ML17256A648

Rev 0 to Fracture Mechanics Evaluation of High Energy Piping Lines at Re Ginna Nuclear Power Plant.

ML17256A648
Person / Time
Site:	Ginna
Issue date:	04/08/1983
From:	Copeland J, Wu Y NUTECH ENGINEERS, INC.
To:
Shared Package
ML17256A647	List: ML17256A646 ML17256A648
References
TASK-03-05.A, TASK-3-5.A, TASK-RR RGE-02-004-R00, RGE-2-4-R, NUDOCS 8304290006
Download: ML17256A648 (59)

Text

RGE-02-004 Revision 0 April 8, 1983 100.2602.0200 FRACTURE MECHANICS EVALUATION OF HIGH ENERGY PIPING LINES AT THE R.E. GINNA NUCLEAR POWER PLANT Prepared for:

Rochester Gas & Electric Corporation Prepared by:

NUTECH ENGINEERS San Jose, California Prepared by: Reviewed and Approved by:

ZF. g/

Dr. J. F. Copeland Dr. LE CD Hsu Project Engineer Applied Mechanics

'anager, Issued by:

Dr. Y. S. Wu D. K. McWilliams Consultant I Project Manager Date:

nutech

>

,

", REmjATORY OOCKET FILE MPV

REVISION CONTROL SHEET Fracture -Mechanics Evaluation RgpoRT NUyBgR RGE-02-004 of High Energy Piping Lines at Revision 0 the R.E. Ginna Nuclear Power'lant NAME/ TITLE INITIALS Dr. L.C. Hsu En ineerin Mana er NAME/ TITLE INITIALS Dr. Y.S. Wu/Consultant I YSW NAMEI TITLE INITIALS NAME I TITLE INITIALS PREPARED ACCURACY CRITERIA REMARKS PAGEIS) REV BY /OATE CHECK BY / OATE CHECK BY / OATE i

t,hru 0 Tf-~  %+6 )SW@~-F3-v l

thru C P"S'-S3 Pstk'ag-Q 53 QEP 34.1.1 3.3.

nutech

TABLE OF CONTENTS Pacae

1.0 INTRODUCTION

/EXECUTIVE

SUMMARY

1.1 Background 1 1.2 Objectives and Technical Approaches for Ginna '

1.3 Conclusions and Recommendations 5 2.0 FRACTURE MECHANICS LEAK-BEFORE-BREAK ANALYSIS 2.1 Critical Crack Sizes for Instability 2.1.1 J-Integral Analysis and Tearing Modulus ll 2.1.2 Net Section Plastic Collapse Criterion 16 2.2 Leak Rates 19 2.2.1 Accumulator Line 20 2.2.2 Pressurizer Surge Line 22 2.2.3 Ginna Leak Detection Capabilities 23 2.3 Subcritical Crack Growth Rates 24 2.3.1 Stress Profiles 25 2.3.2 Cycling Rate 26 2.3.3 Crack Growth Analysis 27

3.0 CONCLUSION

S AND RECOMMENDATIONS 29 4 ' REFERENCES 52 RGE-02-004 Revision 0 nutech

I e-LIST OF TABLES Number Title Pacae 2-1 Parameters for Leak-Before-Break Analysis of 31 Pressurizer Surge and Accumulator Lines 2-2 Level D ASME Code Maximum Allowable Stresses 32 Used for Analysis of Crack Instability for Pressurizer Surge (PSL) and Accumulator Lines (AL) 2-3 Applied J-Integral and Tearing Modulus Values 33 as Functions of Through-Wall Half Crack Length, a, For the Ginna Accumulator Line With An Applied Stress of 30,133 psi 2-4 Applied J-Integral and Tearing Modulus Values 34 as Functions of Applied Stress for the Ginna Accumulator Line with a Through-Wall Half Crack Length of 3.5 in.

2-5 Applied J-Integral and Tearing Modulus Values 35 as Functions of Through-Wall Half Crack Length, a, for the Ginna Pressurizer Surge Line with an Applied Stress of 37,600 psi 2-6 Failure Crack Sizes for Postulated Compound 36 Crack in Ginna Accumulator Line, Based on Net Section Plastic Collapse Criterion 2-7 Failure Crack Sizes for Postulated Compound 37 Crack in Ginna Pressurizer Surge Line, Based on Net Section Plastic Collapse Criterion 2-8 Leak Rate Results for Circumferential Through- 38 Wall Cracks (CTWC) and Longitudinal Through-Wall Cracks (LTWC) in the Pressurizer Surge and Accumulator Lines (PSL and AL) for Normal Operation Pressure Stresses 2-9 Transients Considered in Subcritical Crack '39 Growth Rate Analyses for Pressurizer Surge and Accumulator Lines (Reference 8)

RGE-02-004 iv Revision 0 nutech IINOINQIIRB

LIST OF FIGURES Ficiure Title Pacae 2-1 Pressurizer Surge Line 40 2-2 Safety Injection From Accumulator A 41 2-3 Representation of Postulated Cracks in Pipes 42 for Fracture Mechanics Leak-Before-Break Analysis 2-4 J-Integral Resistance Curves for Austenitic 43 Stainless Steel (Reference 7) 2-5 J-Integral/Tearing Modulus Stability Diagram 44 for Ginna Accumulator Line with Through-Wall Cracks 2-6 J-Integral/Tearing Modulus Stability Diagram for Ginna Pressurizer Surge Line with Through-Wall Cracks 2-7 Failure Analysis Diagram for Postulated Com- 46 pound Crack in Ginna Accumulator Line Line, Based on Net Section Plastic Collapse Criterion 2-8 Failure Analysis Diagram for Postulated Com- 47 pound Crack in Ginna Pressurizer Surge Line, Based on Net Section Plastic Collapse Criterion 2-9 Diagram for Maximum Steam/Water Flow Rate to 48 Determine Flow Rate for Saturated Liquid in the Pressurizer Surge Line (Moody Model Reference 15) 2-10 Stress Profiles for Bounding Case (Reference 3) 49 for Subcritical Crack Growth Predictions 2-11 Cyclic Loading Conditions Assumed for Conserva- 51 tive Subcritical Crack Growth Rate Analysis of Pressurizer Surge and Accumulator Lines 2-12 Predicted Subcritical Crack Growth Rates for 51 Circumferential and Longitudinal Cracks with Assumed Initial Depths (ai) of 0.02 inches and 0.10 inches for the Pressurizer Surge and Accumulator Lines RGE-02-004 Revision 0 nutech

1.0 INTRODUCTION

/EXECUTIVE

SUMMARY

High energy line break (HELB) analyses were completed for the resolution of open items for the NRC Systematic Evaluation Program (SEP) Topic III-5.A for the R.E.

Ginna Nuclear Power Plant. This report addresses leak-before-break fracture mechanics evaluations of the Ginna pressurizer surge and accumulator piping lines.

Back round The SEP was initiated by the NRC to review the designs of older operating nuclear reactor plants to reconfirm and document their safety. The review compared the as-built plant design with current criteria in 137 different areas defined as "topics" (Reference 1). Many of these topics'et current criteria or were acceptable on another defined basis for Ginna. The objective of this study is the resolution of SEP Topic III-5.A for Ginna, as defined in Reference 1.

Appendix A of 10CFR Part 50 reguires that structures, systems and components important to safety (Engineered Safety Features, ESFs) be appropriately protected against the dynamic effects of postulated pipe breaks.

