ML20083N079
| ML20083N079 | |
| Person / Time | |
|---|---|
| Site: | Hatch  |
| Issue date: | 01/31/1983 |
| From: | Charnley J, Reeves P, Riccardella P NUTECH ENGINEERS, INC. |
| To: | |
| Shared Package | |
| ML20083M970 | List: |
| References | |
| GPC-04-104, GPC-04-104-R00, GPC-4-104, GPC-4-104-R, TAC-49156, NUDOCS 8302010550 | |
| Download: ML20083N079 (97) | |
Text
GPC-04-104 Revision 0 January 1983 GPC004.0104 DESIGN REPORT FOR RECIRCULATION SYSTEM AND RESIDUAL HEAT REMOVAL SYSTEM WELD OVERLAY REPAIRS AND FLAW EVALUATION AT E. I. HATCH NUCLEAR POWER PLANT UNIT 1 Prepared for Georgia Power Company Prepared by NUTECH Engineers, Inc.
San Jose, California Prepared by: Reviewed by:
L L%x 1.C.
J Charnley, D.E.
Project Engineer P. E. Reeves Project Quality Assurance Engineer p e y: 4 Issued by:
A, Ni A mv P. C. Ri rdella, P.E.
Senior Director N. Eng Project Manager
'I/
Date: --
- 26. 1983 k
8302010550 830127 0 h PDR ADOCK 05000321 (F) PDR
REVISION CONTROL SHEET Design Report for Recirculation System TITLE: and Residual Heat Removal System Weld REPORT NUMBER: GPC-04-lO4 f Overlay Repairs and Flaw Evaluation Revision 0 at Hatch 1 J. E. Charnley / Principal Engineer ,h P. C. Riccardella/ Senior Director Y. S. Wu/ Consult et I K. W. Benting/Coraultant I D[
J. R. Taylor / Consultant II hg7 T. Lem/ Consultant I [
R. D. Carignan/ Principal Engineer H. L. Gustin / Engineer NAME / TITLE INITIALS PREPARED ACCURACY CRITERIA PAGE(S) REV SY/DATE CHECK BJ/DATE CHEp sp 0 ATE 11 ill o
htc.1/as/n hkrs hkikt3 iv v
vi vii viii ix X
1 2
3 4
5 6
7 8
9 10 11 12 13 U 14 l-8-83 15 l -Af-83 16 l-2543 17 gf(b l-7,5-0 18 yf(M f-%f83 19 20 21 22 y y y CEP 3 3.1.1
' EV A 0 nutggh
REVISION CONTROL SHEET
{ Design Report for (CONTINUATION)
TITLE: Recirculation System and Residual Heat REPORT NUMBER: GPC-04-104 Removal System Weld Overlay Repairs Revision 0 and Flaw Evaluation at Hatch 1 PREPARED ACCbalACY CRITERIA SY/OATE K CHECX SY / OATE CHECK SY /JATE 23 0 hO #!U!U M # "!O 24 KUB //ts/n 25 YW I-Z&O 26 27 28 29 9 30 31 l K-6b 32 E-AS-6b 33 ySW l-JPS 34 35 36 y 37 38
- 72. //2r/t3 YSWI.zy.Q 39 40 41 Y 42 prll1t/93 43 Al'1/2s/12 44 O l'Z5-63 45 'I 46 47 \r 48 % l- M-83 49 Pfb t-25-1}
50 #AA l~JJ-54 51 'fsW I-43-8 52 53
\ /fD6 Ifl5f00 yg I-zPO 56 57 58 f 59 ki-A$2h 60 /5W p.m.t-g 61 f
? Y=*'
'; yh ms 66 Y
!s i 69 70 y y y CEP 3 3.1.2 iii REV O nutp_qh
REVI3 ION CONTROL SHEET Design Report for (CONTINUATION)
TITLE: Recirculation System and Residual Heat REPORT NUMBER: GPC-04-104 Removal System Weld Overlay Repairs Revision 0 and Flaw Evaluation at Hat.ch 1
' PREPARED ACCURACY CRITE RI A II RE"#8*3 EY / OATE CHEM SY / DATE CHEg g gAfg 71 0 f%; Ih5lV3 VYE lh'h&b (!Nb If3 ,
72 73 74
- 75 l 76 l 77 l 78 79 80 81 82 83 84 85 y P u 1' I
i l
I l
~
( CEP 3 3.1.2 REV O nutggh l
l CERTIFICATION BY REGISTERED PROFESSIONAL ENGINEER I
I hereby certify that this document and the calculations contained herein were prepared under my direct supervision, reviewed by me, and to the best of my knowledge are correct and complete. I am a duly Registered Professional Engineer under the laws of the State of California and am competent to review this C actme nt.
Certified by s f w u y f m
[ s;g assic
~
. C H.
a b@.4, i .....ly,'%h t [$ 'g8 7 %... s Cd J. E. Charnley p :.? : R%y; p 3 54 iM 0"q' 4., ,d 4 " ',
/ ': il* Professional Engineer
'A ;, ', Q -l
'is :p *~,*/;"'? v '
State of California
' yt , Q,, ,"C . ,;. : , ;
~ * ' '
Registration No. 16340
' e' h m s .-*..r Date hh Y V
GPC-04-104 v Revision 0 nutggh i
i l
)
TABLE OF CONTENTS Page i
LIST OF TABLES vii LIST OF FIGURES vili
1.0 INTRODUCTION
1 2.0 REPAIR DESCRIPTION 4 3.0 EVALUATION CRITERIA 8 3.1 Weld Overlay Evaluation 8 3.1.1 Strength Evaluation 9 3.1.2 Fatigue Evaluation 9 3.1.3 Crack Growth Evaluation 10 3.2 Flaw Evaluation 11 4.0 LOADS 12 4.1 Mechanical and Internal Pressure Loads 12 4.2 Thermal Loads 12 5.0 EVALUATION METHODS AND RESULTS 14 5.1 End Cap Evaluation 14 5.1.1 Code Stress Analysis 15 5.1.2 Fracture Mechanics Evaluation 17 5.2 Elbow Evaluation 23 5.2.1 Code Stress Analysis 24 5.2.2 Fracture Mechanics Evaluation 25 5.3 Pipe-to-Pipe Evaluation 31 5.3.1 Code Stress Analysis 31 5.3.2 Fracture Mechanics Evaluation 32 5.4 Sweepolet Evaluation 36 5.4.1 Code Stress Analysis 37 5.4.2 Fracture Mechanics Evaluation 38 5.5 Effect on Recirculation and RHR Systems 42 GPC-04-104 vi Revision 0
TABLE OF CONTENTS (Continued)
Page 6.0 LEAK-BEFORE-BREAK GS 6.1 Net Section Collapse 68 6.2 Tearing Modulus Analysis 69 6.3 Leak Versus Break Flaw Configuration 70 6.4 Axial Cracks 71 6.5 Multiple Cracks 72 6.6 Crack Detection Capability 72 6.7 Non-Destructive Examination 73 6.8 Leakage Detection 74 6.9 Historical Experience 75 7.0 SUMMAPY AND CONCLUSIONS 81
8.0 REFERENCES
83 GPC-04-104 vii Revision 0 nutp_qh
i
_ LIST OF TABLES Number Title Page 5.1 Thermal Stress Results 44 5.2 End Cap Code Stress Results 45 5.3 Elbow Code Stress Results 46 5.4 Pipe-to-Pipe Code Stress Results 47 6.1 Effect of Pipe Size on the Ratio 76 of the Crack Length for 5 GPM Leak Rate and the Critical Crack Length (Assumed Stress a= (Sm)/2)
GPC-04-104 viii Revision 0 nutegh
LIST OF FIGURES Number Title Page 1.1 Conceptual Drawing of Racirculation 3
. and RHR Systems 2.1 Schematic of End Cap Meld Overlay 5 2.2 Schematic of Elbow-to-Pipe Weld Overlay 6 2.3 Schematic of Pipe-to-Pipe Weld Overlay 7 I 5.1 End Cap Finite Element Model 48 t 5.2 Weld Overlay Thermal Model 49 5.3 Thermal Transients 50 5.4 Axial Crack Growth Residual Stress 51 5.5 Typicai IGSCC Crack Growth Data 52
=
t 5.6 Ead Cap Axial crack Growth 53 5.7 E d Cap Tearing Modulus 54 I 5.8 Elbow Finite Eleme..t Model 55 m
5.9 Circumferential Crack Growth Residual Stress 56 I 5.10 Elbow Circumferential Crack Growth 57 L
= 5.11 Elbow Tearing Modulus 58 5.12 Pipe-to-Pipe Finite Element Model 59 r
== 5.13 Pipe-to-Pipe Axial Crack Growth 60 5.14 Pipe-to-Pipe Tearing Modulus 61
~
r 5.15 Sweepolet Crack Geometry 62
\"
6- 5.16 Sweepolet Finite Element Model (Outside) 63 Illt--
, 5.17 Sweepolet Finite Element Model (Inside) 64
==
- 5.18 Sweepolet Crack Growth 65 w,
sa-7 ---
GPC-04-104 ix Revision 0 f nutp_qh
i LIST OF FIGURES (Continued)
Number Title Page 5.19 Sweepolet Tearing Modulus 66 5.20 Piping Model 67 6.1 Typical Resut*. of Net Section Collapse 77 Analysis of Cracked Stainless Steel Pipe 6.2 Stability Analysis for BWR Recirculation 78 System (Stainless Steel) 6.3 Summary of Leak-Before-Break Assessment 79 of BWR Recirculation System 6.4 Typical Pipe Crack Failure Locus for Combined 30 Through-Wall Plus 360' Part-Through Crack GPC-04-104 x Revision 0 nutggh
v i
1.0 INTRODUCTION
This report summarizes evaluations performed by NUTECH to azsess weld overlay repairs and unrepaired flaws in the Recirculation and Residual Heat Removal (RHR)
Systems at Georgia Power Company's Edwin I Hatch Nuclear Plant Unit 1. Weld overlay repairs have been applied to address leakage and additional ultrasonic (UT) examination results believed to be indicative of intergranular stress corrosion cracking (IGSCC) in the vicinity of the welds. The parpose of each overlay is to arrest any further propagation of the cracking, and to resters original design safety margins to the weld.
