ML20151R862
| ML20151R862 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 02/04/1986 |
| From: | Croneberger D, Kazanas N, Leshnoff S GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| Shared Package | |
| ML20151R847 | List: |
| References | |
| RTR-REGGD-01.121, RTR-REGGD-1.121 TDR-758, TDR-758-R, TDR-758-R00, NUDOCS 8602060281 | |
| Download: ML20151R862 (17) | |
Text
4PPtNOIX b
(
TDR NC.
750 REVISION NO.
0 ggg SUOGET 123125 paog 1
op 11 TECHNICAL DATA RGPORT ACTIVITY NO.
QEFARTMENT/SECTION OTSG Tube Plugging RELEASE DATE _
REVISION DATE DOCUA4ENT TITLE:
Assessment of 50% IV Repair Limit with respect to Reg. Guide 1.121 Guidelines.
DATE ORIGINATOR SIGNATURE DATE APPROVALISI SIGNATURp g
I D. K. Croneberger M -
2-3-tc S._D. Leshnoff_
/ h. Me>M
'/38/T6 l N. C. Kazanas M dwaA' 1/2J/tt, U
IU APPROVAL FkR EhTERNkL 018TRIBUTION 1
OATE ub __
R. F. Wilson N N h l
\\
Does this TOR include recomenendation(s)7 CYee O No if yes. TFWR,'T R 8 Gl5T3UGUTION AS4 TRACT:
l e
PURPOSE R. F. Wilson Analyt1 Cal results, provided by several different method's, D. K. Croneberger show that tubes with deep local flaws need not be treated H. D. Hukill the same as tubes with shallower but more extensive J. J. Colitz R. J. Toole flaws. A comprehensive plugging criteria should R. O. Barley disposition tubes based on ECT characterization of both G. R. Capodanno depth and circumferential extent for circumferential J. D. Abramovici defects and depth and length for axial defects.
R. J. McGoey N. C. Kazanas A comprehensive tube plugging criteria is developed here F. S. Giacobbe which meets or exceeds NRC guidelines on structural R. L. Miller margin, as per the guidelines of Reg. Guide 1.121, J. Jandovitz addresses ECT accuracy, aad provides a basis that an additional thickness degradation allowance is not S. Ko eabany D. D. Bowman necessary.
METH005:
l 5tructural I
Analytical results, some described in earlier reports, l
were compared against guidelines for establishing a steam generator tube plugging criteria contained in Reg. Guide 1.121. These guidelines addressed, among other things, providing a margin of 3.0 on normal loads and 1.428 on upset to prevent ductile failure (circumferential defect) or burst (axial).
l 8600060201 hD'?
ADOW pDa PDR P
l eGOVun PAGE ONLY
TDR No. 758 Rev. O Page 2 of 11 METHODS:
Cont' d Margin against ~ plastic collapse was met by using a non-linear strain analysis and exceeded by an elastic plastic fracture mechanics analysis. The margin to burst was shown to be exceeded by net section collapse methods and by actual test data.
ECT Accuracy ECT accuracy was demonstieted with metallurgical samples. The recently qualified conversi7n curve was used in conjunction with the.540 SD probe.
RESULTS:
The following results were derived:
Defect size A defect of 50% TW with a continuous length of 0.55" is acceptable.
ECT Accuracy The proposed tube plugging criteria contains a margin of ten percentage points on throughwall extent out of recognition of possible ECT error. For a 50% throughwall defect, this represents a 20% margin for error.
Using the mean and standard deviation obtained from netallurgical sanples the percent error due te undercall is 3.3%.
Additional thickness degradation allowanc'e An additional thickness degradation allowance, as suggested in Reg. Guide 1.121, has not been included because, first, the mechanism for continued chendcal attack from the inner surface has been arrested and, second, the TMI-l OTSG's do not have a history of either significant tube problens because of wear on the outer surface at the elevations of the lateral support plates or secondary side chendcal attack.
Both the NRC staff and the Atomic Safety and Licensing Board have concurred that prinary side corrosive attack is not ongoing.
4104d l
TDR No. 758 Rev. 0 Page 3 of 11 CONCLUSIONS:
1.
Conparison of the present results with the results from previous analysis, albeit that methods differed, shows that they are nearly the same.