The goal is to protect these ESFs so the plant can be RGE-02-004 Revision 0 nutech GNOINQGRG

shut down and maintained in a safe shutdown condition in the event of a postulated rupture of a piping system containing high energy fluid.

Current designs protect ESFs against the consequences of high energy line breaks (HELBs) through the use of pipe whip restraints, jet impingement shields, physical separation and other methods. However, plants designed before the existence of current requirements generally do not have the full complement of such features.

Furthermore, in many cases modifications to incorporate these features may be impractical due to physical plant configurations or other considerations. Therefore, the NRC has given guidance on other acceptable methods for the resolution of Systematic Evaluation Program (SEP)

Topic III-5.A, for High Energy Line Breaks Inside Containment.

1.2 Ob'ectives

and Technical A roaches for Ginna In Reference 1, the NRC advises that breaks in the accumulator line or pressurizer surge line could adversely affect nearby safety-related equipment.

Additionally, guidance for performing fracture mechanics leak-before-break evaluations to resolve this issue was forwarded to Rochester Gas and Electric Corporation RGE-02-004 Revision 0 nutech

(RGE) by the NRC (Reference 2). This approach was employed for these piping lines. It is based on a combination of inservice inspection (ISI) and leak detection, to detect the presence of cracks, and of fracture mechanics analysis to assure that crack instability will not occur for cracks smaller than those detectable by these methods. These detection methods complement each other, since ISI is especially suited to finding long cracks, and leak monitors detect short, through-thickness cracks. Reference 2 provides the methodology to compute crack opening areas for determining leak rates for comparison with detection limits. The ISI involves volumetric inspection in accordance with ASME Section XI for a Class 1 system, regardless of actual system classification. The goal is to detect and limit any service induced flaws to allowable sizes prescribed by the ASME Code,Section XI (crack depth limited to less than approximately 10% of pipe wall thickness). Fracture mechanics subcritical crack growth analyses are employed to assure that this goal for limiting crack growth is met. These limits on crack size imposed by leak monitors and ISI are compared to the critical crack sizes predicted for instability and pipe rupture, computed in accordance with Reference

2. Adequate margin between crack detection and the crack size for rupture must exist. In this way, crack RGE-02-004 Revision 0 nutech

detection and corrective actions will precede any chance for HELBs and subsequent postulated effects on ESFs.

In accordance with the latest NRC guidance (Reference 2), the leak-before-break technique was evaluated for the Ginna pressurizer surge and accumulator lines. The elements of this evaluation include the definition of the following:

a) Largest crack size which will remain stable b) Leak rate resulting from a crack of length 2t (twice the pipe wall thickness) c) Size of crack which will leak at a rate greater than l gpm, if b) results in less than l gpm.

d) Analysis of part-through-thickness cracks for subcritical crack growth rates to establish ISI intervals.

Very conservative analyses were performed to predict the largest stable crack sizes, by using worst case stresses (References 3 and 4), as well as ASME Code maximum allowable stresses. In this way, sets of analyses were performed to envelope all locations in each line, as RGE-02-004 Revision 0 nutech GNOINGGAU

well as to compensate for potential future load increases. The subcritical crack growth rate analyses were also enveloped in a similar manner, by using a conservative load cycling spectrum based on Ginna design transients and the most severe transient loads in the stress reports (References 3 and 4).

1.3 Conclusions Conclusions resulting from the preceding analyses are reflected in terms of critical through-wall crack lengths for instability and crack lengths for 1 gpm leak rates in the following table:

N G INNA STRESS REPORT WORST CASE STRESSES Crack Len th for Instabilit (in.)

Crack Crack Length Orien- Net Section Tearing for 1 gpm Line tation Colla se Modulus Leak Rate Pressurizer circ ~ 7.8 >20 2.6 Surge long ~ >12 <2 Accumulator ClrC ~ 10 22 3.5 long. 16 2.66 Both Net Section Plastic Collapse and Tearing Modulus approaches are used to predict critical crack lengths for instability, based on worst-case stresses. The worst case (Ginna stress report) stresses used for the RGE-02-004 Revision 0 nutechIINOINCGAS

'

~

pressurizer surge line analyses are Pm + Pb = 34,747 psi. The corresponding stresses for the accumulator line are Pm + PI = 30,133 psi. Only normal operating pressure stresses were used to compute leak rates.

Thus, this analysis is considered to be conservative.

RGSE (References 5 and 6) has current bulk leak detection capabilities to detect 1 gpm leak rates for these lines in at least 6.4 hr. Since a margin of at least a factor of 2 exists between the crack lengths for a 1 gpm leak and the "worst actual stress" lengths for instability (consistent with Reference 2 guidance),

these current leak detection systems are considered adequate. Furthermore, subcritical crack growth rate analyses show that inservice inspection intervals of 10 years are appropriate to detect part-through-thickness cracks before they approach instability.

RGE-02-004 Revision 0 nutech

2.0 FRACTURE MECHANICS LEAK-BEFORE-BREAK ANALYSIS Postulated breaks in the accumulator line or pressurizer surge line could adversely affect nearby safety-related equipment. These lines at the Ginna Plant are shown in Figures 2-1 and 2-2, and the piping system parameters are given in Table 2-1. Additionally, guidance for performing fracture mechanics leak-before-break evaluations to resolve this issue was forwarded to Rochester Gas and Electric Corporation (RGE) by the NRC (Reference 2). This approach was employed for these piping lines. It is based on a combination of inservice inspection (ISI) and leak detection, to detect the presence of cracks, and of fracture mechanics analyses to assure that pipe rupture will not occur for cracks smaller than those detectable by these methods. These detection methods complement each other, since ISI is especially suited to finding long cracks, and leak monitors detect short, through-thickness cracks. These types of cracks are represented in Figure 2-3.

In accordance with the latest NRC guidance (Reference 2), the leak-before-break technique was evaluated for the Ginna pressurizer surge and accumulator lines. The elements of this evaluation include the definition of the following:

RGE-02-004 Revision 0 nutechGNOINCRRU

a) Largest crack size which will remain stable; b) Leak rate resulting from a crack of length 2t (twice the pipe wall thickness);

c) Size of crack which will leak at a rate greater than lgpm, if b) results in less than 1 gpm; d) Analysis of part-through-thickness cracks for sub-critical crack growth rates to establish ISI intervals.

Very conservative analyses were performed to predict the largest stable crack sizes, by using ASME Code maximum allowable stresses, and also, by using the maximum stresses in the piping stress reports (References 3 and 4). In this way, sets of analyses were performed to envelope all locations in each line, as well as to compensate for potential future load increases. The subcritical crack growth rate analyses were also enveloped in a similar manner, by using a conservative load cycling spectrum based on Ginna design transients (Reference 8) and the most severe transient loads in the stress reports (References 3 and 4).

RGE-02-004 Revision 0

2.1 Critical Crack Sizes for Instabilit Three methods of analysis to predict critical crack sizes for the Ginna accumulator line and pressurizer surge line were considered according to the guidance given by the NRC in Reference 2. These methods are:

a) linear elastic fracture mechanics; b) J-Integral and Tearing Modulus approaches (Reference 9); and II c) the net section plastic collapse criterion (References 10 and 11).

As seen in Table 2-1, these piping materials are Type 316 austenitic stainless steel, which has a very high level of toughness. Reference 7 reports a critical J-specimen.

Integral value, JIc, for fracture initiation of Type 316 at 600'F, of 5260 in-lb in 2 in a one inch thick test Other results for Type 304 stainless steel (a similar material) are shown in Figure 2-4 and are also at a high toughness level. Figure 2-4 shows crack extension as a function of applied J-Integral loading.