Tha unrepaired welds which had CT examination indicati-cos have been shown by analysis to continue to have the original design safety margins.
The required design life of each weld overlay repair is at least two fuel cycles. The amount that the actual design life exceeds two fuel cycles will be established by a combination of future analysis and testing.
Leakage was observed during overlay welding adjacent to one end cap-to-manifold weld and in addition, crack indications have been detected adjacent to three end GPC-04-104 1 Revision 0 nutp_qh a
5 cap-to-manifold welds, one elbow-to-pipe weld, one pipe-to-pipe weld and one sweepolet-to-manifold weld. All of these welds except the sweepolet-to-manifold weld were repaired with weld overlay design.s evaluated in this ,
report. The analysis of the unrepaired sweepolet weld in also contained herein.
Figure 1.1 shows the end caps, elbov, sweepolet and pipe-to-pipe welds in relation to the reactor pressure vessel and other portions of the recirculation and RHR systems. All of the existing material is type 304 stainless steel.
GPC-04-104 2 Revision 0 nutpsh
ra
&?
En O
$'f INDICATIONS IN INDICATIONS IN INDICATIONS IN INDICATIONS IN
$ fBM 1) END CAP WELD END CAP WEL.)
O #* ( BM (AM 1) r (AMA)
Q hPb INDICATIONS IN .-
i k kw ===::-
/
g' ELBOW WELD , l I (20-BD-3) 8 l
8 p l l k _
- p < ,
' T MjPbd 5 $
l f s 1 X'-
b f \.h ',,Ck N('i INDICATIONS IN y
1 SWEEPOLET I ,f0 WELD (AM-1BC-1)
[ -]N3'/ [ l s d e INDICATIONS W l[, ( PIPE TO PlFE 1 8 , wtLD (24 as ia 1 pp y, (0 LOOP A LOOP E 3 Figure 1.1 g M CONCEPTUAL DRAWING OF RECIRCULATION AND RHR SYSTEMS 17
l I 2.0 r3 PAIR DESCRIPTION The through-woll cracks and other indications around and to both sides of the existing end cap, pipe-to-pipe and elbow weld heat-affected zones have been repaired by establishing additional " cast-in-place" pipe wall thickness from weld metal deposited 360 degrees around and to either side of the existing weld, as shown in Figures 2.1, 2.2, and 2.3. The weld deposited band over the cracks will provide wall thickness equal to that required to provide the original design safety margins. In addition, the weld metal deposition 51'11 produce a favorable compressive residual stress pattern. The deposited weld metal will be type 308L, which is resistant to propagation of IGSCC cracks. GPC-04-104 4 Revision 0 r1 Lit 1)
l E Y A L R M LE U L 8V M 0O 3 I A X N - E D I DX LL PL A UO A YE IW M l H i GP PW 5 Nail 4 O G L NU {y TI OO ST t H EAi GClI
" R D%
3 A N2 Y 3 LI 7
- A L
N 1 4 8
- - - 4 - R I
M MMMM E AABB V IN "5 2
; 2 2 2 2 O M 0 S2- 22 D - - -2 qq L CCCC D 0
6 E WI R I I I R RR L 6 E F O3 1 13 83 83 W pS a 8 888 (1 1 1 1 1 P 2 C A
.
- n~
y s
" e r
1 8 uD g N N i MI F E 0 3
)
P Y yx F O M T I C I D 0 E 8 T I D L EP L IP E A E E W U M L O - DD E P S E t L l E A i O l CF C E W SI N S
- S 2 A 2M H
E S I H i sk i s, f i ( l
=
o4 I ot
.~
Based on Reference 4, the applied moments on these welds are: ! Weight + OBE Seismic = 743,100 inch-pounds Weight + Steady State Thermal = 636,100 inch-pounds l l SSE loads are not limiting for the elbow. , The thermal analysis was performed in the same manner as I for the end cap (Section 5.1.1), with appropriate dimensional changes. GPC-04-104 24 Revision 0 nutggb The results of a code stress analysis per Reference 1 are given in Table 5.3. The allowable stress values for Referenca 1 are also given. The weld overlay repair satisfies the Reference 1 requirements. l A conservative fatigue analysis per Reference 1 was performed. A fatigue strength reduction factor of 5.0 was applied due to the crack. The fatigue usage factor was then calculated assuming 38 startups, 25 small temperature change cycles and one emergency cycle every five years. The results are summarized in Table 5.3. 5.2.2 Fracture Mechanics Evaluation Three types of fracture mechanims evaluations were performed. The allowable crack depth was calculated based on Reference 2. Crack growth due to both fatigue and IGSCC was calculated using the NUTECH computer program NUTCRAK (Reference 8) with material constants and methodology from References 9 and 10. Finally, the ultimate margin to failure for a crack assumed to propagate all the way through the original pipe material to the weld overlay was calculated per Referencea 11 and 12. GPC-04-104 25 l Revision 0 l 1 nutagh 5.2.2.1 Allowable Crack Depth The allowable depth for a 3/8 inch long axial crack was determined using Reference 2. The dimensions of the repaired elbow were used. Thus, the ratio of applied primary stress to Code allowable stress (S ,) was calculated in the following manner: Stress Ratio = g ra P = 1325 psi R = 10.50 inche.< (Outside Radius of Overlay) t = 1.16 inches (Overlaid Pipe Thickndss) S, = 16,800 psi Stress Ratio = .71 l The nondimensional crack length ( A) was' calculated in ( the following manner: i ( t e A" , (rt)1/2 GPC-04-104 26 Revision 0 ! nutagh g = .375 inch (Crack Length) g r = 9.87 inches (Mean Radius of Pipe) t = 1.24 inches ^ i = .11 . Thus, per Table IWB-3642-1 of Reference 2, the allowable crack depth is 75 percent of the overlaid wall thickness or a depth of 0.d7 inch. Emergency and faulted conditions are not limiting. The allowable depth of a l-1/2 inch long circumferential crack was also determined using Reference 2. From Table 5.3, the primary stress at the crack lococion is 16,200 psi. Thus the stress ratio was calculated in the following manner: Stress Ratio = "+ =f680 = .96 g m The nondimensional crack length was calculated in the following manner: a GPC-04-104 27 Revision 0 nutg_qh f
- 1.5 =
A" .02 2xR 2w(10.5) Thus based on Table'IWB-3641-1 of Reference 2, the allowable crack depth is 75 percent of the wall thickness. Emergency and faulted conditions are not limiting. 5.2.2.2 Crack Growth Crack growth was calculated in a manner similar to Section 5.1.2.2, except: 1) the residual stress due to the weld overlay was changed to represent a one-half thickness overlay; 2) the axial residual stress was used for the circumferential crack. The axial crack which is essentially through-wall will ; grow into the IGSCC resistant weld overlay only due to I fatigue. The fatigue crack growth for five years of the thermal cycles shown in Figure 5.3 is less than 0.05 inch. Thus, the axial crack depth after five years would be 0.81 inch which is 70 percent of overlaid wall thickness, which is less than the allowable of 75 percent. GPC-04-104 28 Revision 0 nutggb The circumferential crack will grow due-to both fatigue and IGSCC. The fatigue crack growth due to five years of the cycles shown on Figure 5.3 is less than 0.05 inch. The IGSCC crack growth was calculated using-the upper bound growth curve shown in Figure 5.5 and the ~ residual stress curve shown in Figure 5.9. Crack depth as a function of time is shown in Figure 5.10. Thus, the circumferential crack depth after five years is ! approximately 0.30 inch which is 26 percent of the overlaid wall thickness which is less than the allowable of 75 percent. Thus, both the worst case axial crack and the worst case circumferential crack will not grow to an unacceptable size within the next 5 years. i 5.2.2.3 Tearing Modulus The largest size to which the existing axial crack could reasonably be expected to grow was postulated to be a 0.80 inch radius flaw. This assumes growth of the crack in the radial direction completely through the original i pipe material to the overlay. After such propagation, ( i the assumed crack would be completely surrounded by l IGSCC resistant material: the weld between elbow and pipe, the weld overlay, and*the elbow and pipe. A tearing modulus evaluation was then performed 1 GPC-04-104 29 l Revision 0 i .nutagh for this postulated crack. The normal operating loads t of pressure, weight and thermal