This comparison allows the conclusion that fatigue, plastic collapse and burst concerns are all satisfied. Plastic collapse and burst are addressed for the first time herein.
2.
The proposed tube plugging criteria contains margins to failure equal to or greater than those recommended in Reg. Guide 1.121.
3.
The percent error due to undercall is less than that previously assumed.
4 The tube plugging criteria developed-here is applicable to flaws on the inner surface of the tube only.
In addition, it is applicable to the free span portion of the tube only, away from entrance effects associated with support plates.
Tubes with defects on the OD surface will be dispositioned at
~
40% TW. Tubes with indication of nearby OD and ID flaws will be dispositioned on a case-by-case basis in a conservative manner consistent with the nature of the degradations involved and the uncertainties of the ECT call.
PURPOSE The proposed GPUN OTSG tube plugging criteria provides for the structural integrity of tubes with defects against fatigue failure mechanisns and against failure in single application of large loads.
Tne latter condition, based in ASME Code practice, is recommended in Reg.
l Guide 1.121, Basis for Plugging Degraded PWR Steam Generator Tubes (Ref. 2).
l This source recommends a margin of safety against ductile failure equal to 3.0 l
x norns1 loads and 1.428 x upset loads.
In addition, identification of error l
associated with ECT is also necessary as is a discussion of an additional thickness degradation allowance.
4104d
TDR No. 758 Rev. O Page 4 of 11 Reg. Guide 1.121 does not recognize the demonstrated capability of ECT in characterizing both depth and extent of circumferential defects nor does it nake a distinction between circumferential and axial defects. A conprehensive plugging criteria should disposition tubes based on ECT characterization of both depth and circumferential extent for circumferential defects and depth and length of axial defects.
A comprehensive tube plugging criteria is developed here which meets or exceeds NRC guidelines on structural nargin, identifies a probability of ECT error, and provides basis that an additional thickness degradation allowance it not necessary.
METHODS The approach used to demonstrate structural nargin, as recommended in Reg.
Guide 1.121, is described first. The approaches to address ECT error and additional thickness degradation allowance will follow.
1.
Structural Margin Structural nargin is demonstrated in " Evaluation of GPUN proposed 0TSG Tube Plugging Criteria" (Ref. 3) prepared by Structural Integrity Associates.
Conceptual Overview.
Loads A factor of 3x nornal loads (ASME Code, Sect. III). and 1.4?8 x upset loads (ASME Code, Sect. III, App. F) is recomnended by Reg.
Guide 1.121. The basic loads originate in a B&W generic document on tube plugging (Ref. 4).
That report not only provides identification of loads under anticipated design basis conditions, I
it also provides the thernal/ hydraulic nethodology for deriving those service loads.
The dominant component in the tube axial load is thermally induced, as would occur when the OTSG shell is hotter than the tubes. The resulting load is due to thernal growth difference, or, in other words, displacenent control.
If l
displacements of interacting menbers are reduced, reactions are
(
reduced. This is in opposition to load control where reactions are independent of displacement.
4104d
+
TDR No. 758 Rev. O Page 5 of 11 Non-linear Strain Analysis Applying large factors to relatively large loads produces stresses in the region of the material stress-strain curve where displacement and load are no longer linearly related. Resistance to displacenent decreases as sterial response becomes non-l i nea r.
Reaction loads decrease as the more flexible tubes are stretched, or displaced, to conform to the growth of the OTSG vessel shell. Loads less than what are predicted by linear proportionality are actually generated.
Invoking the tube mterial actual stress-strain response shows that lower internal reactions should be used in the evaluation.
The loads that are actually developed on the OTSG tubes are identified. This is discussed in Ref. 3, Sec. 2-1; please see Fig. 2-1, specifically.
This effect is particularly important when considering circumfer-ential defects.
No such benefit exits for axial defects, however, because large strains are only possible in the longitudinal tube direction.
Failure Criteria: Net Section Collapse, Tearing Instability, and Burst.
Net section collapse (NSC) has been used by EPRI to gauge the structural integrity of pipes with circumferential defects (Ref.
5).
A defect is unacceptably large where a point on the cross-section reaches the material flow stress.
