assumptions of J-controlled growth, and any assessment of the stability of piping would be subject to serious error.

J-resistance curves need to be developed to include extended amounts of crack growth while satisfying the J-controlled growth requirement.

Because of these Ilmitations the assumption of a constant but non-zero value of the T,9 is felt appropriate.

In this report, this assumption was made where desned necessary.

I I

1 24

5.0 SYSTEM DESCRIPTION I

The Isolation condenser piping runs from the reactor vessel to the isolation condonsors.

The piping includes portions that are betwoon the drywalI and the reactor vessel and other portions that are outsido the drywalI and run through the isolation condonsor tanks themselves. As stated bef ore, our concern is with the demonstration of leak-bef ore-break and unconditional stability of def ects at the 75 ft. elevation.

I 5.1 Plo f ng Systan.

I 5.1.1 Peturn lina. The return lines begin at the ends of the Isolation condensor tanks on the 95 ft. elevation. The lines run basically straight down out of the tanks through the 95 f t. elevation floor through a hole in the floor which is f il led with a grout. These linos at this point are 8 Inches In diemoter and have en insulation thickness,that was measured, of approximately 3.25 inches.

Whoroas the supply lines went through the 95 f t. elevation point at only two locations and then branched above the floor, the return lines go through the 95 ft. elevation floor at four different locations and each pair of return linos from one tank are routed along this space in the 75 ft. elevation until they join into a 10 inch lino. Af ter joining into the 10 Inch line, the piping runs parallel to the 75 ft. elevation floor at an elevation of approximately 87.5 ft.

The return iIno piping at the 75 ft.

elevation has a number of spring hangers and an appropriate number of hydraulic snubbers and 3 rigid supports.

The isometrics (21) for these lines are shown in Figures 5.1 and 5.2.

5.1.2 Sunniv line. The supply lino piping comes through the containment penetration at about a 90 f t. elevation and runs parallel to the 90 f t. elevation until it turns upward and penetrates the 95 ft.

olevation floor.

Imediately above the 95 ft. elevation, the lino branches into a Y junction and the 16 inch supply line is branched into 2-12 Inch linos. The 12 Inch lines rise vertically 25

}

, bh 9'g el

\\

..('

-]4, gg Y

_,,', A

,,..r

• s y, m.%

g

)

.x<

i saQ $ t q$

al-($,/c w,

i d

~ ~

'g f as

.A 4@

p s '<

(

r p

\\

w

/

\\

s' s ef f

• /

ja e

K /

s p,L A,

/

\\

/

l 2.p

~

t f:l 1

(, is

.. /

f

)

s

/

13

f W

/

/\\

/

\\

'$N e

O p

1

\\

i

/

yb irs t

,~

,/ ras plq

(

3-i

)

[

g.

/,

itv s

%/

\\

c

.=n w

/

d in

'Q '

4, * //

T s

1

/;+9 3

6*. ;'s)/

%'s

/ ~

i 4i h

'x

's

/.

['

,1 a

/

\\

p' 1

' '. w _

s

.;s

,,/

's

\\

- > +,

/

t l

\\

l 1

1

~

il

~,.J q l'j > g em gure 5.1 Oyster Creek Emergency Condenser Return Line A (s

.'~ y/

+ g e,,

~

~

~

~

c g

p'

~

l'x -

)

as

~

\\

's.

e

'~n

\\

'stN/

x s

m.

\\

/. x m;m _

/ 9.,

s 4

ID

/.

,g,

/

~~

s

-x r

x st2-u x

V

-a f.o.

s g

-x s

m I,,: '^

r I.

• )

4, 7,

s,

s

}

%g

/

%y-

/

i.

'-?> '~

, l' -

,.f

'o e-Figure 5.2 (

[

a sx

/

q'

/

f*

f 6

.~

6

'

/

l

..t...

gi l

l 4"

f

m

- P

/p

/

~c/,

'"~

p/

r\\

'Z=.

/

d s

l

,, /

, /' '

~%

4:6 9, Y, /

N..

-s -

4 s

. v "< e s

f

, \\ 2 ' ! o-

~

[

^

\\

- \\I'.

1f

,,/

sai a s f

-o, p

..,/

<y 1,

1. )*

/

C. Ca

. a eg 6.

v

~

gg f-Ih e

us

,78 e

fi W

f.

QJ

,\\

s,e : -

s

~

y' i

l/

l

-g/~.)

yyj

'x

-o

's;

+-,g,'

\\

N

\\

/

n' e

I l

l t

i i

l I

pstcr Crsek Emergency Condenser Return Line B 3 r 1

i k

l e,, s,,9)$

r >

4 l

e t

l

%*J i

1

-A7 i

/

el X

/

/

NL ? wo j ut I ha s.

4:4pe s

., N t8 7

% "e

. s:. ?+ /

e. <-

/

./

. e,,

)

• 4 *i 4

• J eG,

.y yi

/

5,,-

' /

4,,,d

- =

N t-

. '^sf '

.c ss v.

/

e'

/

,,p-27

\\,.3 :

. / /

N l

~

g

. \\s\\

/

,s-x s

~

I to a point within two f eet of the ceiling in the 95 f t. elevation and then run along the axis of the tanks and enter the ends of the isolation tanks.

The penetration of the supply line is very rigid in construction and for all practical purposes does not permit any displacements; thus, it can be considered an anchor point.

The hole where the supply line piping goes through the 95 ft. elevation floor is grouted with a high strength grout having a compressive strength of some 2,000 ps!.

Thus it can be considered that the movanent of the line is severely restricted as it goes through the floor.

Just above the 95 ft.

elevation floor, the Y fitting can be considered a type of displacement limiter.

For example, if the I!ne were to try to displace downward, the Y fitting is much larger th an the hole and thus would limit the displacement. Similarly, an upward motion of the supply lina would be II.nited by the lines striking the ceiling and columns in the 95 f t. elevation space.- ~ The Insulation on the supply lines is approximately' 31/2 inches as measured during site inspections on December 10, 1981.

Basic observation of the supply I!ne, 75 ft. elevation is that the system appears to be quite stiff and thus should exhibit a very low T The two lines are shown in the piping Isometrics (21) of Figures 5.3 and 5.4.

5.1.3 Surrounding Structmr.n. The spacing between the four penetration points X-3A, X-3B, X-SA, X-58, is on the order of less than 3 ft, center to center.

Numerous ductwork, cable trays, conduits and other smali diameter piping systems run in close proximity to both the supply and return Iines.