expansion were applied. The evaluation was performed using the methodology of Reference 11 with material properties from Reference 12. The postulated flaw and the results are shown in Figure 5.11. The upper dotted line represents the inherent material resistance to unstable fracture in terms of J-integral and Tearing Modulus, T. The line originating at the origin represents the applied loading. Increasing load results in applied J-T combinations moving up this line, and unstable fracture is predicted at the intersection of this applied loading line with the material resistance line. Figure 5.11 shows that the predicted failure load is in excess of 3 times the normal operating loads. Thus, there is a safety factor on normal operating loads of at least 3, which is in excess of the safety factor l inherent in the ASME Code, even in the presence of this worst case assumed crack. i l l GPC-04-104 30 l Revision 0 l nutgsh l 5.3 Pipe-to-Pipe Evaluation The pipe-to-pipe weld number IEll-lRHR-24-BR-13 was determined by ultrasonic examination to have axial crack indications. The largest axial crack is approximately one-half inch Jong with a depth of approximately 47% of the wall thickness. 5.3.1 Code Stress Analysis The weld overlaid regions were assumed to be axisym-metric. That is, a 47% through-wall axial crack was conservatively assumed to be 360 degrees around the pipe and 1/2 inch long centered on the weld. Thus, the assumed crack geometry conservatively envelopes all observed cracks in the pipe-to-pipe weld. In addition, all analyses were conservatively performed using a weld overlay thickness of 0.30 inch which is 25 percent smaller than the actual thickness of 0.375 inch. A finite element model of the cracked and weld overlaid region was developed using the ANSYS (Reference 6) computer program. Figure 5.12 shows the model. Based on Reference 4, the applied thermal, weight and seismic moments on this weld are: GPC-04-104 31 Revision 0 nutg.gb s Weight + OBE Seismic = 1,113,000 inch-pounds Weight + Steady State Thermal = 1,626,000-inch-pounds SSE is not limiting for this weld. The thermal ~ analysis vias performed in the same manner as for the end cap (Section 5.1.1), with appropriate dimensional changes. The results of a code stress analysis per Reference 1 are given in Table 5.4. The allowable stress values for P Reference 1 are also given. The weld overlay repair satisfies the Reference 1 requirements. A cons (rvative fatigue analysis per Reference 1 was performed. An additional fatigue strength reduction fac*.or of 5.0 was applied due to the crack. The fatigue usage factor was then calculated with the thermal transients shown in Figure 5.3. The results are summarized in Table 5.4. 1 5.3.2 Fracture Mechanics Evaluation Three types of fracture mechanics evaluations were l performed. The allowable crack depth was calculated i based on Reference 2. Crack growth due to both fatigue GPC-04-104 32 Revision 0 l nutggb and IGSCC was calculated using the NUTECH computer program NUTCRAK (Reference 8) with material constants and methodology from References 9 and 10. Finally, the ultimate margin to failure for a crack assumed to propagate all the way through the original pipe material to the weld overlay was calculated per References 11 p and 12. . 4 5.3.2.1 Allowable Crack Depth The allowable depth for a 1/2 inch long axial crack was determined using Reference 2. The dimensions of the repaired pipe were used. Thus, the ratio of applied primary stress to Code allowable stress (Sm) was calculated in the following manner: Stress ratio = R/t g m P = 1325 psi (Design Pressure) R = 12.30 inches (Outside Radius of Overlay) t = 1.44 inches (Overlaid Pipe Thickness) Sm = 16,800 psi (Reference 1) GPC-04-104 33 Revision 0 nutggb Substitution yields: i j Stress ratio = .67 i ; The nondimensional crack length (1) was calculated in the following manner: K g (rt)1/2 i { A g = 3/2 inch r = 11.58 inches (Mean Radius of Pipe) t = 1.44 inches i Substitution yields: j Nondimensional Length = .12 l 2 Thus per Table IWB-3642-1 of Reference 2, the allowable crack depth is 75 percent of the wall thickness. Emergency and faulted conditions are not limiting. i l l l GPC-04-104 34 Revision 0 nutagh 5.3.2.2 Crack Growth Crack growth was calculated in a manner similar to Section 5.1.2.2. The fatigue crack growth for five years of the cycle shown in Figure 5.3 is less than 0.05 inch. The IGSCC crack growth calculated with the upper bound growth law and an infinite crack length is shown in Figure 5.13. Thus, the axial cracks in the pipe-to-pipe weld will not grow to an unacceptable size in the next 5 years. 5.3.2.3 Tearing Modulus The largest size to which the existing crack could reasonably b, expected to grow was postulated to be a 1.14 inch radius flaw. This assumes growth of the crack in the radial direction completely through the original pipe material to the overlay. After such propagation, 1 the assumed crack would be completely surrounded by IGSCC resistant material: the pipe-to-pipe weld, the weld overlay, and the annealed piping. A tearing modulus evaluation was then performed for this postulated crack. The applied loads were pressure, seismic, steady state thermal and weight. I GPC-04-104 35 Revision 0 nutagh The evaluation was performed using the methodology of Reference 11 with material properties from Reference 12. The postulated flaw and the results are shown in Figure 5.14. The upper dotted line represents the inherent , 4 material resistance to unstable fracture in terms of J-integral and Tearing Modulus, T. The line originating at the origin represents the applied loading. Increasing load results in applied J-T combinations moving up this line, and unstable fracture is predicted at the intersection of this applied loading line with the material resistance line. Figure 5.14 shows that the predicted failure load is in excess of 4 times the normal loads. Thus, there is a safety factor on normal operating loads of at least 4, i i which is well in excess of the safety factor inherent in
- the ASMC Code, even in the presence of this worst case assumed crack. $,
5.4 Sweepalet Evaluation seven small ultrasonic indications were found in the I weld between a sweepolet and the 22 inch manifold (weld l l number 1B31-lRC-22AM-1BC-1). All the indications are i i GPC-04-104 36 i Revision 0 t nutagh . .- -- .= _ . 1 1 transverse to the weld. The largest indication is 4 approximately 1/2 inch long with a depth of i approximately 12% of the wall. Figure 5.15 shows the approximate location of the indications. ] 5.4.1 Code Stress Analysis 4 i Due to the three dimensional geometry of the sweepolet and the difficulty of performing a repair, a three-I dimensional finite element model was developed using ANSYS (Reference 6). The model is shown on Figures 5.16 ~ and 5.17. 1 l Based on Reference 4, the upplied moments are: i Weight + Seismic n 176,000 inch-pounds f I Weight + Steady State Thermal = 246,000 inch-pounds i i The maximum primary stress intensity in the sweepolet is 16,600 psi which is significantly less than the i l allowable of 25,200 psi. 4 i GPC-04-104 37 Revision 0 l nutagh E______,_____ . . _ . . _ _ _ _ . _ . _ . . _ . _ _ _ _ _ _ _ ,_ _ , _.._ _ ._ _____ _ . 5.4.2 Fracture Mechanics Evaluation Three types of fracture mechanics evaluations were
- performed. The allowable crack depth was calculated based on Reference 2. Crack growth due to both fatigue and IGSCC was calculated using the NUTECH computer program NUTCRAK (Reference 8) with material constants and methodology from References 9 and 10. Finally, the ultimate margin to failure for a crack of the depth equal to the Upper bound predicted depth after an eighteen month fuel cycle was calculated per References 11 and 12.