This condition is equated to ductile failure. The flow stress condition represents the departure from uniform material elongation and the on-set of neck-down deforation prior to reaching the ultimte tensile strength. The analysis of NSC proceeds from principles of solid mecha nics.
The analysis for tearing instability, however, proceeds from principles of elastic plastic fracture mechanics (EPF?!).
A crack l
in a structure my propagate a small distance and then arrest or it my tear through the material without arresting if the j
contination of load and crack size is sufficier.tly damging.
EPFM l
predicts the onset of the latter condition, i.e., teari ng instability.
The tearing modulus and applied J are computed for this purpose.
See Sect. 4.1 of Ref. 3.
Burst is the failure mode i
I for tubes with axial defects.
No benefit can be taken here for actual sterial response to reduce reaction loads because burst is load, not displacement, controlled.
Analytically, flow stress is I
taken to govern prediction of burst.
A comparison of predicted burst behavior with experimental data shows that analysis contains inherent conservatism.
See Sect. 5.3 of Ref. 3.
4104d
TOR No. 758 Rev. O Page 6 of 11 Failure by Fatigue Mechaniscu and the MSLB Analyses d?monstrating the serviceability of flawed tubes against fatigue fai' fre rrechanisms nave been previously reviewed and endorsed by the HRC staff.
These analyses included ASME Section III and Section XI fatigue evaluations, and a solid mechanics single accident load (Main Steam Line Break, MSLB) analysis conducted as part of GPU Huclear's response to the 1981 tube cracking experience, as presented in TR-008 (Ref.1).
GPUN can now take credit for that previous work in identifying minimum required tube wall thickness.
Inherent in the previous work was the capability to establish that al by fatigue analysis that inservice tubes would not develop cracks under norm 1 operating conditions, even in areas of suspected degradation and b) that existing cracks, should they go undetected, would not propagate throughwall under noral operating conditions.
CPUN's evaluation contines the methodology of both ASME Sections III and XI in order to assess the reduction in fatigue resistance caused by identified or hypothetical ECT indications. ASME Section III provides guidance for designing nuclear pressure components against failure; ASME Section XI provides guidance for evaluating the impact of suspected flaws in pressure retaining components inservice.
The Section III fatigue failure analysis uses crack initiation as the criterion for loss of fatigue resistance of the sterial; therefore, design using this approach assumes only a degraded sterial condition and not outright structural failure. The approach used to enter the ASME III design fatigue curve was originally discussed in TDR-421 (Ref.11) and is summarized in TR-008 (Ref.1), which formed a basis for MRC conclusions in NUPEG-1019.
In ASME Section XI, the methods of linear elastic fracture echanics (LEFM) are recommnded.
In this approach the presumed crack is analytically interacted with the local stress field in order to predict enlargemnt and propagation as service loads (both mechanical and therm 1) are cycled in the anticipated m nner. As discussed previously in TDR-388 (Ref.10) and TR-008 (Ref.1), a particular fracture mechanics solution was used by GPUN in order to properly model the response of a thin tube to the presence of an ID circumferential c. rack under applied axial load, internal pressure, and bending stress due to flow induced vibra tion. The aim of this analysis originally was to demonstrate the adequacy of the threshold of ECT cetection sensitivity.
The rupture strength of a flawed tube to the eximum axial load, applied one time only, was evaluated under the faulted condition of a main steam line break (MSLB).
The tube response was analyzed by mthods of solid mechanics, capturing the increased flexibility of the tube at the elevation of the flaw and utilizing the flow stress as the limiting sterial condition.
4104d
TDR No. 758 Rev. O Page 7 of 11 2.
ECT Accuracy ECT accuracy was demonstrated with metallurgical samples.
The recently qualified conversion curve was used in conjunction with the.540 SD probe to generate 6 data points for defects.
The approach taken here utilizes percent error of the ECT call with respect to actual flaw size, as shown by metallurgical examination, to establish relative error.
This approach allows conclusions concerning ECT overcall or undercall.
The margin separating the fatigue analysis results and the proposed plugging criteria is at least ten percentage points (10%)
on throughwall out of recognition of possible ECT error.
3.
Additional Thickness Degradation Allowance Additional material allowance out of recognition of both a prinary side attack combined, at the same elevation, with mechanical wear from the outer surface, as at the elevation of the upper lateral support plate, is addressed in two ways.