In some cases, JIc values can be simply converted to KIc (linear elastic fracture toughness) and used for a RGE-02-004 Revision 0 nuteclh

conservative analysis of fracture resistance in a component. However, this conversion is suspect for materials which do not meet the following validity criterion (Reference 12):

B > 25 Ic y

where, B = Specimen thickness (in.)

= Critical J-Integral Value (

in-lb )

JZc in 2 e = Material yield strength (psi)

By solving the preceding equation for B, using JIc =

5,260 2 and a = 30,000 psi (the room temperature yield strength), it can be seen that B must be greater than 4.4 inches for a valid conversion to linear elastic fracture mechanics parameters (KIc) and analysis. At higher temperature, as the yield strength decreases, the thickness requirement becomes even greater. Thus, for this combination of high material toughness and low yield strength, the approach of linear elastic fracture I

mechanics is not considered valid. Other approaches, based on elastic-plastic fracture mechanics are required. The approaches of J-Integral, Tearing Modulus, and net section plastic collapse criterion are used in this program, as detailed in the following subsections.

RGE-0 2-0 0 4 10 Revision 0 nute@4

2.1.1 J-Integral and Tearing Modulus Analyses These analyses follow the methodology of Reference 9 for the stainless steel accumulator and pressurizer surge lines at Ginna. In accordance with Reference 2, Level D stresses were used for this analysis. ASME Code allowable stresses for these lines are given in Table 2-2. For the case of these elastic-plastic crack stability analyses, only the primary stresses are considered, because of the relatively large deformations accompanying fracture, which would relieve any secondary stresses. This is consistent with Reference 10, where it is recognized that secondary and peak stresses have no effect on the limit load because they are produced by the action of imposed strains or are locally conf'ined and self-limiting.

2.1.1.1 Stresses The Code maximum allowable primary stresses (membrane plus bending) from Table 2-2 are equal to Shy For the accumulator line, this gives Pm + Pb = 37,600 psi. For the pressurizer surge line, Pm + Pb = 56,400 psi.

RGE-02-004 Revision 0 nuts cd KNOINIKAS

Conservative analyses were also done for the most severe stresses at the worst location in each line, according to the stress reports (References 3 and 4). These stresses also include thermal stresses and assume all events occurring simultaneously for conservatism. For the accumulator line, the worst case Pm + Pb = 30,133 psi at node 8,400 (Reference 4), consisting of the sum of dead weight, RHR malfunction, loss of load, and seismic (SSE) stresses, along with the primary membrane pressure stress@ Pm 5gl60 psi. For the pressurizer surge line, the worst case Pm + Pb = 34,747 psi at node 690 (Reference 3), consisting of the sum of dead weight, control rod ejection and seismic (SSE) stresses, along with the primary membrane stress of 5,919 psi.

2.1.1.2 Accumulator Line Both circumferential and longitudinal through-wall I

cracks, as shown in Figure 2-3 a) were evaluated.

From Reference 9, the J-integral (J) and Tearing Modulus (T) are calculated from:

2 0 a (Stress) [ 2]

E Factor and T = (Stress)

Factor

[Y2 + 2]

RGE-02-004 12 Revision 0 nuts ch IKNOINGIXRG

where, Flow stress Crack half-length E Elastic modulus

() = Stress factor (Reference 9)

[] = Geometry factor, defined by geometry parameters X, Y, and Y'Reference 9)

From the ASME Code Section III Appendices, g is com-puted as the average of Sy and Su (minimum expected yield and tensile strengths) to be 52.5 ksi at 100'F and 48 .8 ksi at 300'F. A value of 50 ksi is, thus, used for this analysis. Similarly, E = 28.3 X 10 psi, at 100'F and 27 X 10 psi, at 300'F. Thus, a value of 27.5 X 10 psi was selected for this analysis.

The parameter X is used in Reference 9 to determine the geometry factors. It is defined as:

a (Rt> /

where, R = Pipe radius, 5 inches t = Pipe wall thickness, 1 inch RGE-02-004 13 Revision 0 nuteelhj

Values of J and T for a stress of 30,133 psi and varying crack lengths were computed, as shown in Table 2-3, for circumferential and longitudinal cracks. Another case was computed for a flaw 7 inches long with varying stresses, in Table 2-4. The 7 inch flaw size was selected to be a factor of two greater than the largest flaw which will give a predicted leak rate of 1 gpm (Section 2.2 of this report). The factor of two is consistent with that given in the Reference 2 guidance for the margin between a leaking and an unstable crack.

This was done to define the maximum stresses at which the cracks are predicted to remain stable.

The results of the analyses in Tables 2-3 and 2-4 are plotted in Figure 2-5 for comparison with a represen-tative material J/T curve (Reference 7). Applied J and T values to the left of the material curve are con-sidered to result in stable crack behavior (Reference 9). A ceiling of applied J = 24,000 in-lb/in. was also placed on this analysis, since this is the highest value represented by the material resistance curve in Figure 2-5. It should be noted that the analyses for longitu-dinal cracks are especially conservative, since stresses other than pressure stresses are not as likely to affect this orientation of flaw; yet they are included in this evaluation.

RGE-02-004 14 Revision 0

Conclusions from Tables 2-3 and 2-4 and Figure 2-5 are that large values of stresses and crack sizes are tolerated by these pipes before instability is predicted. In some cases, the validity of this analysis is exceeded before crack instability is predicted. The largest through-wall crack sizes, evaluated for, the worst case stress of 30,133 psi., which remain stable are a 22 inch long circumferential crack and a 16 inch long longitudinal crack. Stresses of at least 55,000 psi, and 50,000 psi, are tolerated for 7 inch long through-wall circumferential and longitudinal cracks, respectively. These stresses are just below the maximum ASME Code allowable primary stress of 56,400 psi, but are well above the worst case stress of 30,133 psi.

2.1.1.3 Pressurizer Surge Line Analyses similar to those done for the accumulator line were also performed for the Ginna pressurizer surge line. Applied J and T values were computed for a stress of 37,600 psi, the maximum ASME Code allowable primary stress which is greater than the worst case stress of 34,747 psi (Reference 3), for varying crack sizes (Table 2-5). Since higher primary stresses are not permitted by the ASME Code, the calculations for higher stresses RGE-02-004 15 Revision 0 nutec4

were not performed in this case. Values of g = 45 ksi and E = 25 X 10 psi are used. The results of these analyses are plotted on the stability diagram in Figure 2-6, as was done for the accumulator line.

Conclusions from Figure 2-6 are that through-wall cracks longer than 20 inches and 12 inches for circumferential and longitudinal orientations, respectively, are stable for the applied stress of 37,600 psi.

2.1.2 Net Section Plastic Collapse Criterion The net section collapse criterion (NSCC) follows Reference 10, with several minor changes in the equations used. Specifically, this criterion assumes that failure is defined by plastic instability which occurs when the stress in the net section at the crack reaches the material flow stress. This approach is con-servative for austenitic stainless steel, since the flow stress is taken as the average between the minimum expected yield and tensile strengths. In reality, strain hardening of this material would continue beyond this flow stress and would give increased resistance to collapse, which is not taken into account in this analysis.

RGE-02-004 16 Revision 0 nutech QNQINQtERG

When a conservative compound crack (a combination of through-wall and part-through wall cracks, as seen in Figures 2-7 and 2-8) is assumed, the analysis becomes slightly more complex because of shifting the pipe neutral axis by the angle g(Figures 2-7 and 2-8).