5.4.2.1 Allowable Crack Depth Due to the three-dimensional state of stress that exists at the sweepolet, the allowable depth was calculated in the same manner as for a circumferential crack. Stress Ratio = g m I l s i GPC-04-104 38 Revision 0 nutagh Pm + Pb = 16,600 psi (Section 5.4.1) S, = 16,800 psi Stress Ratio = 0.99 The nondimensional crack length was calculated in the following manner: f - I = (rt)1/2 l Ag = .50 inch r = 11.0 inches t = .975 inch Thus per Table.IWB-3641-1 (Reference 2) the allowable crack depth is 75 percent of the wall thickness. 5.4.2.2 Crack Growth The existing cracks could grow due to both fatigue and l stress corrosion. Fatigue growth due to the three types l of thermal transients defined in Section 4.2 was l calculated using the material properties from Reference 9. The fatigue crack growth for five years of P GPC-04-104 39 Revision 0 nutagh 1 the cycles shown in Figure 5.3 Was calculated to be less than 0.05 inch. IGSCC crack growth was calculated using the upper bound crack growth law shown in Figure 5.5. The residual stress distribution normal to the crack is unknown. It was judged that the sweepolet weld residual stress would be equal or less than that due to a butt weld. Therefore, the residual stress was conservatively assumed to be 30 kui through-wall bending with tension on the inside surface. The normal stress perpendicular to the crack was determined from the finite element model. The crack growth analysis was performed per Appendix A of Reference 5. All of the observed cracks are oriented transverse to the sweepolet-to-manifold weld. Therefore, the IGSCC crack length is limited to the width of the heat-affected zone. A finite sized flaw of constant length equal to 1/2 inch was assumed. The predicted crack depth as a function of time is shown in Figure 5.18. Maximum crack depth after 5 years is predicted to be 0.38 inch (38 percent of wall thickness), which is well below the allowable of 75 percent of wall thickness. GPC-04-104 40 Revision 0 nutggb 5.4.2.3 Tearing Modulus The largest size to which the existing sweepolet crack could reasonably be expected to grow to within one fuel cycle was postulated to be a 0.50 inch radius flaw. This assumes growth of the crack at a faster rate than the upper bound prediction in Section 5.4.2.2. A tearing modulus evaluation was then performed for this postulated crack. The applied loads were pressure, seismic, steady state thermal and weight. The evaluation was performed usirg the methodology of-Reference 11 with material properties from Reference 12. The postulated flaw and the results are shown in Figure 5.19. The upper dotted line represents the inherent material resistance to unstable fracture in terms of J-integral and Tearing Modulus, T. The line originating at the origin represents the applied loading. Increasing load results in applied J-T combinations moving up this line, and unstable fracture is predicted at the intersection of this applied loading line with the material resistance line. 9 GPC-04-104 41 . Revision 0 1 nutagh Figure 5.19 shows that the predicted failure load is,in excess of 3.3 times the normal operating loads. Thus, there is a safety factor on normal operating loads of at least 3.3, which is well in excess of the safety factor inherent in the ASME Code, even in the presence of this worst case assumed crack. 5.5 Effect on Recirculation and RHR Systems Installation of the weld overlay repairs caused a small amount of radial and axial shrinkage underneath the overlay. Based on measurements of the weld overlays, the maximum axial shrinkage was 1/4 inch (elbow-to-pipe). The effects of the radial shrinkage are limited to the region adjacent to and underneath the overlay. Based on Reference 13, the stresses due to the radial shrinkage are less than yield stress at distances greater than 4 inches from the ends of the overlay. Weld residual stresses are steady state secondary stresses and thus l e p: not limited by the ASME Code (Reference 1). The effect of the axial weld shrinkage on the Recirculation and RHR Systems was evaluated with the GPC-04-104 42 Revision 0 nutagh NUTECH computer program PISTAR (Reference 14) and the - piping model shown in Figure 5.20. The four end cap weld overlays are adjacent to free ends of the recirculation manifold. Thus, axial weld shrinkage will not induce stress in any other section of the piping. The measured axial shrinkage of the elbow weld overlay (.25 inch) and of the pipe-to-pipe weld overlay (.19 inch) were imposed as boundary conditions on this model. Since the AFME Code does not limit weld residual stress, all stress indices were set equal to 1.0. The maximum calculated stress was less than 9 ksi. The location of this stress is shown on Figure.5.20. Steady state secondary stresses of 9 kai are judged to have no deleterious effect on the Recirculation or RHR Systems. 1 GPC-04-104 '43 Revision 0 nutggb NORMAL SMALL TEMPERATURE EMERGENCY STARTUP PARAMETER CHANGE CYCLE CYCL ~e CYCLE (CYCLE 1) (CYCLE 2) (CYCLE 3) U EQUIVALENT 2F 32 F 265 F LINEAR TEMPERATURE AT I 0 PEAK 0 8F 640F TEMPERATURE AT 2 THROUGH 368 PSI 8,840 PSI 72,370 PSI WALL THERMAL STRESE e l l l Table 5.1 THERMAL STRESS RESULTS GPC-04-104 Revision 0 44 nutp_q.h h l t s ACTUAL EQUATION STRESS SECTION III CATEGORY NUMBEP. NB ALLOWABLE OR THICKNESS S N/A N/A S = 16,800 PSI m , 10,590 PRIMARY (9) p 25,200 PSI PRIMARY + 18,950 (10) 50,400 PSI SECONDARY PSI ~ PEAK CYCLE 1 (23,370)5* CYCLE 2 (11) (16,950)5 N/A CYCLE 3 (129,300)5 USAGE FACTOR N/A 0.02 1.0 (5 YR)
- THE FACTOR OF 5 IS THE CONSERVATIVELY ASSUMED FATIGUE STRENGTH REDUCTION FACTOR.
l Table 5.2 END CAP CODE STRESS RESULTS GPC-04-104 Revision 0 45 nutp_qh ACTUAL EQUATION STRESS SECTION III CATEGORv. NUMBER NB ALLOWABLE OR THICKNESS S N/A N/A S, ; 16,900 DSI 16,200 25,200 PSI PRIMARY (9) PRIMARY + 19,600 < (10) 50'400 PSI SECONDARY PSI PEAX (16,200)S* C(CLE 1 8,800)5 N/A (11) CYCLE 2 ((83,900)5 CYCLE 3 USAGE FACTOR N/A 0.01 1.0 (5 YR)
- THE FACTOR OF 5 IS THE CONSERVATIVELY ASSUMED FATIGUE STRENGTH REDUCTION FACTOR.
Table 5.3 ELBOW CODE STRESS RESULTS GPC-04-104 Revision 0 46 nutggh i ACTUAL EQUATION STRESS SECTION III CATEGORY NUMBER NB ALLOWABLE OR THICK!1ESS S N/A N/A S, = 16,800 PSI PRIMARY (9) 12,300 PSI 25,200 PSI (10) 16,000 PSI 50,400 PSI 5 10ARY PEAK CYCLE 1 (19,500)5* N/A (11) (12,950)5 CYCLE 2 CYCLE 3 (125,400)5 USAGE FACTOR N/A 0.018 1.0 (5 YR)
- THE FACTOR OF 5 IS THE CONSERVATIVELY ASSUMED FATIGUE STRENGTH REDUCTION FACTOR.