First, prinary side chemical attack was arrested by chemical cleaning and is prevented from reoccurring by plant chemistry procedures involving pH and lithium addition (Ref. 8).
Second, plant engineering records of the tube plugging on account of wear on the outer surface (Ref. 7) indicate.that cross-flow patterns for the generators at TMI-1 do not promote this mode of degradation. Six lane tubes were plugged on account of wear at the 15th lateral support plate as a precautionary measure. ECT techniques now in place will be employed to examine these areas.
RESULTS 1.
Structural Margin The results of the non-linear. strain analysis are shown in Figure 1.
Tube load versus displacement, assuming linearity, is shown as the bold straight line. The parallel dashed line is the 0.2%
offset yield line.
The curved dashed lines are the actual material temperature dependent engineering stress-strain curves.
As the sterial strains, the predominately thernal loads are reduced.
Dropping down from the pseudo-elastic response to the actual non-linear material response (intersection at circles) gives the true tube load.by reading back to that axis. The applied axial loads are shown multiplied by the factors of safety recommended by Reg. Guide 1.121.
i l
l l
4104d l
TDR No. 758 Rev. O Page 8 of 11 The results of the NSC and EPFM structural analyses for circum-ferential defects in tubes are shown in Fig. 2.
The analytical results are shown with respect to a piece-wise linear expression of the proposed plugging criteria. The two NSC curves (dotted) reflect the two conditions of flawed tube structural response; that the tube is flexible (triangle) and that it is inflexible.
The EPFM result is indicated as: J-T, 42 KSI. The 42 KSI follows from the industry practice for 360*, 40% TW defects.
In the area of the proposed plugging criteria, both NSC models produce results well removed from the 10% TW zone. The EPFM results are in a region well removed from the proposed plugging criteria. These results are nowhere within 10% W of the cri teria. The NSC results for a flexible tube model (triangles),
where the centroids of the defective and non-defective cross-sections tend to line-up under load reducing the internal moment reactions, come within 10% TW of the plugging criteria only for defects of very large circumferential extent.
Results for an inflexible tube (squares) come within 10% TW over a broader region of circumferential extent.
Inflexible tube response is less likely than flexible response.
The results shown in Fig. 2 are all within the proposed plugging criteria.
Figures 3 and 4 show NSC results for tubes with axial defects.
The noral and upset loads are multiplied by the factors of safety recommended by Reg. Guide 1.121. The figures indicate that the proposed plugging criteria bounds the analytical results. Figures 5 and 6 compare actual burst tests results for INC0 600 with analytical prediction.
The latter are always con 3ervative when compared to burst test results.
Except for a smll region, these results are not within 10% TW of the plugging criteria. Where there is a smil discrepancy there is argin in the analysis methods to compensate. For example, using Figures 5 and 6 and equations 5-4 and 5-5 in Ref. 3, the actual burst pressure by test, is about 22.5% greater than predicted burst pressure.
The results of the previous fatigue and MSLB analyses are provided in (Ref.1) TR-008.
The proposed plugging criteria bounds the results cf these analyses.
In the area of the proposed plugging criteria, there is at least a mrgin of 20% TW or greater.
(The argin increases with decreasing length.) Margin of this egnitude occurs when stable crack growth and not fatigue resistance are governing.
4104d
4 TDR No. 758 Rev.- 0 Page 9 of 11 ECT Accuracy ECT accuracy was demonstrated with metallurgical samples using the recently qualified conversion curve (6). The mean of six data points (Ref. 6) was 13.4% overcall. The standard deviation was +
16.7%. On this basis, a 3.3% undercall was observed.
This is -
less than the 20% undercall, on a 50% throughwall indication, already included in the proposed criteria.
Previously, (Ref. 6), a statistical presentation was nade regarding differences between ECT sizing and metallurgical resul ts. The approach taken here utilizes percent error of theECT call with respect to actual flaw size as shown by metallurgical examination. The approach allows conclusions concerning ECT overcall or undercall. Previous work discusses accuracy in terms of per _ cent throughwall units.
That approach gauges error against the total throughwall dimension. The previous work does not include an assessment of relative error, as presented above.
