ML20096A892
| ML20096A892 | |
| Person / Time | |
|---|---|
| Site: | Perry  |
| Issue date: | 05/08/1992 |
| From: | Cricler P, Tyler M CENTERIOR ENERGY |
| To: | |
| Shared Package | |
| ML20096A888 | List: |
| References | |
| NUDOCS 9205110273 | |
| Download: ML20096A892 (10) | |
Text
..
PY-CEI/NRR-1491 L
)
Page 1 of 9 1
SUMMARY
TECilNICAL REPORT EVALUATION OF FLAV INDICATION (S) IN Tile PERRY FEEDVATER N07.ZLE TO SAFE-END VELDS EXAMINED DURING RFO-3 Revision 0:
May 8, 1992 Cleveland Electric Illuminating Company Prepared by:
3<
~
Dste
~
W Reviewed by:
/dN
[
_6
/92 D6te 5[B[9L
/f - [<
I Approved by:
y f j
~ ~ -
v a
gg, g,
l l'
L 9205110273 920500
'PDR -ADOCK 05000440
- P PDR-
. ~. _. - - -.
=
.. ~. _
4.
PY-CEI/NRR-1491 L Page 2 of 9 Table of Contents P,agg Stress Field Discussion...........
3 Acceptance Criteria.....
4 Crack Growth Discussion........
5 Conclusions.............
6 Table 1 - ASHE XI IVA 3200 Reported Size..
8 Figure 1 - Flav Acceptance Criteria.
9 l
(?
.~.. - - -
~ _. - - - -
I PY-CEI/NRR-1491 L Page 3 of 9 STRESS FICLD DISCUSSION No supporting analysis is necessary for flavs in the region a/t < 30% since the Hechanical Stress Improvement Process (HSIP) has been applied for flavs in
' this-envelope, thus introducing compressive stresses (negative stress intensity (K )) which prevent further flav grovth. Growth of any flav that y
may exist within the envelope a/t <= 30% and (L/2nr) <= 10% is assumed to be fully mitigated by application of MSIP as discussed previously in letter PY-CEI/NP.R-1463L dated March 4, 1992, and in Generic Letter 88-01 and NUREG-0313, Rev. 2.
l Concerning the NRC staff concern involving shrinkage effects on the piping rystem from application of mechanical stress improvement (reference Generic Letter 88-01,-Supplement 1. Item #5), i t is actually the case that HSIP produces elongation. The change in axial length is hardly field measurable, i
vith typical elongation in the range of 10 to 15 mils, producing negligible impact on the piping system.
The two feedvater nozzle safe-ends are loaded by the feedvater pipe which vas E
analyzed in GE Report 23A6987, Revision 1.
The GE steady state piping r
analysis results were used and included the effects of snubber reduction and provide axial, bending and thermal expansion effects on the nozzle safe-end.
The GE generated piping reaction loads-vere previously considered in
' determination of the Acceptance Criteria.
t a
a
,.,e_ _,.,,
..._.,c. _ _ _, _
,m,
,,,._,.___,,,.4
PY-CEl/NRR-1491 L Page 4 of 9 ACCEPTANCE CRITERIA The flaws in the safe ends are at or close to the Inconel 182 buttering and SA 500 class 1 interface.- Figure IVB-3641-1 in ASHE Section XI indicates flavs in oximity of the veld are considered to be in the veld.
The affected velds vere made by the SHAV process which dictater using ASME Section XI, Paragraph C3320 (c) for the acceptance criteria for the flav.
The equatinns found there are based on a relationship between the collapse load and flav size at incipient plastic collapse.
The acceptance level a/t (flav depth to thickness) is set-at a maximum 60% by the ASHE so that any calculated allovable values of a/t greater than 60% default to-this value, while a/t values calculated to be less than 60% retain-their calculated values. The horizontal a/t acceptante line in Figure 1 herein for all flav lengths illustrates that all calculated values of allovable a/t, from (L/2rx)
- 0% to 100%, exceed the 60% limit and defaulted to the a/t - 60% limit.
Figure 1-shows the acceptance envelope of 60% a/t for all service levels, since' emergency and faulted conditions produce negligible effects on the final projected flaw depth. The acceptance envelope is based on the combination of
-Primary Hembrane Stress _(P,), Primary Bending Stress (P ) and Thermal b
Expansion Stress (P,), including seismic loadings obtained from the feedvater piping analysis.