T l l Table 5.4 PIPE TO-PIPE CODE STRESS RESULTS GPC-04-104 Revision 0 47 nutggb l I Rpm.M l : jf.N.j . .w {a'e i ___ l l - 1- _ Figure 5.1 END CAP FINITE ELEMENT MODEL GPC-04-104 Revision 0 48 nutggh i INSULATION +A , , , , , , , , b0 . .. . .. . .N ,,,,,, n S\\\\\\\\\\\\\\\\\\%\\\\\\\\\\\\\\\\\\\\\\\\M " A a 0.99"J 11.25" 10.01" 'I l' 4A 1 - x h h== l' ~ i ~ k = 10 BTU /hr-ft- F 't' \ s s l SECTION A-A Figure 5.2 WELD OVERLAY THERMAL MODEL GPC-04-104 Revision 0 49 nutggh l l e EMERGENCY - SMALL TEMPERATURE - 0 CHANGE N STARTUP SHUTDOWN - NORMAL - 1 OPERATION RESIDUAL - 38 25 5 CYCLES CYCLES YEARS ; ; E'; TIRE 'l' SE0VENCE CiCLE , REPEATS
- e. TIME 4
Figure 5.3 THERMAL TRANSIENTS GPC-04-104 Revision 0 50 nutggb 40 30 -. PRELIMINARY RESULTS (1/4T) EPRI PLUS WORST CASE WITHOUT OVERLAY 20 - 1C C .0 ~. m g 0 $ , a ID PIPE OD OVERLAY e m E WORST CASE WITHOUT OVERLAY -40 Figure 5.4 AXIAL CRACK GROWTH RESIDUAL STRESS GPC-04-104 Revision 0 51 nutggb 10-3 , 6 - P. Ford,1.5 ppm 02 O - R. Horn, 0.2 ppm 02 ; 104-s 5 - Uppee Bound /l [ da/dT= 1.843 x 10-12K4.615 K4.615 / h = 4.116 x 10-12 * - I e / x I F3 / 0 39 : , f Lower Bound l Ij I 1 I / 10-6 , , ,,,, , , , , , , , , ,, ,,1 i 1 10 20 50 100 1000 Stress intensity Factor (ksi /In) l Figure 5.5 TYPICAL IGSCC CRACK GROWTH DATA (WELD SENSITIZED 304S$ IN BWR ENVIRONMENT) GPC-04-104 Revision 0 52 nutgrb V A!!- - - - - - . . . . m- . . . . . . iiiA '# # # /&7 # ## a n -a -t Y L ? /####### Niii iiiV
- 1. 0 --
.;;;;;;;; 'UELDUVNRLdY" 0.8 0.6 - a CRACK GROWTH 2 WITH OVERLAY RESIDUAL STRESS 0.4 -
- 0. 2 -
0.0 , , , , 0.0 1.0 2.0 3.0 4.0 5.0 1 TIME (years) l l f ' l Figure 5.6 END CAP AXIAL CRACK GROWTH GPC-04-104 Revision 0 53 l i nutggh l 240 - 200 - I" b ~ 150 - JC = 14,500 2 x , N .5 120 - 2 ~ PRESSURE = 5500 PSI inib b UIC=6000 in2 , 80 - 40 - PRESSURE = 5000 PSI O . , , , , , 0 40 80 120 160 200 240 T OVERLAY WELD / \ ANNEALED MATERIAL ANNEALED MATERIAL WELD / 1" RADIUS ptAw I t ( l Figure 5,7 END CAP TEARING MODULUS GPC-04-104 , Revision 0 54 l nutg,gh ~ x \ - \ \ 'x \ \ \ \\ \\ , \ s k \\ \ \ \ 1 a \ \ ' / \ s \ . \ (\
- s. s y
CRACK / N LOCATIONS j/ / g . / / j 1 1 j ), 0 / / Figure 5.8 ELBOW FINITE ELEMENT MODEL vi io 55 nutRGb
- 40 i
PRELIMINARY R.ESULTS (1/2 T) EPRI PLUS WORST CASE WITHOUT OVERLAY 30 - 20 - 10 - n 3 - 0 ID m bDELB0W OD y OVERLAV i G-- u 8 m u WORST CASE WITHOUT OVERLAY l -40 , Figure 5.9 CIRCUMFERENTIAL CRACK GROWTH RESIDUAL STRESS h GPC-04-104 Revision 0 56 nutggb s L -a ~ [ _t ) 1.0
- ws 0.8-i:
=;<,a:L.
- 0. 6 J ESSENTIALLY NO IGSCC GROWTH PREDICTED a FOR CRACKS OF THIS, DEPTH N
- 0. 4 -
PREDICTED IGSCC GROWTH NEXT 5 YEARS 0.2-0.0 . . . . 0.0 1.0 2.0 3.0 4.0 5.0 1(inches) Figure 5.10 ELBOW CIRCUMFERENTIAL CRACK GROWTH GPC-04-104 Revision 0 57 nutggh 240 - 200 - inib J = 9,700 ~ 160 - in x N k1 f 3 x NORMAL LOADS $ 80 - 2.5 x NORMAL , LOADS I" 4p . JIC = 6000 jn 2 0 , . . . . . . 0 40 80 120 160 20" 240 T OVERU ? WELD / \ ANNEALED MATERIAL ANNEALED MATERIAL WELD f 2 0.60" RADIUS FLAW Figure 5.11 ELBOW TEA 9ING MODULUS GPC-0 4 -ly14 Revisiont0 58 y , nutEh. ) F _.s ____s Figure 5.12 PIPE-TO-PIPE FINITE ELEMENT MODEL GPC-04-104 Revision 0 59 nutp_qh 7 A:!: . .s.+ .. :!:ih. 'f##Ist# ;L JL .a -t ? - L ? N;r :!:V 1.0 - ;.;g;;. 'UELO'UVERLNY , 0.3 INITIAL DEPTH FOR CRACK TO GRCW TO OVERLAY IN 5 '.' EARS 0.6 - Ce 0.4 - PREDICTED IGSCC GROWTH NEXT 5 YEARS
- 0. 2 -
0.0 -- , , 0.0 1.0 2.0 3.0 4.0 5.0 TIME (years) Figure 5.13 PIPE TO-PIPE AXIAL CRACK GROWTH i GPC-04-104 Revision 0 60 nutg,gh -41 240 - i 200 - J = 12,200 ' nib , c in" ~ 160 - S N 1 4 x NORMAL LOADS .5 120 - 2 .5 w , 80 - 3 x NORMAL LOADS I" b 40 - J = 6000 2 0 . . . . . . . 0 40 80 120 160 200 240 T OVERLAY WELO / \ ANNEALED MATERIAL ANNEALED MATERIAL WELD j 1.14" RADIUS FLAW Figure 5.14 PIPE-TO-PIPE TEARING MODULUS GPC-04-104 Revision 0 61 nutggh - - - - - - - - - - - _ - - - _ - _ - - _ _ _ - _ 1 +^ ^+ l I I I I I s \ g_______/ s ._ z _ _ _ ._ _ \ \ LARGEST INDICATION APPROXIMATELY 12% THROUGH WALL \g i I i / R 8 e D HOOP ,# s STRESS / , CRACKS AXIAL STRESS II VIEW A-A Figure 5.15 SWEEPOLET CRACK GEOMETRY GPC-04-104 Revision 0 62 nutggb , / -s ' l w ~ \\ ~ \\ w \ \ ~ A\ \ / /,l/ / , / / ,/ / ~~/ ,l \ N/ Figure 5.16 ~ SWEEPOLET FINITE ELEMENT MODEL (OUTSIDE) Retisio 0 63 nutp_qh Am~ ? 3#- . ;_t = S ? k : 7 1 ys r,- ; ; u - / .s 225 --SS$5' '- /0 - - f ///// / / / / A //# / / / / Q A / // // / / / / -3 / A / // ///// / / / / // // / / // / / 5J- / / / / / / / /// / / / / ~ _=~t / / / / / / //// / / / / I I I I I I Illi l l II JII f I I I I Ill! I I I I : a ) il i! 3- l ' l I \\ \ \ \ \ \ \\\\ \ \ \ \ b \ \ \ \ \ \\\\\ \ \ \ \ L x\\ \\\ \\ \\\\\\ \ \ \ \ J \\\\\\\\\\\ xxxNNxxxxNNNNN ? xxxxxxxxxxxxxx 'hh?hXWO???,- ? ', ,' l l lll , l , Figure 5.17 SWEEPOLET FINITE ELEMENT MODEL (INSIDE) GPC-04-104 - Revision 0 64 nutpah f, pa ce f // / / / Pk/ #1/ / r i i i + _, z . + / / / / / / / /
- 1. 0 -
m-- 0.8-ys 0.6-0.4- IGSCC CRACK GROWTH 0.2-0.0 , , , . _ 0.0 1.0 2.0 3.0 4.0 5.0 TIME (years) _ 1 1 1 .. Figure 5.18 SWEEPOLET CRACK GROWTH _ GPC-04-104 Revision 0 65 nutp_qh 240 - l
- 200 -
u 150 - - x N { 120 - inib =- 3 JIC = 6000 in g 80 - Jc= 6000 i- in 2 3 x NORMAL I LOADS - l 40 - , I V 3.3 x NORMAL f # LOADS 0 40 80 120 160 200 240 h T 3 E ANNEALED MATERIAL ANNEALED MATERIAL % 0.5" e WELD] RADIUS FLAW Figure 5.19 SWEEPOLET TEARING i.10DULUS GPC-04-104 Revision 0 66 m_._. . _. .. i ji 5 3- < d f (- : _ g f.2 . 2 .-
- = - -
=
- l. l
/ i / : , . . . , c t E =- ( 5 :
- - :- :- :- j's-t- .