5.2 Ploing fada Stress Analysis.

The piping stress summary given in Tables 5-1, 5-2 & 5-3 is excerpted from Reference (12). All Stresses are in ksi units and are to be used for the analysis of circumferential flaws.

For the analysis of longitudinal flaws, only the hoop 28

s

+g s

1 s

_4 h

ers_

• /

/

/

V

/; m;;,,

x

~

!M-N g.-

f

- x

=

A

.p

/

t i..

p p

A

.0

5,

4 i tsA pt l,la s y 9

a j

.s'

}?

% i ") p*;" /

s

• ?, ~

./ a

• {

p,%,

~l

G. ; &p %'

~

.,, ~ v/

/

a

,.s.~.

./

,h s'

/

y 7

, /, '

,a'

/

/

. g ze,\\ r-/ cr/ g

/

e'

/

/

..,zf

.c

. ;,e s

't-

.s.-

s.

, ',4 -

g V,L @-

g-w/

/

1 y

see cr e :wr s.c t c'"t

/

i.,

.. ~

(f'b

(

l

l N

cC y

L&

y

,/

'a

/

y

\\

h d

,,$,/,

, ' '.e e _.

/

s N

y' l

t 1

y t

[i

,' N Qgo -

e

.... O

/

x N,

N

<sNN'S %,

s h,

,V

/

y T

l-

)

+:~u#

l

/

l

/

jb

+~ #

t /j\\

A

/ f t

+

~

{

'/

j

/]

/

~

/

d.

i o

Figure 5.3 Oyster Creek Emergency Condenser Steam Line A I

l.

,m

</

331;

/

THE HEC Mil

/

ef L.,

SHAD M Sulk g

g/ j l

q PIPE WELDS M OTHERwlSE, il M vt$UAQY l 3..

4

/

I' eEGS DURi%

h..

l u

\\,%

s-s

• i

\\

y

/

x \\

24

~

) I s

N

~. "~

1;h x-.

e s.

x x

s

'\\

q-s m

e ar_-

',,'./,

a,<.

Mi I e s, p

.1.. s 4

,g

'9, y.

x

+

Se, b

V.,

x

..< r.

N.,

j N

8 l

o' s

sg x

//

N j'u

+

y

$o f

If,

,/p d

n t

t '* t, e';

y

/s

'y t *,a

,, 'st,,

w p

u Q,, '

V

.- p,

I

?;x a

3

,/>

~; v f.

N,

,i '

,fj jf

~~ \\(,..,

gh'

~~

~'

s i

[p.,

/

,~

, "f t

, - / r -v

/

Y '

/, 'l/

./e

.'o

• • Y

'/

N%,

/.#

"/

r I

/ y:+

,/

s y

_ [ g.f-j

• s.

s m

4

' %l-

./

f

' _/,

c**,.-

N s

..s

\\

s' s

s

/^

s av

<6V s

s.

fD 1

s

zg f,' 0

/-

s v

-/

%sh b

L)

)

+

..'qg.

7 s

O..

{f

,; s l

}

L

~*

J...

~.

s i

. u =. '.s j

,.r/

'v p

y s ';

4

' L i

i s.

',/

6.

s s

.~

+'

p'

/j/.

<m 4 x/

i p

>l ' -'I

/

/,

'x,

//

~/,'

y p

s.

4 1

/

,/

._r

u

...w

)'

\\

Figure 5.4 Oyster Creek Emergency Condenser Stearn Line B J

I r.

I E

NE elt BEidEE4 THE %PPORT AC TW

'J 54LL EE SLBJEET TO %RFAEE EWIW 4

P!PE wil05 arf SUB4CI TO Y0 unit"

/

g OT*ER.15E, THE 5:JPPORT-TO-PRE 15eEE k' e

/

• ,g.8 EE v15GAtLV EWlhED 14 THE SAFE fue 3

p

.n:s ausm wr.,a0sTAfit, PatsSaRE.

/,

e sie r...t r o.,n a

h, atp. pts,6.

.: r.,u..

\\

A*,

A I

.50 t

4, u.,

.', '.'. T s

\\a

'n 9.

I".

i 4

I stresses need be considered. Because these are constant along the piping and only depend on the pipe size and internal pressure, they are not summarized until used in Section 6.

Tabl e 5-1, 8 INCH RETURN LINE STRESSES, 75 FT. ELEY.

Level A Level D Cmoonent Line A*

Lina B*

Line A**

Lina E*

Pressure 4.4723 4.4723 4.4723 4.4723 Dead Wt 0.120 0.1'10 3.562 2.093 Seismic 0.0 0.0 4.667 3.102 Thermal 0.482 2.1 85 5.723 11.550

• Stresses given are: Min. Dead Weight and Thermal of all locations along the line at the 75 ft. elevation.

• • Stresses given are: Max. Dead Weight, Seismic and Thermal of alI locations along the line at the 75 ft. elev.

Table 5-2,10 INCH RETURN LINE STRESSES, 75 FT. ELEY.

l Level A Level D

)

Cmoonent Line A*

L in a BJL Llne A!L*

Line BH Pressure 4.7364 4.7364 4.7364 4.7364 Dead Wt 1.32 0.151 2.8 83 1.898 Seismic 0.0 0.0 3.533 2.017 Thermal 1.267 1.315 3.379 7.048

• Stresses given are: Min. Dead Weight and Thermal of alI locations along the iIne at the 75 ft, elevation.

• • Stresses given are: Max. Dead Weight, Seismic and Thermal of alI locations along the line at the 75 ft. elev.

Table 5-3,16 INCH STEAM LINE STRESSES, 75 FT. ELEV.

Level A Level D Cmoonent Line A*

Lina BJL Lina A*

• Line BjL*

j Pressure 5.011 5.011 5.011 5.011 i

Dead Wt 0.211 0.335 0.794 5.311 Seismic 0.0 0.0 1.294 1.839 Thermal 0.572 1.013 2.439 7.665

• Stresses given are: Min. Dead Weight and Thermal of alI locations along the lIno at the 75 ft. elevation.

• • Stresses given are: Max. Dead Weight, Seismic and Thermal of alI locations along the line at the 75 f t. elev.

31

I 6.0 RESULTS & DISCUSSION 6.1 Leak Detectabil Ity.

I The USNRC Oriteria A) requires the demonstration of stability of a crack that length that would result in a detectable leakage rate.

For this analysis, has a rates of 0.1 and 1.0 g.p.m.

under normal or Level A loads were selected as being representative of a detectabl e leak rate.

Such rates are readily detecable by existing sensors or by visual inspection.

There are no particular load assumptions required as the, normal operating loads and associated stresses at various points in the piping system were determined as a part of the stress analysis of this plant. The stresses were summarized in Tables 5-1, 5-2 & 5-3.