This effect is included in the following equations to compute critical compound crack sizes for instability:

( m- v) (1--)d t (

"Pm a

)

(1- d 2

t) pb = t 2

m (1-) [2 sin g-sin v]

where:

Shift of the neutral axis Half-crack angle for through-wall crack Depth of part-through wall crack Pipe wall thickness a Flow stress Pm Primary membrane stress Pb Primary bending stress These parameters are defined further in Figures 2-7 and 2-8 RGE-02-004 17 Revision 0 nutech GNOINCGRQ

Using the same Pm and Pb stresses discussed in sub-section 2.1.1.1 of this report, the preceding equations were solved simultaneously to produce the failure diagrams in Figures 2-7 and 2-8. Numerical values of the critical crack sizes thus computed are also given in Tables 2-6 and 2-7 for the accumulator and pressurizer surge lines at Ginna.

I Conclusions for the accumulator line are that circum-ferential through-wall cracks of 0.318 and 0.116 fractions of the circumference are stable for the worst case stress of 30,133 psi and the maximum ASME Code allowable primary stress of 56,400 psi, respectively.

These are cracks 10 and 3.6 inches long, which are more conservative than the Tearing Modulus results of subsection 2.1.1. Such conservatism has also been shown in Reference 7, where net section collapse was analyzed by the J/T approach. Thus, the net section collaspe criterion is truly a conservative estimate of austenitic stainless steel flaw tolerance.

Part-through wall cracks equal to 59.61% and 21.83% of the wall thickness were defined for the onset of instability for the accumulator line with stresses of 30,133 psi and 56,400 psi, respectively.

RGE-02-004 18 Revision 0 nutech

Conclusions based on the net section collaspe criterion for the Ginna pressurizer surge line are that circumferential through-wall flaws equal 'to 0.248 and 0.223 fractions of the circumference are stable for the worst case stress of 34,747 psi and the maximum ASME Code allowable primary stress of 37,600 psi. These are cracks greater than 7 inches long.

Part-through wall cracks equal to 47.5% and 43.5% of the wall thickness were defined for the onset of instability for the pressurizer surge line with stresses of 34,747 psi and 37,600 psi, respectively.

2.2 Leak Rates In accordance with the guidance of Reference 2, crack opening areas and leak rates were computed for through-wall cracks 2 inches (2 t) long. Crack sizes required to give leak rates of 1 gpm were also calculated for the Ginna accumulator and pressurizer surge lines. The leak rates are quite conservative since only pressure stresses are considered. In reality, other stresses will also tend to open the cracks for leakage, especially for circumferential flaws.

RGE-02-004 19 Revision 0 nutech

2.2.1 Accumulator Line Crack opening areas were computed for circumferential through wall cracks from the following equations (Reference 2):

Ap

~ E (2mRt) G P

( X)

(Rt) ~

Gp(Z) = X + 0.16K g (0< X < 1)

G ( X) = 0.02 + 0.81 X + 0 ~ 3X + 0.03Z i (l<)<5) where, Ap = crack opening area a = pressure stress, in axial direction p

1/2 ~,

t for internal pressure, p The minimum normal operating pressure of 750 psi (Table 2-1) was employed in this calculation. Crack opening area results are given in Table 2-8.

RGE-02-004 20 Revision 0 nutech RNOINGGRS

For a longitudinal through-wall crack the following equations are used to compute Ap (Reference 2):

A P

= ~a E

(2mRt) G p

(),)

Gp( X) + 0.625),, (0< ><1)

G ( g) = 0.14 + 0.36 g + 0.72 Z + 0.405 X, (1< X<5) where, a is the hoop pressure stress and the other terms p

are as defined for the circumferential through-wall crack. The leak flow rates for the accumulator line are computed for the non-saturated liquid at 120'F, using the Bernoulli equation (Reference 13):

2g ap 1/2 G = p[ )

P where, G = Flow rate, (ibm/(sec. in. 2 ))

Density at 120'F = 61.71 ibm./1728 in.

gc = (32.2) (12) in./sec.

Pressure difference, (750 14.7) psi.

The results, in gpm, are shown in Table 2-8. For the AL, leak rates of 0.236 gpm and 0.516 gpm result from pressure stresses for 2t long circumferential and longitudinal through-wall cracks. Circumferential and RGE-02-004 21 Revision 0 nutech GNOWCRRQ

longitudinal through-wall crack half-lengths of 1.752 and 1.328 inches are required for leak rates of 1 gpm.

All these results in Table 2-8 include a leak friction coefficient factor of 0.6 (Reference 14).

2.2.2 Pressurizer Surge Line A similar analysis to that for the accumulator line for Ap values was done for the pressurizer surge line.

Results are given in Table 2-8. For the saturated liquid the flow rate, G, is obtained from Figure 2-9 (Reference 15), for the pressure of 2235 psi and the temperature of 612.2'F (Table 2-1). Again, a flow friction coefficient of 0.6 is employed (Reference 14).

The results, in gpm, are shown in Table 2-8. For the pressurizer surge line, leak rates of 0.567 gpm and 1.245 gpm result from pressure stresses for 2t long circumferential and longitudinal through-wall cracks. A circumferential through-wall half-crack length of 1.314 in. is required for a leak rate of 1 gpm.

RGE-02-004 22 Revision 0 nutech

2.2.3 Ginna Leak Detection Capabilities For primary coolant leak detection (the pressurizer surge line), RGaE has provided the NRC documentation (Reference 5) supporting a 1 gpm leak detection in 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. Discussion with RGGE (Reference 6) gave further details regarding these leak detection capabilities.

The methods consist of 1) an airborne particulate radio-activity monitor, which can ideally detect 0.013 gpm within 20 minutes, 2) a monitor of condensate flow rate from the air cooler, which can detect 1 gpm within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and, 3) a chemical volume control system (CVCS) monitor, which can detect 0.25 gpm within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.

I RG&E has two systems of leak detection for the accumu-lator line (Reference 6). Level detectors consist of high and low level alarms set for 1108 ft and 1134 ft The difference is 26 ft or 194 gal This results

~ ~

in a time interval of 3.23 hr to detect a leak of 1 gpm for the worst-case, where the initial level is just below the high level alarm. There also is a sump pump

("A" pump) which is activated from a level alarm 30.5 in. from the floor of the 4.5 ft x 4.5 ft sump area.

This gives a required volume of fluid of about 51 ft or 385 gal. This results in a time interval of 6.4 hr. to detect a leak of 1 gpm.

RGE-02-004 23 Revision 0 nutech

Prom the preceding leak detection systems, it is apparent that RG&E currently has the capability to detect a 1 gpm leak at Ginna for the pressurizer surge and accumulator lines. As seen in Table 2-8, those crack lengths corresponding to a l gpm leak rate give significant margins against crack instability, when actual worst-case stresses are used to predict instability. The smallest margin is a factor of 2.86 on crack length for a circumferential crack in the accumulator line. This margin is above the factor of 2 given by the NRC guidance (Reference 2). Thus, the existing RGSE leak detection capabilities appear adequate to support this leak-before-break approach.

2 ~3 Subcritical Crack Growth Rates The preceding fracture mechanics analyses and leak rate analyses provide information for protecting against HELB by the leak-before-break approach. However, such information must also be generated to protect against pipe rupture resulting from the growth of long part-through wall, non-leaking cracks. This is accomplished by detecting and preventing such cracks with augmented inservice inspection (ISI). Adequate ISI intervals to prevent postulated part-through wall cracks from becom-RGE-02-004 24 Revision 0 nutech

ing critical (unstable) are established 'by predicting subcritical crack growth rates. Such analyses are presented in the following subsections, for the Ginna pressurizer surge and accumulator lines.