j: f f a \ : W e - 2o 'f .. a0 (f ' 2 - - I,_ : * ._ _ _3 _ 4- ~ : f 3 5 5 .y _- ,/ w ../ W w O'. . , . ~ ~ m,
- W oe Bw
.- w =< CU =- -o =a Ns \ GPC-04-104 Revision 0 67 nutggb 6.0 LEAK-BEFORE-BREAK 6.1 Net Section Collapse The simplest way to determine the effect of IGSCC on the structural integrity of piping is through the use of a simple " strength of materials" approach to assess the load carrying capacity of a piping section after the cracked portion has been removed. Studies have shown (References 10 and 12) that this approach gives a conservative, lower-bound estimate of the loads which would cause unstable fracture of the cracked section. Typical results of such an analysis are indicated in Figure 6.1 (Reference 10). This figure defines the locus of limiting crack depths and lengths for circumferential cracks which are predicted to cause failure by the net section collapse method. Curves are presented for both typical piping system stresses and stress levels equal to ASME Code limits. Note that a very large percentage of pipe wall can be cracked before reaching these limits (40% to 60% or circumference for thrcagh-wall cracks, and 65% to 85% of wall thickness for 360' part-through cracks). GPC-04-104 68 Revision 0 nutg,gh Also shown in Figure 6.1 is a sampling of cracks which have been detected in service, either through UT examination or leakage. In each case there has been a comfortable margin between the size crack that was observed and that which would be predicted to cause failure under service loading conditions. Also, as discussed below, there is still considerable margin between these net section collapse limits and the actual cracks which would cause instability. 6.2 Tearing Modulus Analysis Elastic-plastic fracture mechanics analyses are presented in Reference 12 which give a more accurate representation of the crack tolerance capacity of stainless steel piping than the net section collapse approach described above. Figures 6.2 and 6.3 1 graphically depict the results of such an analysis (Reference 12). Through-wall circumferential defects of arc-length equal to 60* through 300* were assumed at _- various cross sections of a typical BWR Recirculation System. Loads were applied to these sections of sufficient magnitude to produce net section limit load, and the resulting values of tearing modulus were - compared to that required to cause unstable fracture -~ GPC-04-104 69 Revision 0 (Figure 6.2). Note that in all cases there is substantial margin, indicating that the not section collapse limits of the previous section are not really f ailure limits. Figure 6.3 summarizes the results of all such analyses performed for 60' through-wall cracks in terms of margin on tearing modulus for stability. The margin in all cases is substantial. 6.3 Leak Versus Brer.k Flaw Configuration of perhaps more significance to the leak-before-break argument is the flaw configuration depicted in Figure 6.4. This configuration addresses the concerns raised by the occurrence of part-through flaws growing, with respect to the pipe circumference, before breaking through the outside surface to cause leakage. Figure 6.4 presents typical size limitations on such flaws based on the conservative, net section collapse method of Section 6.1. Note that very large crack sizes are predicted. Also shown on this figure are typical detectability limits for short through-wall flaws (which are amenable to leak detection) and long part-through flaws (which are amenable to detection by UT). The l margins between the detectability limits, and the conservative, net section collapse failure limits are GPC-04-104 70 Revision 0 nutp_qh substantial. It is noteworthy that the likelihood of flaws developing wi.ich ate characterized by the vertical axis shown in Figure 6.4 (full 360* circumferential with no thr'ough-wall component) is so remote as to be considered impossible. Material and stress asymmetries always tend to propagate one portion of the crack faster than the bulk of the crack front, which will eventually result in " leak-before-break". This observation is borne out by extensive field experience with BWR IGSCC. 6.4 Axial Cracks The recent IGSCC occurrences at Monticello and Hatch 1 were predominately short, axial cracks which grew through the wall but remained very short in the axial direction. This behavior is consistent with expectations for axial IGSCC since the presence of a sensitized weld heat-affected zone is necessary, and this heat-affected zone is limited to approximately 0.25 inch on either size of the weld. Since the major loadings in the above net section collapse analysis are bending moments on the cross section due to seismic loadings, and since these loads do not exist in the circumferential direction, the above leak-before-break arguments are even more persuasive for axially ' oriented GPC-04-104 71 Revision 0 nutp_qh cracks. There is no known mechanism for axial cracks to __ lengthen before growing through-wall and leaking, and . the potential rupture loading on axial cracks is less .= than that on circumferential cracks. . 6.5 Multiple Cracks Recent analyses performed for EPRI (Reference 15) indicate that the occurrence of multiple cracks in a weld, or cracking in multiple welds in a single piping - line do not invalidate the leak-before-break arguments _ discussed above. 6.6 Crack Detection Capability IGSCC in BWR piping is detected through two means: non-destructive examination (NDE) and leakage detection. Although neither is perfect, the two means complement one another well. This detection capability combined _ with the exceptional inherent toughness of stainless steel, results in essentially 100% probability that IGSCC would be detected before it significantly degraded the structural integrity of a BWR piping system. GPC-04-104 72 Revision 0 nutp_qh 6.7 Non-Destructive Examination The primary means of nondestructive examination for IGSCC in BWR piping is ultrasonics (UT). This method has been the subject of considerable research and development in recent years, and significant improvements in its ability to detect IGSCC have been , achieved. Nevertheless, recent UT experience at Monticello, Hatch 1, and elsewhere indicate that there is still considerable r. ..a for improvement, especially in the ability to distinguish cracks or crack-like indications from innocuous geometric conditions. Figure 6.4, however, illustrates a significant aspect of UT detection capability with respect to leak-before-break. The types of cracking most likely to go undetected by UT are relatively short circumferential or axial cracks which are most amenable to detectien by leakage. Conversely, as part-through cracks lengthen, and thus become more of a concern with respect to leak-before-break, they become readily detectable by UT, and are less likely to be misinterpreted as geometric conditions. This argument is further enhanced by the usual practice of supplementing the UT inspection with radiography (RT) when large UT indications are GPC-04-104 73 Revision 0 nutgrh . . . . . , . . . . .. . . . ~ . - - . . . - . . N Ed%1 i N 4? N3 observed. If a long UT indication is truly a geometric ig[ == condition, it will be observable as density differences ; on the radiograph. If, on the other hand, no $-c significant RT density differences are observed in the -C vicinity of the UT indication, (or if the density differences are abrupt and crack-like), the observed 7? ~ indication is usually diagnosed as IGSCC. 