6.1.1 circumferential Flaws. The leakage rate computation is based on Level A (or normal) stresses that are equal to the sum of the pressure component plus dead weight component plus a thermal stress component. No dynamic loads are used in developing the normal stresses. At each point of Interest, the demonstration is made using the minimum of the stress components after considering all locations.

This is conservative because the actual combined of fective stress at any location would be greater than that computed using the minimums.

And, the lI lower the of fective stress, the lower the leak rate and the longer the crack must be in order to have a detectable leak.

Leakage rates are then computed for a series of crack sizes based on this particular stress condition and the crack l

lengths which result in 0.1 and 1.0 g.p.m.

rates of leakage are thus determined.

The results are presented in Table 6-1.

l 6.1.2 l_ongitudi nal Fl aws.

For the longitudinal flaws, the 8 inch line has a hoop stress of 8.9 ksi.

For this stress, a 4 inch crack results in a leakage rate of I gpm. Because this hoop stress is lower than either the 10 or 16 inch

lines, it is used for computing a crack length, A t=4 inches, that is conservative for j

all lines. Thus, A =8 and the flaw size for the Level D analysis is equal to g

l 32

8t+2 t= 10 t.

I Table 6-1, Circumferential Flaw Size Assumptions vs. Leak Rate Pipe Size Sef fu A t(.1gpm) A,t+2t A t(1gpm) A t+2t j

j j

Dnom t

ksi inch inch Inch inch 16 0.843 5.6 3.6 5.286 7.7 9.3 86 10 0.593 6.0 3.3 4.486 7.2 8.386 8

0.500 4.9 4.0 5.000 8.2 9.200 (S press + (Sdw + Sther)2).5

• Seff

=

6.2 Level D Loads.

6.2.1 Longitudinal F1aws.

Conservatively, alI locations can be bounded by 10t/2 and F(X) from Table 6-2 and using the maximum values of S,

c =

h substituting these values into Equation 3-9.

In this manner we find a J 235

=

in-lb/in (from J=K /E), which is much less than J ic, thereby Insuring ciack I

stability.

Table 6-2, Longitudinal Crack Stability Under Level D Loads D

I'I" 8,ksi 10t,in X(10t) nom h

8.0

.50 8.9 5.0 1.76 10.0

.59 9.5 5.9 1.71 16.0

.84 10.0.

8.4 1.67 6.2.2 Circumferential Flews. The solution of Equations 3-1 th ru 3-8 for circumferential flaws was obtained using a computer program which perf ormed the necessary iterations on K to obtain the plasic zone corrected K values.

From the K(a+r )

values, the appropriate J,pp estimates were determined.

This evaluation was perf ormed using the conservative load combination given In Tables 5-1, 5-2 and 5-3 and the results are presented In Tables 6-3, 6-4 and 6-5.

33

I J

insuring stabil ity and complianco ulth the J,pp<

lc In all casss, requirements of Criteria B.1.

Thus, no crack extension is Indicated fcr those conservative load assumptions.

Even if crack extension did occur, exce l l er.t margin for stabliIty Is evident because T,pp<< T,,,.

I Table 6-3, OYSTER CREEK RETURN LINES A & B, WORST CASES, 8 lN. LINE LOADS = LEVEL D 57079.

Mapplied = 0.29943E+06 Poper =

1250. ps!

Faxial

=

11550.

Smem 4472. psi,

Saxial 4472.

Sbending

=

=

=

PlPE OD =

8.625 THICKNESS =

0.500 Sflow =

45000. psi CRACK LEAK AREA Japp Tapp LENGTH lH IN#*2 IN-LB/lN##2 4.40 0.027 0.17642E+03 0.21753 E+01 5.40 0.046 0.26329E+03 0.29155E+01 6.40 0.073 0.38146E+03 0.38764E+01 7.40 0.110 0.54139E+03 0.51213E+01 8.40 0.159 0.75692E+03 0.67388E+01 9.40 0.223 0.10466E+04 0.88545E+01 I

I I

I I

I 34

-g---

I Tcblo 6-4, OYSTER CREEK RETURN LINES A & B, WORST CASES,10 IN. LINE LOAD = LEVEL D I

898C1.

Mapplied = 0.52661 E+06 Poper =

1250. psi Faxial

=

4746. pst 10960.

Smem 4746.

Sbending Saxial

=

=

=

PIPE OD =

10.750 THICKNESS =

0.593 Sflow =

45000 psi

)

i e,;

b CRACK LEAK AREA Jepp Tapp LENGTH,IN IN#52 IN-LB/IN**2 4.40 0.024 0.15047E+03 0.16765E+01 5.40

,0.040 0.21508E+03 0.21343E+01 6.40 0.063 0.29831 E+03 0.27006 E+01 7.40 0.093 0.40507 E+03 0.33970E+01 8.40 0.132 0.54149E+03 0.42509E+01 9.40 0.182 0.71526E+03 0.52971E+01 Table 6-5, OYSTER CREEK STEAM LINES A & B, WORST CASES,16 IN. LINE I

LOADS = LEVEL D 201151.

Mapplied = 0.22542E+07 Poper =

1250. ps!

Faxial

=

5011. pst 14820.

Smem 5011.

Sbending Saxial

=

=

=

PIPE OD =

16.000 THICKNESS =

0.843 Sflow =

45000, psi I

CRAM LEAK AREA Japp Tapp i

LENGTH,1N IN**2 IN-LB/IN**2 1

4.40 0.026 0.18556E+03 0.17459E+01 5.40 0.042 0.25106E+03 0.20693 E+01 6.40 0.063 0.32960E+03 0.24515E+01 7.40 0.089 0.42380 E+03 0.28998E+01 I

8.40 0.123 0.53 674E+03 0.34233 E+01 9.40 0.165 0.67203E+03 0.40329E+01 I

I l

l 35

I 6.3 Fully Plastle Analysis.

6.3.1.

Aool led imd. The intent of the USNRC Alternative Criteria is to insure that brittle behavior of the piping system does not occur. This is accomplished by demonstration of structural ductility in the presence of cracks.

It is assumed that the cracked section of the pipe must be capable of absorbing large amounts of energy. This is proven by applying a mment to the cracked I

limit moment, (M )p and demr... ating that ductility is section equal to the preserved through a 1 rotation of the cracked sect sn, i.e.,

O

=1.

This cr is shown schematically in Figur,e 6.1.