In the region of the proposed plugging criteria, the margin against ECT is at least 10% on throughwall as seen by inspection '
of Figure 2 and TR-008 (Ref.1). These analyses represent distinctly different solutions but allow the same conservative conclusions with regard to margin against ECT error.
DISCUSSION Application of Plugging Criteria In a strict sense, the structural model used here was that for a ID surface flaw. The applicability of the results will be limited to that geometry only.
Defects on the OD surface will be dispositioned in accordance with the existing Tech Spec repair cri terion.
The ECT sizing accuracy is established for defects on the ID surface in the free span. Applicability of these structural results will be limited to these regions.
The structural problem of OD and ID surface flaws at the same elevation has not been solved here. Tubes having this type of defect conbination will be dispositioned on a case-by-case basis in a conservative nanner consistent with the nature of the degradations involved.
4104d
TDR No. 758 Rev. O Page 10 of 11 IGA /IGSAC Previous work (Ref. 8) provided an explanation of the Novenber, 1984, tube defects. What was proposed was, essentially, that previously existing IGA /IGSAC was mechanically exercised into ECT detectabili ty.
Additionally (Ref. 9), it was found from pulled tube specinens that IGA could exist apart from IGSAC.
The structural results discussed above apply to defects whose origination is from either mechanism.
Inability to call IGA defects would inpact only the statistics associated with ECT.
If necessary, the issue of ECT margin will be revisited.should there be a deficiency in ECT with regard to detection and sizing of IGA alone.
4 CONCLUSIONS:
1.
Comparison of the present results with the results from previous analysis, albeit that methods differed, shows that they are nearly the sane.
Th'is comparison allows the conclusion that fatigue, plastic collapse and burst concerns are all satisfied. Plastic collapse and burst are addressed for the first time herein.
2.
The previous tube plugging criteria contains margins to.
failura equal to or greater than those recommended in Reg. Guide 1.121, 3.
The percent error due to undercall based on an assessment of netallurgical data is less than that assumed in GPUN structural analyses.
4.
The tube plugging criteria developed here is applicable to flaws on the inner surface of the tube only.
In addition, it is applicable to the free span portion of the tube only, away from entrance effects associated tube support plates.
Tubes with defects on the OD surface will be dispositioned at 40% TW. Tubes with indication of nearby OD and ID flaws will be dispositioned on a case-by-case basis in a conservative manner consistent with the nature of the degradations involved.
i 4104d
. - - ~.,, -,. _.
TDR No. 758 Rev. 0 Page 11 of 11 References 1.
TDR 008 Assessment of THI-l Plant Safety for Return to Service af ter OTSG Repair, Rev. 3, 8/83.
2.
Regulatory Guide 1.121, Basis for Plugging Degraded PWR Steam Generator Tubes, 8/70.
3.
J.F. Copeland and T.L. Gerber, Evaluation of GPUN Proposed OTSG Tube Plugging Criteria, S.I. Report No.: SIR-85-017, May, 1985.
4.
BAW 10146, Determination of Minimum Required Tube Wall Thickness for 177-FA Once Through Steam Generators.
5.
B.J.L'. Darlaston,. Some Aspects of Leak-Before-Break; Their Quantification and Application, Nuclear Eng'g & Design 84 (1985) 225-232, North-Holland, Ansterdam.
6.
TDR 642, Qualification of Conversion Curve for Inner Dianeter Discontinuities, Evaluation of Eddy Current Indicators During the 1984 Tech., Rev. 2.
7.
Record of Telephone Conversation, R. O. Barley to S. D.
Leshnoff, OTSG Tubes Plugged Because of Wear on the 0.D.
Surface, 5/7/85.
8.
TDR 638, Rev. 0 Evaluation of Eddy Current Indications Detected During the 1984 Tech. Spec. Inspection.
9.
TDR 686, Rev.1, Characterization of IGA in TMI-l 0TSG Tube Sa nples.
- 11. TDR 421, Rev. O, Steam Generator Adequacy of Tube Plugging and Stabilizing Criteria.
4104d
l OTSG TUBE LOADS
- e am amne PSEU00-ELASTIC,
,1.428 ($F) p.