1 d-6.6 Leakage Detection _i_ M - Typical leakage detection capability for BWR reactor ,2 coolant system piping is through sump level and drywell activity monitoring. These systems have sensitivities on the order of 1.0 gallon per minute (GPM) of unidentified leakage (i.e., not from known sources such 3 as valve packing or pump seals). Plant technical _g specification limits typically require investigation / 2f corrective action at 5.0 GPM unidentified leakage. F. _ Table 6.1 provides a tabulation of typical flaw sizes to . cause 5.0 GPM leakage in various size piping ' (Reference 10). c Also shown in this table are the critical crack lengths for through-wall cracks based on the net section 6 GPC-04-104 74 _ Revision 0 L- == Ilu - collapse method of anaiysis discussed above. For conservatism, the leakage values are based on pressure stress only, while the critical crack lengths are based on the sum of all combined loads, including seismic. (Considering other normal operating loads in the leakage analysis would result in higher rates of leakage for a given crack size.) Note that there is considerable margin between the crack length to produce 5.0 GPM leakage and the critical crack length, and that this margin increases with increasing pipe size. 6.9 Historical Experience The above theories regarding crack detectability have been borne ost by experience. Indeed, of the approximately 400 IGSCC incidents to date in BWR piping, all have been detected by either UT or leakage, and none have even come close to violating the structural integrity of the piping (Reference 15). GPC-04-104 75 Revision 0 nutagh 1 l ., NOMINAL CRACK LENGTH FOR CRITICAL CRACK gfg , PIPE SIZE LENGTHci O n. ) c . 5 GPM LEAK (in.) 4" SCH 80 4.50 6.54 0.688 10" SCH 80 4.86 15.95 0.305 - 24" SCH 80 4.97 35.79 0.139 ri Table 6.1 . EFFECT OF PIPE SIZE ON THE RATIO OF THE CRACK LENGTH FOR 5 GPM LEAK RATE AND THE CRITICAL CRACK LENGTH (ASSUMED STRESS a = Sm /2) GPC-04-104 Revision 0 76 nutp_qh -f r 1 17 N - d 6b - t m 1 'r I m 1.0 N O \ - \ p \ _ \ 0.8 -3 s \ g #m = 6 ksi, Pb=0 i 9 % k %'*= ~~~ 5 @@ - r 0.6 - P m = 6 ksi, Pm+Pb = 1.5 S m j:: .z ~ g @@@ __ 0.4 @ Field Data - Part-9 Through Flaws 'O j 3 o Field Data - Leaks _? C S m = 16.0 ksi 2 s of = 48.0 ksi 0.2 - Values at 5500F _ g 0- -- 0.6 0.8 1.0 0 0.2 0.4 . ~ Fraction of Circumierence, shr Figure 6.1 TYPICAL RESULT OF NET SECTION COLLAPSE ANALYSIS OF CRACKED STAINLESS STEEL PIPE GPC-04-104 Revision 0 77 .-- nut %h. _ .. .. . , . - . . ~ , _ , , . _ . . . , _ . ,- - - , ., . , . . , - . 550 - " ' 0 t
- p . & r " ~ 210 M -
Sucuan WW ! Discherve 300 T=0 250 - Material \ N 200 - (Unstable) S X N $ 150 - (Stable) a 29 =1200 29 = @ 723 240a 1 100 - , p 50 - i Jge 29 = 3000 0, , , , , , , 50.0 12.5 25.0 82.5 100.0 137.5 175.0 212.5 250.0 T Figure 6.2 STABILITY ANALYSIS FOR BWR RECIRCULATION SYSTEM (STAINLESS STEEL) GPC-04-104 Revision 0 78 nut 29_h. o a ~$ Nm 6 55 s 08 zo 3 mE w* o ~* $$ 0 52 = ~ $ g$ Es G g 5" 5s =g 5 o g5 g= de o x x m ; ;: un =5 "s g =z$ oH de ma g eo m gA <o o 4 &$g 20 * .o w xm-55 w *b E8 - CE y h* oN $E " 2 [5 8=$ >o c ~$ h ~ ? t i N'x'%
- i -o SNOIIVAH2SEO JO H3EWGN J,
GPC-04-104 Revision 0 79 nutp_qh 2= 1 t a PIPE CROSS SECTION ~ 0.7 0.6 - - - - - - - - - -- l < 0.5 - l l 1 0.4 - g t 1
- 0. 3 - l 1
l 0.2 - I 0'1 -[Ist LEAK MONITOR l I I 0.0 , 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7' = /n Figure 6.4 TYP! CAL PIPE CRACK FAILURE LOCUS FOR COMBINED THROUGH-WALL PLUS 3600 PART-THROUGH CRACK GPC-04-104 Revision 0 80 nutggh ; i l s 7.0
SUMMARY
AND CONCLUSIONS The evaluation of the sweepolet flaws and the repairs to the Recirculation and RHR Systems reported herein shows that the resulting stress levels are acceptable for all design conditions. The stress levels have been assessed from the standpoint of load capacity of the components, fatigue, and the resistance to crack growth. Acceptance criteria for the analyses have been established in Section 3.0 of this report which demonstrate that:
- l. There is no loss of design safety margir. over that provided by the current Code of Construction for Class 1 piping and pressure vessels (ASME Section III).
- 2. During the design lifetime of each repair, the observed cracks will not grow to the point where che above safety margins would be exceeded.
l Analyses have been performed and results are presented which demonstrate that the sweepolet flaws and the GPC-04-104 81 Revision 0 nutagh
repaired welds satisfy these criteria by a large margin, and that:
- 1. The design life of each repair is at least five !
years. j 2. The sweepolet flaws will not grow to an unacceptable size within five years. Furthermore, it is concluded that the recent IGSCC experienced in the reactor recirculation system at Hatch 1 does not increase the probability of a design basis pipe rupture at the plant. This conclusion expressly considers the nature of the cracking which has been repaired at Hatch 1, and the likelihood that other similar cracking may have gone undetected. The conclusion is based primarily on the extremely high inherent toughness and ductility of the stainless steel piping material; the tendency of cracks in such piping to grow through-wall and leak before affecting its structural load carrying capacity (which indeed was the case in the defects observed at Hatch 1); and the fact that ae cracks lengthen and are less likely to " leak-before-break", they become more amenable to detection by other NDE techniques such as UT and RT. GPC-04-104 82 Revision 0 l nutgfj)
l l l l l
8.0 REFERENCES
1
- 1. ASME Boiler and Pressure Vessel Code Section III, Subsection NB, 1974 Edition with Addenda through Summer 1975.
- 2. ASME Boiler and Prossure Vessel Code Section XI, Paragraph IWB-3640 (Proposed), " Acceptance Criteria for Austenitic Stainless Steel Piping" (Presented to Section XI Subgroup on Evaluation Standards in l
November 1982).
- 3. General Electric Design Specification 22A1344, Revision 3.
- 4. General Electric letter G-GPC-2-511, "IE Bulletin 79-14 for E. I. Hatch Nuclear Plant Unit 1-Transmittal of Preliminary Results of Recirculation System Analysis and Design Drawings,"
Dacembe:: 17, 1982.
- 5. ASME Boiler and Pressure Vessel Code Section XI, 1980 Edition with Addenda through Winter 1981.
I t GPC-04-104 s 83 Revision 0
i I
- 6. ANSYS Computer Program, Swanson Analysis Systems, Revisions 3 and 4.
- 7. Schneider, P. J., " Temperature Response Charts,"
John Wiley and Sons, 1963.
- 8. NUTCRAK Computer Program, Revision 0, April 1978, File Number 08.039.0005.
- 9. EPRI-2423-LD, " Stress Corrosion Cracking of Type 304 Stainless Steel in High Purity Water - a Compilation of Crack Growth Rates," June 1982.
- 10. EPR.I-NP- 2 47 2, "The Growth and Stability of Stress Corrosion Cracks in Large-Diameter BWR Piping,"
July 1982.
- 11. NUREG-0744, Volume 1 for Comment, " Resolution of the Reactor Materials Toughness Safety Issue."
- 12. EPRI-NP-2261, " Application of Tearing Modulus l
Stability Concepts to Nuclear Piping," February 1982. i GPC-04-104 84 Revision 0 nutagh
- 13. NUTECH Report NSP-81-105, Revision 2, " Design Report for Recirculation Line Safe End and Elbow-Repair, Monticello Nuclear Generating Plant,"
December 1982.
- 14. NUTECH Computer Program PISTAR, Version 2.0, Users Manual, Volume 1, TR-76-002, Revision 4, File Number 08.003.0300.
- 15. Presentation by EPRI and BWR Owners Group to U.S.
Nuclear Regulatory Commission, ' Status of EWR IGSCC Es De'velopment Program," October 15, 1982. s GPC-04-104 85 Revision 0 nutggb
. m %.-e, m
, 4 i r. en. .t 7, ,,i..
, 4 +, n.w: .i o<s Ta 1, ..