I This approach is more conservative than using J as he J, 6 le Ic*

6.3.2 Crack Stah t i ltv Calcul ations.

Using the approach described in Section 6.3.1, the stabil Ity of postulated cracks throughout the piping systen was examined. The analysis was perf ormed using the JTPIPE (14) program and the value was computed using the fcregoing load assumptions at the crack J,pp section.

6.3.2.1 Return LJna A.

Return Line A was idealized as shown in Figure 6.2.

The results, presented in Figure 6.3 and Table 6-6, Indicate marginal stability I

for a number of locations and good stability for the renalnder of the lIne.

6.3.2.2 Return Llna B.

Using the piping Isometric of Figure 5.2, Return Line B was idealized for analysis with JTPIPE. The idealization is shown in Figure 6.4 and the analysis results are summarized in Figure 6.5 and Table 6-7.

Excellent to good stability margins are evident throughout the lIne.

6.3.2.3 Steam Line A. This line was Idealized based on the geometry of the piping Isometric shown in Figure 5.3 and it is shown in Figure 6.6.

It is obvious f rom the results shown in Figure 6.7 and tabulated in Table 6-8, that this line 36

I has excellent crack stabil Ity characteristics.

I 6.3.2.4 Stem Line a. A simplified idealization of this

line, based on Figure 5.4, is shown in Figure 6.8.

As with Steam Line A, this line is very stabl e.

The results of the JTPIPE analysis are sumtrarized in Figure 6.9 and Table 6-9.

6.4 Discussion.

i The USNRC Alternative Criteria was completely satisfied and it is concluded that no pipe-whip restraints are required at the 75 f t. elevation of the systm.

I It is recommended that a visual leak monitor Ing system be uttiIzed as the iIno exhibits excellent stability for 90 circumferential Iength cracks under Level D loads. A special leak sensor system is not f elt to be warranted.

I I

l I

I I

II

I I

I

,I I

l" /'/M////

/

l bs re ll/

/

e 1.0 Occ, d grees figuro 6.1 Energy Absorbtion Capability of Cracked Section I

I I

I iI se

M M

M M

M M

M M

M M

M M

M M

M M

M M

M

$los V

OYSTER CREEK RETURN LINE "A" Figure 6.2 Idealization of Return Line A tot 00 i

i

[I 5

F6 4

3 l

53 3

I q

32 0

6 7

11 5

0 V,

Nu N

=

l

M M

M M

M '~ \\

TO O.

T1 o

OYSTER CREEK RETURN LINE "A"

)

O_

CRACK RNGLE = 90 MATERIAL = 59358 TP315 Figure 6.3 Stability Diagram for Return Llos A m

g o

m s

E H

m be n,

O $_

- ru x

RTL o

u No o

x go Z -

(UNSTABLE 1

]"

ISTABLE1

-1 zmm_

l i.

l i

l i

re

.e_

e WW w

w W

._____ _________ R T_fifff!EtF39KW Y_Y E M_F_-__--_---------

Ju oo

-50.00

-12.50 2'5.00 6'2.50 l'00.00 l'37.50 l'75.00 2'12.50 2'50.00 T

I I

i y

i

'l i

i OYSTER CREEK RETURN LINE

'B' Figure 6.4 Idealization of Return Line B i

kt&

53 3

32 1

l 31 q

50 B

9 2

5 0

b 48 97 15 18 40

m M

M M

W W

W oo OYSTER CREEK RETURN LINE "B"

-oo_

CRRCK ANGLE = 90 MATERIAL = A358 TP316 Figure 6.5 Stability Diagram for Return Line 8 m

g m'

i m_

s m

t~e a

eu 08_

-*cu x

RTL a

No 8

x x o Z cy-(UNST ABLE) ss IST ABLE!

CD l

m J n zmm_

r-e

.e_

e mW ww w

w

.___ M R _X X5__Y A 5y__________1_2_________________________Ja oo

-50.00

-12.50 25.00 62.50 1'00.00 l'37.50 l'75.00 2'12.50 2'50.00 o

i i

T

W M

M M

M M

M M

M M

M M

M M

M M

1 l

V a

j OYSTER CREEK STERM LINE "A"

\\

Figure 6.6 Idealization of Stem Line A 1

32 6

33 I

W 8

1 O

O 4

5 58 73 58 0

s y

6 54 g

1 l

l

. - - _ _ =.._.

W H

M oo OYSTER CREEK STERM LINE "R"

-oo_

CRACK ANGLE = 90 en HRTERIAL = R358 TP316 O

Figure 6.7 Stability Diagram fcr Stean Line A o

o e

in_

s N

RTL oo eu 08_

d "N

X (UNSTABLE) o No ISTABLE)

X X O z $-

g l

N i

OJS i

l zo l

~ 3-l 3

i 8

Wxx m

W W

I o_

tn 1

.._______________________________________________________________Ja oo 6

i i

i i

i

-50.00

-12.50 25.00 62.50 100.00 137.50 175.00 212.50 250.00 T

l t

l l

l l

V i

OYSTER CREEK STEAM LINE "B" Figure 6.8 Idealization of Steam Line B i

l 8

4 i

l 6

4 j

7 8

25 9

I w

M M

M M

M M

M M

M oo I

OYSTER CREEK STERM LINE "B" I

oo_

CBRCK RNGLE = 90 MRTERIAL = R358 TP316 Figure 6.9 Stabil Ity Diagre for Steam Line B o

o o

in _

s N

RTL

.i o

I o

n.

I C O_

-m x

(UNSTABLE) o No (ST ABLEl A

X m

yo z$-

\\

CD

~ O-A S

WWW WG W

W

.o_

LD

._______________________________________________________________Jie oo

-50.00

-12.50 2'5.00 6'2.50 l'00.00 1'37.50 l'75.00 2'12.50 250.00 T

Tcble 6-6, R: turn Lins A, 75 f t. Elevation and J for 0 1

and 20 = 90 Computation of T,pp

=

pp 00NN." NO.