2.3.1 Stress Profiles Loading conditions for subcritical fatigue crack growth analyses were defined by considering the bounding case in the stress reports (References 3 and 4). The bounding case (most severe stresses) is at node 690 for the pressurizer surge line (Reference 3). These stress profiles were used to envelope both the pressurizer surge line and accumulator line for Ginna.

The stress profiles through the pipe wall thickness are shown in Figure 2-10 for both circumferential and longitudinal cracks. In fatigue analysis, the cycling between minimum and maximum loads is considered. The minimum load condition is for plant shutdown, where loading consists of piping dead-weight and weld residual stresses for circumferential flaws, and is assumed as zero for long longitudinal flaws. The maximum load condition, for all cycles is conservatively assumed to be the case for pressurizer surge line stresses with control rod ejection (Reference 3). For circumferential RGE-02-004 25 Revision 0 nutech ENIRINGliRQ

cracks, the maximum load condition consists of dead-weight, weld residual stresses, pressure (3015 psi stresses, and thermal stresses. For longitudinal cracks, the maximum load condition is comprised of pressure loading. Again, the stress profiles associated with these loadings are shown in Figure 2-10. Welding residual stresses are included in the NUTECH crack growth computer model, NUTCRAK (Reference 16).

2.3.2 Cycling Rate The transients considered to develop the frequency of cycling (between the preceding maximum/minimum loads) are shown in Table 2-9 (Reference 8). These transients were only used to estimate the number of cycles expected during the plant life, since the assumed loadings are more severe than those associated with these transients.

Only transients with significant loads or which were associated with the subject piping lines were con-sidered, as shown in Table 2-9. This resulted in a total of approximately 1200 significant cycles in a 40 year plant design life, or 30 cycles per year. Thus, the load cycling spectrum assumed for the subcritical crack growth analyses are shown in Figure 2-11.

RGE-02-004 26 Revision 0 nutech

2.3.3 Crack Growth Rate Analysis The preceding information was input to the following crack growth law (Reference 17):

'da dN

( g() n where, da d

= Crack growth rate (in. /cycle)

C. = Material constant = 2.74 x 10 n = Material constant = 3.97 bK = Range in applied stress intensity factor for each cycle.

The equation was solved by the NUTCRAK program (Reference 16) for crack depth as a function of number of cycles. Stress intensity factors for circumferential and longitudinal cracks are a part of the calculation output, and include the effects of maximum-to-minimum loading ratios.

The results of this analysis are shown in Figure 2-12.

Two initial crack sizes (depths) were assumed for part-I through wall cracks of infinite length: ai = 0.02 inch RGE-02-004 27 Revision 0 nutech

and ai = 0.10 inch. If it is assumed that a crack of depth 0.02 inch can be detected by ISI, and this is used as the initial flaw size, then insignificant flaw growth occurs over 40 years (1200 cycles).

As an extreme case, it is assumed that.a flaw 0.10 inch deep (10% of the wall) is the limit of ISI detect-ability, and is assumed as the initial crack size. Even for this large crack size, longitudinal crack growth is insignificant. However, circumferential crack extension does occur, growing from 10% of the pipe wall thickness to 20% in about 10 years. This is still well below the crack size of almost 50% of wall thickness (d/t in Tables 2-7 and 2-8 for v/m = 0) predicted for instability on the bases of worst case stresses and the net section plastic collapse criterion. Thus, it is recommended that even for this extreme case, an ISI interval of 10 years should be adequate to detect part-through wall cracks prior to approaching pipe rupture.

RGE-02-004 28 Revision 0 nutech GNOINCERG

3.0 CONCLUSION

S AND RECOMMENDATIONS The leak-before-break approach for resolution of HELB for the Ginna pressurizer surge and accumulator lines is shown to be feasible and practical.

Critical crack sizes for rupture of the pipes were predicted conservatively by Tearing Modulus and net section plastic collapse criterion approaches. The applied loads were based on the worst case stresses, using the stress reports. Through-wall circumferential cracks 24.7% of the pipe circumference (7.75 inches) are shown to be stable for these loads, based on the net section plastic collapse criterion. Longitudinal through-wall cracks 12 inches long are shown to be stable for these loads, using the Tearing Modulus approach. Circumferential part-through wall cracks equal to almost 50% of the pipe wall thickness are shown to be stable by the net section collapse criterion.

Thus, an ample margin of fracture resistance exists in these pipes. The analyses based on maximum ASME Code allowable stresses still predict significant resistance to fracture, but with less margin for leak detection and ISI.

RGE-02-004 29 Revision 0 nutech

Leak rates, based on internal operating pressure stresses only, were computed for through wall flaws of length 2t (2 inches). Leak rates range from 0.236 gpm to 1.245 gpm for circumferential and longitudinal through-wall cracks in both lines. The worst case crack length to give a minimum leak rate of 1 gpm, is 3.5 inches for a circumferential crack in the accumulator line. This still provides margin against reaching the crack length required for instability (10 inches, Table 2-6) for the accumulator line with worst case stress report stresses. RGGE currently has bulk leak detection systems at Ginna capable of detecting 1 gpm leaks for these lines in less than 6.4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, and are considered adequate.

ISI intervals of 10 years are found to be adequate to prevent part-through wall longitudinal and circumfer-ential cracks from reaching instability. A substantial margin against rupture exists even when large initial flaws (10% of wall thickness) are assumed.

RGE-02-004 30 Revision 0 nutech

TABLE 2-1 PARAMETERS FOR LEAK-BEFORE-BREAK ANALYSIS OF PRESSURIZER SURGE AND ACCUMULATOR LINES 0 PRESSURIZER SURGE LINE (REFERENCE 3)

Size = Outer Dia. = 10.75 in., Thickness = 1 in.

Material A376 TP316

=

Normal Mode = 612.2'F, 2235 psi. pressure Control Rod Ejection Mode = 697.2'F, 3015 psi. pressure 0 ACCUMULATOR LINE (REFERENCE 4)

Size = Outer Dia. = 10.75 in., Thickness = 1 in.

Material = A376 TP316 Normal Mode = 120'F, 2235/750 psi. pressure Loss of Load Mode = 120'F, 2628/750 psi. pressure RHR Hx Malfunction Mode = 120/300'F, 2235/750 psi. pressure RGE-02-004 31 Revision 0 nutech

TABLE 2-2 LEVEL D ASME CODE MAXIMUM ALLOWABLE STRESSES USED FOR ANALYSIS OF CRACK INSTABILITY FOR PRESSURIZER SURGE (PSL) AND ACCUMULATOR LINES (AL)

(ASME CODE SECTION III g NC 3600 g 1980 EDITION)

~allowable ~x + ~b

~t PSL 600 3 i 015 65,350 5,919 59,431 14,853 AL 100 2,628 84,600 5 I 160 79,440 12,947 where:

allowable hy A

'Shy sma 1 1 er of ( 3Sh, 2Sy )

Sh = 17 ksi., Sy = 18.8 ksi. at 600'F Sh = 18.8 ksi., S Y = 30 ksi. at 100'F SA f (1.25 Sc + 0.25 Sh)

Sc = 18.8 ksi. at 100'F f = 1 for thermal cycles ( 700 a

b allowable pa pr /(r -r ) (ri = inside radius, ro = outside radius)

P(r +r. )/(r -r. )

RGE-02-004 32 Revision 0 nutechGNOINGGRtl

TABLE 2-3 APPLIED J-INTEGRAL AND TEARING MODULUS VALUES AS FUNCTIONS OF THROUGH-WALL HALF CRACK LENGTH, a, FOR THE GINNA ACCUMULATOR LINE WITH AN APPLIED STRESS OF 30,133 PSI.