' ,o ATTACINENT 7
. - ; *(
o . Georgia Power J. T. Beckha.ti. Jr. ' ##" Pr n r
.r2. mon **st' .r
.ne n ov r,
,cjg4 U. S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Division of Licensing Washington, D. C. 20555 i
ATTENTION: Mr. Darrell G. Eisenhut, Director NRC DOCKETS 50-321, 50-366 OPERATING LICENSES DPR-57, NPF-5 EDWIN I. HATCH NUCLEAR PLANT UNITS 1, 2 NUREG-0313, REVISION 1 IMPLEMENTATION RESPONSE Gentlemen: Georgia Power Company hereby submits the follcwing information in response to NRC Generic Letter 81-04 dated February 26, 1981, regarding the implementation of MUREG-0313, Revision 1, " Technical Report On Material Selection and Processing Guidelines For BWR Coolant Pressure Boundary ' Piping". Please be advised that Technical Specification changes will not be submitted for either unit for implementation of the NUREG, but, by this letter, Georgia Power Company commits to performing augmented inservice inspection on " nonconforming, nonservice-sensitive" and " nonconforming, service-sensitive" piping as discussed herein. In addition, Technical Specification changes pertaining to leakage detection systems will not be made as existing Technical Specifications for both units and the leakage detection systems meet the intent of NUREG-0313, ., Revision 1 and Regulatory Guide 1.45. The Hatch units were built before issuance of NUREG-0313, and stainless steel piping and safe ends at that time were Type 304 material. This material, including weld material, as defined in NUREG-0313, Revision 1 is i considered as nonconforming. Class 1 " nonconforming, nonservice-sensitive" and " nonconforming, service-sensitive" piping and safe ends (as defined 'in NUREG-0313, Revision 1) found in each of the Hatch units are as follows: Nonconformino, Nonservice-Sensitive 22" Reactor Recirculation (Recirc) Manifold 28" Reactor Recirculation (Recirc) Suction and Discharge 6" Reactor Water Cleanup (RWCU) Suction 1
@70ftTNI
j
., . A ueorga Power mA U. S. tAJclear Regulatory Commission Office of Nuclear Reactor Regulation Washington, D. C. 20555 Page Two June 29, 1981 Nanconformino, Service-Sensitive 4" Reactor Recirculation (Recirc) Bypass 12" Reactor Recirculation (Recirc) Risers Recirc Inlet Nozzle Thermal Sleeve Attachment Welds 20" Residual Heat Removal (RHR) Suction 24" Residual Heat Removal (RHR) Return The 10" Core Spray and 3" Control Red Drive (CRD) hydraulic return piping would normally have been identified as nonconforming, service-sensitive =
material; however, modifications have been performed either prior to or during commercial operation and conforming materials have been installed. Therefore, those particular lines are exempt from augmented inservice inspection. It should be noted that there is not any Class 2 stainless steel on either unit under the jurisdiction of the inservice inspection program. Georgia Power Company takes exception to the designating of Recirc Riser piping as being service-sensitive. The NUREG itself indicatus that there has not been any incidence of intergranular stress corrosion cracking (IGSCC) in domestic plants. In addition, by designating the Riser pipinq accordingly, inspection personnel would be faced with much higher radiation exposure in order to meet NUREG-0313, Revision 1 examination requirements. Therefore, Georgia Power Company will consider the Recirc Riser piping as having the same inspection requirements as nonservice-sensitive piping and will perform augmented inservice inspection on an 60-month frequency in accordance with the requirements of Section IV.B.l.b of NUREG-0313, Rev.1. This position is considered justifiable since there has not been any incidence of IGSCC of this particular piping in domestic BWR's. In, addition, radiation levels were aporoximately 2 R/hr. on several of the Recirc Riser piping welds on Hatch Unit 1 during the last outage and ALARA concepts must be considered and adhered to. If examination of the Piser piping during the 80-month period reveals no incidence of IGSCC, the examination frequency thereafter shall revert to 120 months as prescribed in Section XI of the ASME Eoiler and Pressure Vessel Code. l
4 A Georgia Power zA U. S. Nuclear Regulatory Commission , Office of Puclear Reactor Regulation Washington, D. C. 20555 Page Three June 29, 1981 Augmented inservice inspection will be performed on both Hatch units for the nonconforming, nonservice-sensitive and nonconforming, service-sensitive piping. Class 1 welds will be selected in accordance with NUREG-0313, Revision 1 Sections IV.B.l.b and IV.B.2.b for nonservice-sensitive and service-sensitive piping, respectively, except that high stress welds will be identified from the unit stress report. , Hatch Unit 1 service-sensitive piping was fully examined during the inservice inspection performed daring the 1978 refueling outage per original commitment to NRC. No indications of IGSCC were detected. Examination areas included 4" Racirc Bypass (capped), 3" CRD Hydraulic Return (capped), 20" RHR Suction, 24" PHR Peturn, and 10" Core Spray, Care Spray piping and , safe ends have subsequently been replaced with conforming material. The CRD Hydraulic Return was cut and capped curing the previous refueling outage and is of conforming material also. The Recirc Inlet Nozzle Thermal Sleeve attachment welds were not examined at that time as they were not considered '. service-sensitive. Cecrgia Power Company met its original commitment to NRC and after more than three years of operation found no indications of IGSCC and considers performing examinations during three successive refueling outages as unnecessary. The next refueling outage for Unit 1, which is tentatively scheduled for Soring 1982, will be considered the first of the three 36-month + 12-month examinations. Examination areas will include 4" Recirc Bypass (capped), Recirc Inlet Nozzle Thermal Sleeve attachment welds, 20" RHR Suction, and 24" RHR Return. Georgia Power Company commits to performing a 50% examination of the Recirc Inlet Nozzle Thermal Sleeve attachment welds, dependent upon radiation levels encountered, during the next refueling outage. In case indications are found, the other 50% will be examined. In the event these 36-month + 12-month period examinations reveal i no unacceptable indications for three sIJccessive inspections, the frequency l of examination will revert to 80-month periods. This schedule for performing the examinations of the service-sensitive lines is considered justified due to good inservice inspection and operating histories. l The examination of the Unit 1 nonservice-sensitive lines, which will l include the Recirc Riser piping, will be examined on an 80-month frequency l as required by the NUREG. Examinations on this frequency are considered to have started during inservice inspection activities in March 1981 (during the refueling outage). In the event there are no indications, examination frequency will revert to 120 months as prescribed by Section XI of ASME Code. i l Examination of the Unit 2 service-sensitive lines will continue to be
~
performed in accordance with commitments already made tc NRC to examine the lines during three successive refueling outages. The examination schedule is included in the existing Unit 2 Long-Term Inservice Examination Plan for l t
.,.y. -
e A GeorgiaIbwer n U. S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Washington, D. C. 20555 Page Four June 29, 1981 Class 1 Components. The Recirc Inlet Nozzle Thermal Sleeve attachment welds will be added to those examinations to be performed. It should be noted that Georgia Power Company will take credit for the 100% examination of the attachment welds performed during January 1979 and commits to performing examinations on a representative sample of 50% of the attachment welds, dependent upon radiation levels, during the next two refueling outages. If examinations of service-sensitive lines are free of IGSCC indications, the examination frequency will be extended to three 36-month + 12-month period examinations, and later to an 80-month period examination, as appropriate. Unit 2 nonservice-sensitive line examinations, including the Recirc Riser piping, will be performed on an 80-month frequency. These examinations shall be considered to have started at commerical operation (i.e. , September 1979) of the unit. The examination frequency will revert to a 120-month frequency if the lines are free of IGSCC indications after the 80-month examination period. With regard to specifying a replacement schedule for the nonconforming material as requested by NRC Generic Letter 81-04, Georgia Power Company cannot justify the indiscriminate replacement of piping that has not shown signs of IGSCC in the Hatch plant. The inspection program described above should identify development of IGSCC in the systems involved. Due to this and the high radiation exposure involved in the replacement of the piping, Georgia Power Company does not plan to replace piping that has not shown evidence of IGSCC at Plant Hatch. In the event that repairs or replacement of nonconforming material is required, at such time the affected component will be replaced with conforming material and processed in accordance with Section III of NUREG-0313, Revision 1. If you have any questions in this regard, please contact this office. Sincerely yours, hr = J. T. Beckham, Jr. Vice President and General Manager Nuclear Generation Qgi JAE/mb xc: M. Manry j R. F. Rogers, III i l L}}