Tapplied Japplied 53 0.934E+02 0.116E+05 54 0.112E+02 0.435E+04 55 0.180 E+02 0.435E+04 56 0.302E+02 0.435 E+04 57 0.512E+02 0.435 E+04 58 0.765E+02 0.435 E+04 I

59 0.627 E+02 0.348E+04 60 0.711E+02 0.348E+04 61 0.117 E+03 0.348E+04 I

62 0.802E+02 0.348E+04

, 63 0.865 E+02 0.348E+04 64 0.103E+03 0.348E+04 I

65 0.124E+03 0.348E+04 66 0.133E+03 0.348E+04 67 0.163E+03 0.348E+04 68 0.162E+03 0.348E+04 I

69 0.174E+03 0.348E+04 70 0.173E+03 0.348E+04 71 0.126E+03 0.348E+04 I

72 0.135E+03 0.348E+04 73 0.165E+03 0.348E+04 74 0.141 E+03 0.348E+04 I

75 0.110 E+03 0.348E+04 76 0.969E+02 0.348E+04 77 0.154E+03 0.348E+04 78 0.119E+03 0.348E+04 I

79 0.894E+02 0.348E+04 80 0.151E+03 0.348E+04 81 0.135E+03 0.348E+04 E

l l 88 0.970 E+02 0.348E+04 89 0.901 E+02 0.348E+04 90 0.101 E+03 0.348E+04 I

91 0.933E+02 0.348E+04 92 0.114 E+03 0.348E+04 93 0.125E+03 0.348E+04 94 0.123 E+03 0.348E+04 I

95 0.119E+03 0.348E+04 96 0.120E+03 0.348E+04 97 0.128E+03 0.348E+04 I

98 0.120E+03 0.348E+04 99 0.894E+02 0.340E+04 100 0.660E+02 0.348E+04 101 0.123E+03 0.348E+04 102 0.115E+03 0.348E+04 I

47

I I

Table 6-7, Return Line B, 75 f t. Elevation

~ I J,pp for 0 1

and 20 = 90 Computation of T,pp and

=

CONN. NO.

Tapplied Japplied 34 0.478E+00 0.435E+04 l

35 0.770E+01 0.435E+04 36 0.174E+02 0.435E+04 37 0.454E+02 0.435E+04 38 0.228E+02 0.435E+04 39 0.920E+02 0.435E+04 40 0.769E+02 0.348E+04

E 41 0.317E+02 0.348E+04

. E 42 0.870E+02 0.348E+04 43 0.744E+02 0.348E+04 44 0.749E+02 0.348E+04

', le 45 0.510E+02 0.348E+04 53 0.128E+03 0.348E+04 i

54 0.327E+02 0.348E+04 55 0.215E+02 0.348E+04 I

56 0.557E+02 0.348E+04 57 0.674E+02 0.348E+04 58 0.288E+02 0.348E+04 59 0.821E+02 0.348E+04 60 0.142E+03 0.348E+04 l

61 0.430E+02 0.348E+04 i

I I

I I

I 48

Table 6-8, Steam Line A, 75 f t. Elevation Computation. of T,pp J,pp for O 1

and 20 = 90 and

=

cr CONN. NO.

Tapplied Japplied 45 0.159E+02 0.649E+04 1

46 0.223E+02 0.649E+04 1

47 0.247E+02 0.649E+04 48 0.477E+02 0.649E+04 49 0.819E+02 0.649E+04 50 0.829E+02 0.649E+04 51 0.805E+02 0.649E+04 52 0.588E+02 0.649E+04 53 0.630E+02 0.649E+04 54 0.456E+02 0.649E+04 Table 6-9, Stesn Line B, 75 f t. Elevation J,pp for O 1

and 20 = 90 and Computation of T,pp

=

cr CONN. NO.

Tapplled Japplled I

16 0.877E+01 0.649E+04 17 0.174E+02 0.649E+04 18 0.248E+02 0.649E+04 19 0.416 E+02 0.649E+04 20 0.845E+02 0.649E+04 21 0.037E+02 0 649E+04 22 0.647E+02- - 0.649E+04 -

23 0.553 E+02 0.649E+04 24 0.576E+02 0.649E+04 25 0.386E+02 0.649E+04 26 0.451 E+02 0.649E+04 I

I 49

7.0 REFERENCES

1.

Paris, P.C.,

et.al, "A Treatment of the Subject of Tearing Instability,"

NUREG-0311, Aug.

1977.

2.

Paris, P.C.

and Tada, H.,

"Further Results on the Subject of Tearing 1,"

NUREG/CR-1220, Vol.

1, Jan.

1980; and Zahoor,A. and Instability 11,"

Paris, P.C., " Further Results on the Subject of Tearing Instability NUREG/CR-1220, Vol.

II, Jan.

1980.

3.

Tada, H.

and Paris, P. C.,

" Tearing instability Analysis Handbook,"

NUREG/CR-1221, Jan.

1980.

4.

Paris, P.C., "A Method of Application of Elastic-Plastic Fracture Mechanics to Nuclear Vessel Analysis", USNRC Report NUREG/CR-1947, June 1981.

5.

Johnson, R.E.,

et.

al., " Resolution of the Reactor Vessel Materials Toughness Saf ety issue," USNRC Report NUREG-0744, April,1981.

6.

Paris,P.C., Tada,H. and Baldind,S.E.,

" Fracture Proof Design," in "CSNI Specialists Meeting on Plastic Tearing instability," USNRC Report, NUREG/CP-0010, Jan.1980.

7.

ASE Boller & Pressure Vessel Code, Seei ton 3,1974 ed.

8.

Rice, J.R.,

Jour. Appd.

Mech., Vol. 35,1968, pp. 379-388.

9.

Hutchinson, J.W.

and Paris, P.C., "Stabil ity Analysis of J-Controlled Crack Growth," ASTM STP-668,1979, pp.

37-64.

10. Zahoor, A.

and Kanninen, M.F., "A Plastic Fracture Mechanics Prediction of Fracture Instability in a Circumfrentially Cracked Pipe in Bending", July I

1980, ASE Paper 80-WA/PVP-3, Accepted for Publication in the ASE J.

of P.

Vessel and Technology,1981.

11. Gudas, J.P.

and Joyce, J.A.,

" Degraded Pipe Experimental Program", HSST Review - VIRG Meeting, 23& 24 July 1980, Silver Spring, m.

12. Tada, H.,

Paris, P.C.

and Gamble, R., "Stabil Ity Analysis of Circumferential Cracks in Reactor Piping Systems," NUREG/CR-0838, June 1979.

13.

Cotter,K.H., Private Communication with D.M.Norris, EPRI.

14.

"JTPlPE," A Finite Elment Program for Computing Piping Syste Crack Stability Parmeters,". Version 1.5, Fracture Proof Design Corporation, St. Louis, MO.

i (PROPRIETARY) 15.

Follas,E.S., "A Circumferential Crack In a Pressurized Cylindrical Shell,"

Intl. J. of Fracture Mechanics, Vol. 3, pp1-11,1%7.

16.

Erdogan, F. and DeLal e, F.,

" Ductile Fracture of Pipes and Cylindrical

,g Containers with a Circumferential Flaw," ASE J.

of Pressure Vessel

,3 Technology, Vol.103, May 1981, pp160-168.

17.