CIRCUMFERENTIAL CRACK:

~a( ai) a (in) a Stress

[Factor ,[Y ] [Y +2' Y ]

J(

'" P) in 30,133 2 0.89 0.6 1.38 1.2 1.8 301 2.5 30,133 4 1.79 0.6 1.38 1.8 3.2 904 4 ~ 4 30,133 6 2.68 0.6 1.38 2.6 4.8 1,959 6.6 30,133 8 3.57 0.6 1.38 3.3 6.2 3,315 8.6 30,133 10 4 46 0.6 1.38 4.1 8.0 5,i49 11.0 30,133 ll ~

4.91 0.6 1.38 4.5 8.4 6,216 11.6 LONGITUDINAL CRACK:

30, 133 2 0.89 0.6 1.38 2.0 4.1 502 5.7 30,133 4 1.79 0.6 1.38 4.89 12. 2,456 16.6 30,133 6 2.68 0.6 1.38 9.07 23 ~ 6,834 31.7 30,133 8 3.57 0.6 1.38 14.53 38. 14,597 52.4 30,133 10 4.46 0.6 1.38 21.22 58. 26,648 80.0 RGE-02-004 33 Revision 0 nutech

TABLE 2-4 APPLIED j-INTEGRAL AND TEARING MODULUS VALUES AS FUNCTIONS OF APPLIED STRESS FOR THE GINNA ACCUMULATOR LINE WITH A THROUGH-WALL HALF CRACK LENGTH OF 3.5 IN.

CIRCUMFERENTIAL CRACK:

Stress J( in-lb) a(ksi.) a (in.) Factor ~Y 1 [Y2+2 gY.Y' in 2 T 0.7 35 3.5 1.56 2.05 1.65 2.85 1,077 5.8 1.0 50 3.5 1.56 16.1 1.65 2.85 8,461 45.9 1.1 55 3.5 1.56 37.9 1.65 2.85 19,917 108.0 1.2 60 3.5 1.56 89.1 1 ~ 65 2.85 46,824 253.9 LONGITUDINAL CRACK:

0.7 35 3.5 1.56 2.05 4.0 11.0 2i612 22.6 1.0 50 3.5 1.56 16.1 4.0 11.0 20,511 177.1 1.1 55 3.5 1. 56 37. 9 4.0 11.0 48,285 416.9 RGE-02-004 34 Revision 0 nutechGNOINGGRG

TABLE 2-5 APPLIED J-INTEGRAL AND TEARING MODULUS VALUES AS FUNCTIONS OF THROUGH-WALL HALF CRACK LENGTH, a, FOR THE GINNA PRESSURIZER SURGE LINE WITH AN APPLIED STRESS OF 37,600 PSI.

CIRCUMFERENTIAL CRACK:

Stress J(in-1b) ai ~Factor) ill

~a( ai. ) a( in. ) Y2) (Y2+2 ~Y Yi ) T 37,600 2 0.89 0.84 4.6 1.2 1.8 894 8.28 37,600 4 1.79 0.84 4.6 1 ' 3.2 2i683 14.72 37,600 6 2.68 0.84 4.6 2.6 4.8 5,813 22.08 37,600 8 3.57 0.84 4.6 3.3 6.2 9,837 28.52 37 i600 10 4.46 0.84 4.6 4.1 8.0 15,277 36.80 LONGITUDINAL CRACK:

37,600 2 0.89 0.84 4.6 2.0 4.1 1,490 18.86 37,600 4 1.79 0.84 4.6 4.89 12. 7,288 55.20 37,600 6 2.68 0.84 4.6 9.07 23. 20,277 105.80 37,600 8 3.57 0.84 4.6 14.53 38. 43,311 174.80 RG E-0 2-00 4 35 Revision 0 notech IINQINI1GRQ

TABLE 2-6 FAILURE CRACK SIZES

• FOR POSTULATED COMPOUND CRACK IN GINNA ACCUMULATOR LINEg .BASED ON NET SECTION PLASTIC COLLAPSE CRITERION Pm+ Pb 30,133 PS'm 'b v/m 56,400 PSI.

v/n 0.318 0.116 O.l 0.291 0.05 0.094 0.2 0.257 O.l 0.070 0.3 0.216 0.15 0.043 0.4 0.165 0.2 0.012 0.5 '.096 0.2183 0.5961 0

• TERMS DEFINED IN FIGURE 2-7 RGE-02-004 36 Revision 0 nutech GNOINGGRG

TABLE 2-7 FAILURE CRACK SIZES

• FOR POSTULATED COMPOUND CRACK IN GINNA PRESSURIZER SURGE LINEg BASED ON NET SECTION PLASTIC COLLAPSE CRITERION P + Pb P + Pb 34 747 PSI. 37,600 PSI.

v/n v/m 0.248 0.223 0.1 0.215 O.l 0.189 0.2 0.176 0.2 0.148 0.3 0.127 0.3 0.096 0.4 0.063 0.4 0.029 0.4749 0 0.4350

• TERMS DEFINED IN FIGURE 2-8 RGE-02-004 37 Revision 0 nutech GNQINEKGRQ

TABLE 2-8 SA 8 Q 4 W LEAK RATE RESULTS FOR CIRCUMFERENTIAL THROUGH-WALL CRACKS (CTWC)

I Ol O I AND LONGITUDINAL THROUGH-WALL CRACKS (LTWC) IN THE PRESSURIZER I h) SURGE AND ACCUMULATOR LINES (PSL AND AL) FOR NORMAL OPERATION 0 I PRESSURE STRESSES O

OA Crack Area Ap ***

Pressure '

Leak Rate Crack Half-length Stress (in. ) for Half- (gpm. ) for a (in.) for ***

Line Crack a P

(psi) Length a = 1 in. a=1 in. 1 gpm Leak Rate PSL* CTWC 4,889 0.00113 0.567 1.314 LTWC 9,778 0.00248 1.245 <1 AL** CTWC 1,640 0.00038 0.236 1.752 LTWC 3,281 0.00083 0. 516 l. 328 Saturated Liguid Non-saturated Liquid

• • • Based on Friction Factor Cf 0.6 C

'0

TABLE 2-9 TRANSIENTS CONSIDERED IN SUBCRITICAL CRACK GROWTH RATE ANALYSES FOR PRESSURIZER SURGE AND ACCUMULATOR LINES (REFERENCE 8)

Operating Occurrences in C cle 40 r. Desi n Life

1. Startup and Shutdown 200

2. Large Step Decrease in Load (with steam dump) 200

3. Loss of Load (without immediate turbine or 80 reactor trip)

4. Loss of Power (blockout with natural circulation 40 in Reactor Coolant System)

5. Loss of Flow (partial loss of flow, one pump only) 80

6. Reactor Trip from Full Power 400

7. Hydrostatic Test (before initial startup, and 55 post operation)

8. High Head Safety Injection 50 1105 Assume 1200 Significant Cycles in 40 yr.

Design Life (30 cycles/yr.)