3140 L8 x I
4,s.
l 1107(Bx3.0($F) l 39'.4 R$1
- *
- I'
~1, -
l 42 KSI x.0649 s-L
/
r 34.2 Est nada {p2 4r a-s0,e i,>
's i g,g.,
', #1 600er a.a
/
235eF e. a
/
12-
- NSCC For
/
A/T = 0.4,
.002 IN/IN x 650 IN,/
8A
~
.'i - i i
i
~
i i
i -.
0A S.4 8.8 1J 1.4 2A 14 2A TW SWudEWENT (31)
OTSG Tubing Elastic.ity Calculated and Expected Load Ofsplacement Behavior f/G /
OTSG TUBE CIRC. CRACKS NSCC. (OulVALENCY TO 40s 340 DES CPACM i.e
.\\
'\\.
\\
C.S A.
j g,,
0.7
-N
, sX y
=*
g 0.5 S
c
\\
N 0.4 --
0.3 --
0.2 o.i O.0 9.0 a.:
0.4 a..
o.s i.o c
NSCC.4:K51 A GPQN x
J-T.4:K$l v
NSCC (N0 NEUTRAL AXIAL $NIFT',
AZ K51 Net Section Collapse Criterion and J.T Instability Assults for Circumferential Cracks in Tubes with an Axial Stress of 42.0 Xsi which Permits a 3600, 405 Through the Tube Wall Crack
/~/ 6 t
t OTSG TUBE AXIAL CRACKS 3 X MCPWAL 09. STPESS WAX.
1.0 0.9 J
0.8 9.7 4A
%g 0.8 = -
D h
h 0.8 <m--
0.4
- L
-O OJ 0.t 0.1 0.0 0.0 0.4 0.8 1.2 1.4 2.0 2.4 CPACK LENGTM (IM) a 34.2 KSI A C= fun Net Section Collapse Criterion Results for Axial Cracks in Tubes at three Times Normal Operating Stress (36.161 Xst)
Ct6 3
,,m-
.n------
7
~
OTSG TUBE AXIAL CRACKS 1.429 if FAUL?t0 STPE55 WAX.
1.0 g
0.9 0.0 0.7 4
0.8 - --
mt== :: ; -
h 0.8 8""+
+---
l-04 P-
~
0.3 t
l 0.t l
Q.1 0.0 0.0 0.4 0.0 1.2 1.4 2.0 2.4 CPACK LtNOTM (IN) e 34.1 KSI A GPUN l
Net Section Collapse Criterion Results for Axial Cracks in Tubes at 1.428 Times Faulted Stress (34.068 Ksi) k-f/C. 4
i le
=
i 4 )
~
1l, 5 s 's*
@ N+ *.
s
\\
l 4
g
" O WsM7ttTts 9
5a
- * '/" 887 Y
0 II
- suf a
t
& l I/8* utt 8 a t us s' sut
,4 "iegvation 5 ),
t,t t
- N ed eted avior' L-t t
e_'
9 IS 39 N et M e4 ?$ At te 6e l.a 'na se t
(equat10A 5-5) manisessNaassifies($mut)
EftST PfttSsJIES FOR 0.428 a 0.034 IN. EDI S,073 le
' \\
f.= '1 \\$
~
.N I.
x 1, +
N i
- 0 W MFittts ji -oeici,u.a \\.N.
_ e er
., ina w.
' + San N iequation 54)
I predictied behavior.
O it 29 30 se W 64 re se se les 1.5 in, defect assat e etsassafles (5 va n)
(equation 5-5) 1 SUP.5T PRE 55URZ5 FCpt 0.625 a 0.034 IN. CLLIPTICAL mA5TAE l
I
\\
4 Burst Test Results (3) and Predictions for Tubes with Part Through Wall Defects f/4 5
80 B
e 40 -
D
,f a
+
+5 E-E
,L h
\\
3
\\
\\
E,
\\
I-c 10 i
0 0.4 0.8 tot 1.5 2
14 L8 semAL CfNE LD0fW (Met)
Q 0.050 inch wall thickness data (11]
0.0$0 inch well thickness prediction 0.030 inch wall thickness data [12]
+
--- 0.030 inch wall thickness prediction Comparison of Thru-wall Crack Surst Test Data with Predicted Senavior
, ~. 6. 6
'