Mayf ield, M.E., et al., " Cold Leg Integrity Eval uation - Final Report," USNRC I

Report NUREG/CR-1319,1980.

50

I 18.

Zahocr, A., Monthly Progress Report, EPRI Project T118-9-1 iI

19. GPUN Letter No.

ME/FA 117, Dated 10 Dec.1981, to FPDC; transmittal of EDS Stress Analysis.

20. GPUF

+ter No. E/FA 121, Dated 12 Dec.1981, to FPDC; transmittal of USNRC I

Alt.

-!ve Criteria.

21. Gr No. E/FA 091, Dated 16 Nov.1981, to FPDC; transmittal of Piping I

I I

I

/

I I

I I

I I

I I

t I

I l

51

I APPENDIX A PROGRAM: JTPIPE A FINITE ELEMENT PROGRAM FOR COMPUTING PIPING SYSTEM CRACK STABILITY PARAETERS I

I A-1

I CONTENTS A-1 INTRODUCTION A-2 APPROACH A-3 ANALYSIS AND IDEALIZATION OF THE STRUCTURE A-3.1 ELEMENT TO STRUCTURAL MATRICES A-3.2 BOUNDARY CONDITIONS A-3.3 COMPLI ANCE COMPUTATION AT CRACK SECTION A-4 PROGRAM ORGANIZATION A-4.1 N0DAL POINT AND ELEM DATA INPUT A-4.2 ASSEMBLAGE OF STIFFNESS MATRI)r A-4.3 COMPLIAN E CALCULATIONS A-4.4 COMPUTATION OF J,pp A-4.5 00MPUTATl0N OF T,pp I

A-2

A-1 INTRODUCTION I

in NUREG/CR-0838, Tada,et al., applied tearing modulus stabil Ity concepts to a sel ected nuclear reactor piping systm geometry and concluded that the piping was shown system was "f racture proof", that is, unstable ductile crack extension to be unlikely.

This was a major breakthrough for the inelastic f racture mechanics analysis of piping.

However, in Tada's analysis, the piping system was idealized as a straight beam with simple boundary conditions and the value of was spectfled.

In general, the geometry and the boundary conditions of a

J,pp nuclear piping system are complicated.

To extend the application of Tada's approach to actual piping systems, it became necessary that a finite el ement program be developed to overcome the structural complexitles of typical piping f or the case of Interest. The JTPlPE compute the value of J,pp systems and to program was developed for that purpose.

This Appendix summarizes the capabilItles of the current version of the JTPIPE computer program. The detailed theory and the numerical techniques used in JTPIPE are not presented in this Appendix.

The piping systens to be analyzed with JTPIPE can be modelled by combinations of four different types of finite elenents. The four elenent types are:

a) 3-d straight beam elenent b) 3-d curved beam elenent c)

Flexible connection elenent d)

Specic! elenent l

A-2 APPROACH The program determines the elastic compliance of the piping systen at specified locations for use in the crack stability analysis. The location of the i

maximum compliance is also Ident! fled. The computed compliance values are then i

A-3

used to dstermina principal stif fnssses at ecch location to be analyzed.

From the minimum stiffness at each

location, the L,ff/R is determined.

The L,ff/R data is stored for post-processing.

I Using the aforementioned L,ff/R data, J

and T,pp are computed p

using Equation (3-3) and (3-5) for each postulated crack location in another program. These latter values a r>3 tabulated for a series of circumferential through-wal l cracks having included angl es of 60 to 300 degrees in 60 degree -

Increments. Alternately, specific angles can be selected. All J vs.

T data is saved and later utilized for computer plotting the stability diagram where T,9 is also corresponding material resistance in the fwm of J,9 vs.

included.

I A-3 ANALYSIS AND IDEALIZATION OF THE STRUCTURE in this section, a brief description of the method of idealization of the structure is presented.

The direct stiffness method is used to analyze the structural systems.

A-3.1 Formulation si Structural Matrices A piping system is basically a three dimensional frame, it can be idealized as a number of discrete beam (straight or curved) elments, flexible connection elments and special elments. The beam elments are two node elments with six degrees of freedom at each node. The stif fness matrices of the elments are 12 x 12 symmetrical matrices which can be directly formulated from beam theory.

After the transf ormation from the local el ment coordinate system to the global coordinate system, the total systen stif f ness matrix can be formed by direct addition of the elanent matrices according to the index of the degree of freedom.

It can be expressed in the following manner.

I I

A-4

[K K

=

IJ g,,,

IJ is the stif f ness matrix component of the total system, K is the K,y where stIffnest, matrix component of the mth elenent and N Is the total number of elunents in the system.

The external force can be expressed in the f em I

F, K,y y

• U

=

where F,

is the external force applied at the Ith degree of freedom and U is y

the displacement at the jth degree of freedom.

A-3.2 Boundary conditions To simplify the programing problans associated with the specific displacements on the boundary, a spring, that is very stif f in comparison with the structure, is assumed to connect the boundary nodal point to a fixed point.

If the applied nodal displacement component is zero, the node will be restrained by

[

the stiff spring.

If a non-zero displacement component is specified, it can be I

l replaced by an equivalent force applied at that nodal point. The equivalent force is evaluated by the specified displacement applied on the stif f spring with the system structure stiffness ignored.

Since the spring is much stif fer than the structure, the error introduced is negligible.

A-3.3 cmollance Comoutation At cracked sectron in the stability analysis of a through-walI circumferential crack in a piping system, the rotational compliance at the pipe cracked section is required for the applied tearing modulus, T,pp.

This is because of the fact computation of the that the cracked section of the pipe is idealized as a plastic hinge. The rotational compliance at the pipe cracked section is due to the flextural rigidity A-5

of two clastic piping ssctions joinnd by tho hingsd section.

From the total systs stiffness, including the boundary conditions,as formulated in section A-3.1 and A-3.2, the rotational compilance at the pipe cracked section can be obtained by applying unit mments on opposite sides of the hinged section.

The principal rotational compliance at that section and the maximum rotational compliance of the selected locations in the piping syste sre i

both calculated.

1 A-4 PROGRAM ORGANIZATION The computation process in the JTPIPE program is basically divided into three distinct phases plus post-processing.

A-4.1 Nodal Point And Elment Data Input in this phase, the control information and nodal point geometry data are input and nodal points are generated by the program as required. The Indices of the degrees of freedom at each nodal point are established. The elenent data are input and element groups generated, the elenent connection arrays and the element coordinate transformation matrices are calculated and all element Information is stored in a disc file for use in the second and third phases.

A-4.2 Assembl age Di Systm Sti f f ness Matrix JTPIPE uses a compacted storage scheme in which the systs stiffness matrix is stored as a one-dimensional array.