RGE-02-004 39 Revision 0 nutech

10 Ice@If<<oo Oo ILOO SOC LO Coa<< Neo CCLla recit Lfo CACO Nfaarw CO<<SINO otfetc<<L cceA e ceca Kc af Iota

'a ercasoc tope,aw OKIILI<<OIo Aralo 6 ooloelslf Ql l ~

ttf~ esses I <<t Vl wlI

~ ~ ~o

>'J 0

eol ta O 8 C)

O wa I<< ~ I a&<< ~ I

~a tws<<

Clsaftso '

tC I ta'fora'co I Lt' llw S~

sww ew<<

r -

-akca.coc IIL'.

IOC' Issf<<f 0 MtaWII ~ ~ ~ 0 CW<IO<<ae<<e

)g( eaaa<<fee<<ow 8 <<OIWI Q cwsw so<<<<o f + e<<t<<n C) Noew ~+

Iowa<<

Mls<<le f SS +NI c~

KIIII 4 ~, Iwoe<<I POO NNN W NCO<<l N ~ NSS gILC NI'I IO NllollO INN C.l ~

ILCOINCWIIINN N ONW NINWI L IN NINI ~ c<<ltNN Nl INWI rolllsoc N IIOOv<<C sso P

ICtltfLICL are<<saaec sstca<<L'sa aac ctaa octo<<f LN<<cceocwof Iescaot 0 ll'L o<<s ccoo ~ sssooew

<<Ito<<e cia ~ oe was ~ ae C I Aea <<IL atCC. Ia<<esca. SO%C OSO KIOLINO

~ ltsa oa<<o<<L Ia<<etc o eo Ka~ taowasaL

~ soaeoso to<<.LML <<C COO C III fels<<O Wflf Cat\1 ~

OCCSICQ CA OLLI CLOC1 ISIC. COAT CteoIL OCLteCsa L1ucttLA toLCEtt OtILttO 1'

~ I (eL I

~ CONCNC ACCCCIL1CS.OIC C7fjl et3og . 4 4 4S24 W3SI-35 3 I yK) lt o Ll t 1 FR GE82.01 Figure 2-1 PRESSURIZER SURGE LINE

>>'rsawll rso KSW Csee Ol

~rwe ~ot

~ 50 C 405

'

ea r.r. (UCO ss IllUhe coo toe

~ Ot OstlaT>>lo to>>wt>>>>L sos leo 440 P.l Ll>>wc >>sec Lwo teeeta>>eacsar>>a Mlle Tsa>>LC>>T j ooco atl ~

Vl O I bJ IL Iss les rr salle>> I ~ Ue ro~ I 0 I I ~

MI poO ~

as*seep '

>>>>

ase ~ l>>

ala ll>>

ar ~ It>> ~ I' 4 I' OICR Ur W ~ I \ar 4>> tt lss IPP OIS 5( csee rst >> law>>

rlfSi Ia "'

cl let~

~

~ r>>r4 4' WIOU Cise v laa

~ t

~

Ill

~s steel t, ~ OM>>We r ~ ~ 9 ce>>sos>>ear

~ O.I

~ ar 't ILI IU ~ oseslUUO se

)('ceerwaeew WU W wss Lt 'P ~r

+IMUU rss 400 ~< +leo>>sr QSP Q>>l

~i fQC DCrAllS CCC Cars UIIII I, ~ ISO ICS QILISWI IS SCUD 0 SSS te SUWI 0 I US g Cs0 II SsISISCIWI cccsm.sect s SSS 0 Seel W I SUSI L ss 'Kss ~ IU>>ss w sws sslssss, Otttetectl 54>>0 la>>t Ire.

swtsa>>et woe>>araM Lcl I

oas ecsoet tel tsse>>l>>T lilt erMULI>>s>>

~ sso tls eal tteeel 11OI tes wec Sale,>>tea>>ar>>l>>

sts osl tlteet cslt >>0 ~ 44

~ TUS teaw>>L ree>>II wre w>> Is>>raw se>> jwa sas

'ec'1>>ow>>st UlaLlws lee ~

Lwt 'LIS 100 eatev

~

P eocssstt ~ sit >>0 cLttteae Owl>>a SILT>>M cote SRICRSJt tetfC'rt REBEC;rEh oawcrwcl>>>>el40 eeoIMIC ~ raa stsclsr Aloccsert s. Ic'.

rsct c ssl 555 cc I I r I 4 I I t Ct F RGE82.02 Figure 2-2 SAFETY INJECTION FROM ACCUMULATOR A

2a 26 a) THROUGH-THICKNESS CIRCUMFERENTIAL AND LONGITUDINAL CRACKS OF LENGTH 2a.

b) PART-THROUGH THICKNESS CIRCUMFERENTIAL CRACK OF DEPTH a.

FGRE82.03 Figure 2-3 REPRESENTATION OF POSTULATED CRACKS IN PIPES FOR FRACTURE MECHANICS LEAK-BEFORE-BREAK ANALYSIS RGE-02-004 42 Revision 0 nutech SNOINGSRS

480 TP304 J-R CURVES

( ) IS REFERENCE FOR DATA PTS (given 400 in reference 7) g (10) 320

)12) g 240 (16)

Y 9" (1 7) 160 M

I 80 JIc = 5260 in-lb/in. 2 FOR TP 316 at 600 F, FOR COMPARISON TO TP 304 DATA BASE (reference 7) 0 0.0 0.4 0.8 1.2 1~6 2.0 2.4 CRACK EXTENSION ha, in.

FR 0 E82.04 Figure 2-4 J-INTEGRAL RESISTANCE CURVES FOR AUSTENITIC STAINLESS STEEL (reference 7 RGE-02-004 43 Revision 0 nutech GNOINGGAG

280 AL LONG.

CRACK, 30,133 AL CIRC.

MATERIAL 240 PSI CURVE (ref.7)

AL LONG.

/ CRACK a = 3.5" 200

/

/ 'NSTABLE

//

x 160 8n // / =

// cr 50,000 PSI

~ Q

/

//

a = SS,OOO 120 PSI I

AL CI RC IVI

~s CRACK, / / STABLE 80 30, 133 P PSI CONSTANT STRESS 40 a = 10" (30, 133 psi),

I CRACK S ZE VARIES CONSTANT CRACK SIZE (a .= 3. 5"),

0 STRESS. VARIES 0 40 80 120 160 200 240 CR G E83.01 Figure 2-5 J-INTEGRAL/TEARING MODULUS STABILITY DIAGRAM FOR GINNA ACCUMULATOR LINE WITH THROUGH-WALL CRACKS RGE-02-004 44 Revision 0 nutech

280 240 TP304 'J-T MATERIAL CURVE (reference 7)

PSL, a = 37,600 PSI LONGITUDINAL'CRACK, 200 STABLE UNSTABLE PSL CIRCUMFERENTIAL a= 6" CRACK a =

37,600 PSI 160 a = 10" 120 80 o,b.CONSTANT STRESS 40 (37,600 psi),

I CRACK S ZE VARIES 40 80 120 160 200 240 CRGE83.02 Figure 2-6 J-INTEGRAL/TEARING MODULUS STABILITY DIAGRAM FOR GINNA PRESSURIZER SURGE LINE WITH THROUGH-WALL CRACKS RGE-02-004 4S Revision 0 nutechENOINOGRO

CRACK N

1.0 R

g 0.8

~ tQ

'Q ~

H M H P~ + Pb = 30,133 PSI

~ (ACTUAL WORST CASES'ROM 0~ 6 STRESS REPORT)

~RH

~

r ~

o~

+ O 0.4 H g og 56 / 400 PSI Pm + Pb (ASME SCT MAXIMUM ALLOWABLE)