In the second phase, the index of the I

storage is established, then the systan stif fness matrix is assembled and modified to satisfy the boundary conditions.

I A-6

A-4.3 Comollance Calculations In the third phase, the locations of the hinge points (post ul ated crack locations) desired for the compliance computation, are input. The rotational compliances and minimum stif fnesses at each cracked nodal point is calculated.

Next the L,ff/R are calculated and stored for post-processing.

A-4.4 Computation d J app This procedure is PROPRIETARY / SENSITIVE.

I A-4.5 cmnutati on d Tapp Finally, a post-processor is used to compute T,pp for specified crack sizes and crack rotations. The data is displayed in tabular form and is stored on a disk for subsequent post-processing;

namely, the generation of J vs.

T diagrams.

I I

I I

I A-7

I I

I APPENDIX B 4

MATERIAL PROPERTY DATA I

l B-1 TEARING RESISTANCE DATA As part of this project, en of fert was made to collect all the available J-R curve data for typical reactor piping materials. This Appendix includes data for j

TP304 and 316 stainless steels.

The J-R curve data that are compiled in this appendix were taken in thei r original form from the literature. Data f rom several specimen geometries were I

compared. The specimen geometries included compact specimens, center-cracked panels, 3-point bend bar and four-point bend pipe specimens. Both the room temperature and elevated tanperature data were included in the comparison.

The J-resistance curve data for stai nl ess steels were collected from I

References B-L through 11-13.

For computational convenience, alI data were curve l

fitted to a second order polynomial using a least squares technique.

All the i

i fitted curves so generated were then plotted on one figure to Investigate the i

!g scatter in the data. This is shown on Figure B-1.

The main source for scatter is lg l

apparently the specimen geometry and the fact that some specimens did not completely satisfy the J-cc '

-led growth and W requirements.

It is clearly seen t

l (Figure B-1) that the istance curves obtained from the pipe bend geometry give the lowest tearing resistancs I

lI s1

I Because of the wide scatter in the J-R curves shown in Figure B-1, it was decided to select the lowest curve for use in the stability assessment (J-T) diagram.

For this purpose, the resistance curve taken from Reference B-li was selected.

This curve is shown on Figure B-2.

The corresponding J-T curve is shown on Figure B-3 (T,9 values in this Figure were obtained from the fitted polynomial equation using material property data at 550F f rom Ref erence B-12.)

Table B-1 summarizes the J and T,9 data extracted from various sources for lc TP304 and 316 stainless steels.

A Iist of references indicating the original source of the data is included at the end of this Appendix.

Flow stress values are those reported in the Ref erence cited or assumed if not given.

J-R curves for TP304 stainless steel are felt to be quite simil ar to those for TP316. Thus, the use of a lower bound for TP316 piping that is based, for the most part, on TP304 data is valid.

I I

I I

I I

I I

e-2

I I

I I

I o

3 c

e n.-

b L

i 3

o E

o nE

_9 8

cu w

C

>C 9

CO ti 1

8' o

w 8

,T E o

x e

a:

,e n

aO e

m-L n

i

's s

z

=

=

l~

o a

n

-c h

g o

1 I

o o

c

>1

-d

+e I

~

o.

8 d d o

s o

4 4

I o

4 4%

5 e

L I

o 3

o 2

v a u.

,.~

00*08h 00'00h 00'02E 00*0h2 00*091 00'08 00 *f

,0l*

2**NI/97-NI

'I I

I I

B-3

I I

I t

i3

>=

o 2

n a

L i

E O

O m

i 5 E

-J u

So

>C G

O' CD D6 1,

y 1

L L

E laJ O

g i

e C

+

to n

0

.:I e==

O O

1 m-L L

0

> = =

$g' W

a m

v 3 E I

-o t

1 e

Q

~e I

m O

e 3

L 3

O I

tr I

7 N

t E

i

~.

O e

L o

I O

LA.

O 00*0h2 00*002 00*09'l 00*021 00*09 00*0h 00*f i

i v

l

,0 l a 2*aNI/97-NI

'I

I

'I

i

! I I

I 4

iiI 1

?!I I

O c

e O

N 4

! I n

O O O e

L 4.D O O

^

'l L

l M

g g O

s O

c) g g

w g

O LL E

O

_O mQ N

E kJ

>E O

CD L

3L w

i o

j n

u

_co

~

e

.I g g-O

!g m-D~

O e

O.

L i

O U

W O

4

_g i

0 1

E I

F-I oO I

.O ao 4-I OE 7

O 4

l O

M

_O I

I

=r CD e

i L

3 I

O

_U 00*0h'2 00*0d2 00*09'l 00*0d1 00*dB 00*dh 00*

l

,01" 2""NI/81-NI T

-I l

i I B-5 4

't

I I

I TABLE B-1 I

TEARING PROPERTIES FOR STAINLESS STEEL PIPING MATERI ALS Reference Material Specimen Thickness, Test O,ksi J

T Type in.

Temp,F g*

lbfT,**

n B-1 TP 304 CT 0.38 RT 66.00 4000 293 i

B-2 TP 304 CCT 0.38 RT 66.00 4000 550 B-3 TP 304 4 pt bend 0.34 RT 66.00 6200 1070***

B-4 TP 304 4 pt bend 0.34 RT 66.00 7400 1090***

i B-5 TP 304 4 pt bend 0.34 RT 66.00 8000 700***

B-6 TP 304 3 pt bend 0.34 550 45.95 6000 750 B-7

'IP 304 3 pt bend 1.00 550 45.60 6000 1000 I

B-8 TP 304 3 pt bend 0.34 RT 67.00 1000 280 B-9 TP 304 3 pt bend 1.00 RT 89.55 6500 1020 B-10 TP 304 CCT 0.20 RT 66.00 5500 87 i Eg B-11 TP 304 CCT 0.40 RT 66.00 7400 95 B-12 TP 304 CCT 0.40 RT 66.00 5300 185 B-13 TP 304 CCT 0.40 RT 66.00 8000 230 I

B-14 TP 304 CCT 0.32 RT 66.00 8000 205 B-15 TP 304 3 pt bend 0.34 RT 66.00 6600 160 B-16 TP 304 4 pt bend 0.35 RT 66.00 6000 100 B-17 TP 304 4 pt bend 0.35 RT 66.00 5800 104 I

B-18 TP 316 CT 1.00 600 40.00 5260 609 Properties at test temperature The units of J are Ib/in; 5710 lb/in = 1 MN/m lc

• • • Surf ace crack

I 4

I I

I

.I B-6

I I

I REFERENCES B-1 Landes, " Size and Geometry Effects on Elastic-Plastic Fracture Characterization," in CSNI Specialist Meeting on Plastic Tearing I nstabil ity, CSNI Rep. No.39, NUREG/CP-0010,p.215, CTS,B=.375 in.