III CL 2 0.2 0.0 0 ~ 0 0.2 0.4 0.6 0.8 1.0 FRACTION. OF CIRCUMFERENCE, v/m (THROUGH-WALL CRACK)

CRGE83.03 Figure 2-7 FAILURE ANALYSIS DIAGRAM FOR POSTULATED COMPOUND CRACK IN GINNA ACCUMULATOR LINE, BASED ON NET SECTION PLASTIC COLLAPSE CRITERION RGE-02-004 46 Revision 0 nutech

CRACK v

N 1.0 R

Z U

ra N

0.8

~

'0 ~

H MH

~ ~

ra R

~ ra 0.6 p~ + pb = 34s747 PSI

~

oH (ACTUAL WORST CASES

+g H FROM STRESS REPORT)

~ O P o 0 ~ 4 H g + Pb 37 I 600 PSI og Pm (ASME SCT III MAXIMUM ALLOWABLE)

CL 2 R

E 0.2 0.0 0.0 0.2 0.4 0.6 0.8 1.0 FRACTION OF CIRCUMFERENCE, v/m (THROUGH-WALL CRACK)

CR 6 E83.04 Figure 2-8 FAILURE ANALYSIS DIAGRAM FOR POSTULATED COMPOUND CRACK IN GINNA PRESSURIZER SURGE LINE, BASED ON NET SECTION PLASTIC COLLAPSE CRITERION RGE.-02-004 47 Revision 0 nutech GNCIN4GRG

QTlltlTCDLPV NNOART 5lllltlTlDYAtOI KV~lT I

SXO

~ ~

~o 0 lN ZO XO 4XI RO N0 M NO SCI Nk llN NO LÃ0 La xraaunoa awun q<~g>

FRGE82.05 Figure 2-9 DIAGRAM FOR MAXIMUM STEAM/WATER FLOW RATE TO DETERMINE FLOW RATE FOR SATURATED LIQUID IN THE PRESSURIZER SURGE LINE (MOODY MODEL REFERENCE 15)

RGE-02-004 48 Revision 0 nutech GNOINQGRD

8. 103 ksi 1.485 ksi ID OD ID OD a) PRESSURE (p = 3015 psi) b) DEAD WEIGHT, CIRCUMFERENTIAL CRACK CIRCUMFERENTIAL CRACK 22.952 ksi 34 ksi OD ID ID OD

-34 KSI c) THERMAL, CIRCUMFERENTIAL d) RESIDUAL (re ference 16),

CRACK CIRCUMFERENTIAL CRACK 16.206.ksi ID OD')

PRESSURE (p = 3015 psi)

LONGITUDINAL CRACK FRGE82,06 Figure 2-10 STRESS PROFILES FOR BOUNDING CASE (reference 3)

FOR SUBCRITICAL CRACK GROWTH PREDICTIONS RGE-02-004 49 Revision 0 nutech

DEAD WEIGHT + RES1DUAL +

PRESSURE + THERMAL STRESS DEAD WEIGHT + RESIDUAL STRESS TIME (months) a) CIRCUMFERENTIAL CRACK (30 cycles per year)

PRESSURE STRESS TIME (months) FRGE82.07 b) LONGITUDINAL CRACK (30 cycles per year)

Figure 2-11 CYCLIC LOADING CONDITIONS ASSUMED FOR CONSERVATIVE SUBCRITICAL CRACK GROWTH RATE ANALYSIS OF PRESSURIZER SURGE AND ACCUMULATOR LINES RGE-02-004 50 Revision 0 nutech

0.5 CIRCUMFERENTIAL

~

0.4 CRACK gl 0.3 W

Q o 0.2 a = 0.1044" a-. = 0.10" f AT 1200 CYCLES 3.

0.1 a 0 02"

l. LONGITUDINAL CIRCUMFERENTIAL LONGITUDINAL CRACK 0.0 0 1200 100 200 300 400 500 CYCLES (10 cycles/year) CYCLES (10 cycles/year)

FRGE82.08 Figure 2-12 PREDICTED SUBCRITICAL CRACK GROWTH RATES FOR CIRCUMFERENTIAL AND LONGITUDINAL CRACKS WITH ASSUMED INITIAL DEPTHS (ai) OF 0.02 INCHES AND 0.10 INCHES FOR THE PRESSURIZER SURGE AND ACCUMULATOR LINES

4.0 REFERENCES

NUREG-0821, "Integrated Plant Safety Assessment, Systematic Evaluation Program, R.E. Ginna Nuclear Power Plant", RGE Docket No. 50-244, Draft Report, U.S. NRC, May 1982.

2. Enclosure 3 of the Safety Evaluation Report for R.E. Ginna, forwarded to RGE by NRC letter, dated February 22, 1982.

3. W. A. Massie and M. J. Harper, "Piping Stress Analysis Report, Pressurizer Surge RC-200", R. E.

Ginna Nuclear Power Plant Seismic Upgrading Program, Westinghouse Report SDTAR-80-05-10 Rev. 1, January 1981.

4 ~ W. A. Massie and V. H. Mehta, "Piping Stress Analysis Report, Safety Injection System, Section 200", R. E. Ginna Nuclear Power Plant Seismic Upgrading Program, Westinghouse Report SDTAR-80 26, Rev. 0, March 1981.

5. SEP Topic V-S, "Reactor Coolant Pressure Boundary Leak Detection," Ginna SER, February 8, 1982.

6. Telecon, G. Wrobel (RGE) to J. F. Copeland (NUTECH), "RGE-02, Leak Detection Sensitivities/

Approach at Ginna," February No. 100.2602.0001.

ll, 1983, NUTECH File 7 ~ K.H. Cotter, et. al., "Application of Tearing Modulus Stability Concepts to Nuclear Piping," EPRI Report NP-2261, February 1982.

8. "Thermal Transients and Categories," Ginna Nuclear Power Plant, Appendix H, RGGE, July 15, 1975.

9. R. Johnson, "Resolution of the Reactor Vessel Materials Toughness Safety Issue," Volume 1, for comment, Appendix B, NUREG-0744, September 1981.

10. D. A. Hale, et al., "The Growth and Stability of Stress Corrosion Cracks in Large-Diameter BWR Piping," Volume 2, Appendix A, EPRI NP-2472-SY, July 1982.

H. Tada, et al., "Stability Analysis of Circumferential Cracks in Reactor Piping Systems,"

NUREG/CR-0838, June 1979.

RGE-02-004 52 Revision 0 nutech ENGINEERS

12. J. D. Landes and J. A. Begley, "The Effect of Specimen Geometry on JIC," Fracture Toughness, Proceedings of the 1971 National Symposium on Fracture Mechanics, Part II, ASTM STP 514, 1972, pp. 24-39 P.H.F. Pao, Fluid namics, c. 1967.

'3.

D

14. I. H. Shames, Mechanics of Fluids, McGraw-Hill, New York, c. 1962, p. 162.

15. F. J. Moody, "Maximum Flow Rate of a Single Component, Two-Phase Mixture," Trans. ASME, February, 1965, pp. 134-142.

16. "NUTCRAK User's Manual," NUTECH File No.

08.039.0005.

17. R. Huet, et al., "Stress Corrosion Cracking of Type 304 Stainless Steel in High-Purity Water: A Compilation of Crack Growth Rates," EPRI NP-2423-LD, June 1982, Figure 2-15.

RGE-02-004 53 Revision 0 nutech GNOINGGRG