B-2 Landes, Ibid., NUREG/CP-0010,p.215, CCT, B=.375 in.

B-3 Zahoor, et al., EPRI Project T118-2, 4PB pipe, Spec.10S B-4 Zahoor, et al., EPRI Project T118-2, 4PB pipe, Spec. 3S B-5 Zahoor, et al., EPRI Project T118-2, 4PB pipe, Spec. 6S B-6 Battel le data, EPRI Project T118-2, 3PB, B=.34 in., 2900 i

B-7 Battelle data, EPRI Project T118-2, 3PB, B=1.02 In., 2900 B-8 Battel le data, EPRI Project T118-2, 3PB, B=.34 in., 24C B-9 Battelle data, EPRI Project T118-2, 3PB, B=1.02 in.,24C B-10 Yagawa,et al, " Theoretical and Experimental Study en Unstable Fracture for Type 304 Stainless Steel Plates with a Soft Tensile Testing Machine, "NEUT 81-04, CCT, B=.09 In.

B-11 Yagawa,et al., Ibid. CCT, B=.18 in.

B-12 Yagawa,et al., Ibi d. CCT, B=.18 I n.

B-13 Yagawa,et al., Ibid. CCT, B=.13 in.

B-14 Zahocr,et al., ASE Paper 80-WA/PVP-3, EPRI T118-2, CCT, B = 0.32 in.

B-15 Zahocr,et al., ASE Paper 80-WA/PVP-3, EPRI T118-2, 3PB, B = 0.33 in.

B-16 Zahocr,et al., ASE Paper 80-WA/PVP-3, EPRI T118-2, PlPE, B = 0.35 in.

B-17 Zahocr,et al., ASE Paper 80-WA/PVP-3, EPRI T11E-2, PIPE, B = 0.35 in.

B-18 Page 204 of Ref.1, CTS, J test g

B-19 ASE, B&PV Code, Section li l, NA, App. A,1974 ed.

I I

B-7

_-u.

a--

-+-- - - - -- '

a-

.---.na3,.---

m-L-

!I a

lI f

APPENDIX C

'I 1!.

l ll il 4

1 1

I t

l l,l a

il

!I J

!I i

!I i

l

!I 4

!I l

c-1

USNRC ALTERNATIVE SAFETY ASSESSENT FOR SELECTED

+

HIGH ENERGY PlPE BREAK LOCATIONS AT SEP FACILITIES

• s This assessment is required only if a LWR high energy piping system (i.e., 275 psi or higher; or,200 R or higher, etc.) is being considered.

It to only required, I

If a postulated double ended pipe break would Impair saf e systs shutdown by pipe whip (lacking pipe whip constraints) consequences, or by the consequences of the implied leakage or its jet action.

The following guidance is for a safety assessment that may be permitted as an alternative to other system modifications I.

or alterations for locations where the mitigation of the consequences of high energy pipe break (or leakage) have been shown to be Impractical.

Guldance hr Alternate Saf ety Assese. ment The suggested guidance are as f ollows:

A.

Detec+abilItv Reaufrements Provide a leak detection systs to detect through-cracks of a length of twice the wall th ickness for minimum flow rates associated with normal (Level A)

ASE B&PV Code operating condition.

Both circumferential and longitudinal cracks must be considered '

~I critical break or leak locations. Methods for estimation of crack ope aas are attached in Appendix 2.

Surface roughness of the crack shoui.

.,a considered.

l B.

Integrity Reau f rements l

(1) Loads for Which Level D is Specified (a) Show that circumferential or longitudinal through-cracks of four walI th icknesses in length subjected to maximum Level D loading conditions do not exhibit substantial' monotonic loading crack growth (e.g.,

staying below J or K by plastic zone correctpd linear-elastic fracture mechanics, methods h a suitable alternative.

For' 4t flaws that are calculated to be greater than K or J consideration flaw growth arguments, fE) pos O,lation of small will be given to; (1) flaw sizes than 4t if justified by leak detection sensitivity. Also assure that local or general plastic Instability does not occur for these loading conditions and crack sizes.

(b) Under conditions in "B.(1)" show that the flow through the crack and the action of the jet through the crack will not impair saf e shutdown

, I of the system. Acceptable methodology for the estimationnnn of crack opening area for a circumferential through crack in a pipe in tension and bending and for longitudinal cracks subject to internal pressure are attached.

(2) Extreme Conditions to Preclude a Double-Ended Pipe Break I

C-2

~

I Using olastic-plcstic frccture-mechenics or suitcblo citernative show that circumferential through-cracks will romain stable for local f ully plastic I

large-deformation bending conditions under the following additional l l

conditions:

,I (a) Fully plastic bending of the cracked section Is to be assumed, unless other load limiting local conditions (such as elbow collapse) dictate j

maximum bending loads, for all critical locations.

(b) Assume all syst m anchors are effective.

To simplify the analysis, I

supports may conservatively be considered Inoperative.

If supports I

are included, consideration should be given to the adequacy of the support to resist large loads.

(c) Other as built displacement limits or constraints may be assurred as j

expecially justifled (such as displacoment Iimits of a plpe running is through a hole in a suf ficiently strong concrete wall or floor, etc.).

(d) Assume a through-crack size of 4t or 90 total circumferential length whichever is greater; or a Iarger crack only if especially justifled.

(e) Assume large deformations means deformations proceeding to as buil t displacement iImits or other especially justifled iImits.

(3) Material Properties Conservative material properties should be used in the analyses.

Sufficient justlitcation must be provided for the properties, both I

weldment and base metal, used in the analyses.

C.

suberttical Crack Develcoment Consideration should be given to the types of subcritical cracks which may be developed at all locations associated with this type of analysis.

From prior I

experience and/or direct analysis it should be shown that:

(1) There is a positive tendency to develop through-walI cracks.

(2) If there is a tendency to develop long surface cracks in addition to through-walI cracks, then it should be f urther demonstrated that the long surf ace crack w11I romain suf fIclently shalIow.

D.

Aunmented Inservice insoection Piping syste locations for which corrective measures are not practicabl e should be Inspected volumetrically in accordance with ASE Code,Section XI for a Class I system regardless of actual system classification.

• USNRC Letter to Consumer's Power dated 12/12/81.

I I

C-3

_ _ _ _ _ _ _ _