Included in these stresses are the effects of snubber reduction analyses performed by GE on the feedvater piping.
P
~....
,4mm.
.,i,
,r,......,,.,,.-,..,...,,,,,.,o.w%.,
e.m...
,.,,e,.c.,ew
.w.c.#.,.gw.,.,,,m,-%,...
.-n.-.-w.,-.-.4w,,-
PY-CEI/NRR-1491 L Page 5 of 9 CRACK GROVTil DISCUSSION
-The two initial flavs detected during RF0-2 vere both 0.15 inches deep (12.5%
of vall thickness) and oriented circumferentially on the inside surface of the nozzle safe end. The longest and therefore bounding flav vas detected in i
nozzle N4C and moasured 2.9 inches in length (7.5% of circumference). The sizing results of both the pre-HSIP and post-MSIP examinations conducted during RF0-3 are included in Table 1.
The RFO-3 pre-MSIP examination detected insignificant change in the nozzle N4C bounding flav and insignificant change in the initial flav detected in nozzle N4E, as compared to RFO-2 sizing results.
Figure 1 demonstrates the ample margin that exists between the size i
=!
of the existing bounding flav (Nozzle N4c) and the ASHE Code Section XI acceptance criteria.
t RFO-3 examination of nozzle N4E detected one additional nonreportable (small) flaw, not previously detected, epproximately 160' circumferentially distant from and independent of the initial flav detected during RFO-2.
This nonreportable flav is also bounded by the N4C nozzle indication.
All the "intergranular stress corrosion cracking (IGSCC) non-resistant" reactor vessel nozzles vere inspected during RFO-3 by ultrasonic examination.
I Feedvater nozzles N4C and N4E vere the only nozzles in which indications were detected.. These flavs, located in the feedvater nozzle to safe-end veld areas on nozzles N4C and N4E, are characterized as planar circumferential defects on the inside diameter.
As shown on Figure 1, the size of the bounding indication remained within the 30% depth /10% circumference envelope provided
-within NUREG-0313, Rev. 2.
MSIP has been performed on the nozzles containing I =.
f__
4 PY-CE1/NRR-1491 L 1
Page 6 of 9 CRACK GROVTH DISCUjSJON (continued) these indications.
HSlP is assumed to be effective and curtails flav growth inside the envelope a/t <- 30% and (L/2 rtr) <- 10%.
A net compressive stress occurs inside the envelope when HSIP is applied, thus producing a negative stress intensity K so that flaws in this region vould not continue to grov.
g Therefore, no further crack growth calculations are necessary.
CONCLUSIONS h
i The tvo reportable flaws, at resized in RFO-3, one in N4C and one in N4E, are vell within the NUREG 0313 Revision 2 allovable flav size " envelope" for stress improvement by MSIP to be considered ful'ly effective in flav mitigation.
MSIP was applied this outage (RFO-3) and post-HSIP UT examinations of nozzles N4C and N4E indicate no appreciable changes in flav size due to the HSIP application (see Table 1).
Since the flavs remained in the 30% depth /10% circumference envelope during the full Operational Cycle 3, and since the HSIP application can thus be considered fully effective in further crack mitigation, no crack growth calculations are required to justify a full Cycle 4 operation (12,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />).
Per NUREG 0313 Revision 2, Paragraph 5.3.1.5, the inspection category for these two feedvater nozzles is upgraded to Category E for subsequent inspections as a result of stress improvement.
Figure 1 illustrates the position of the bounding flav from RFO-2 and RFO-3 flaws (nozzle N4C) in
- he' 30%. depth /10% circumference envelope of the flav acceptance criteria.
The change of the flav size from RFO-2 to'RFO-3 is insignificant so the two points are coincident on Figute 1.
Figure 1 also shows the. ample design margin that exists between the size of the current A."nding flav and the ASME Code acceptance limit.
. -. ~.. -..
PY-CEI/NRR-1491 L
)
Pege 7 of 9 CONCLUSIONS (continuod)
Although no crack growth calculations were required this outage, it should be noted that snubber optirization has been implemented on the NSSS portion of the feedvater lines duri:.g this outage (RFO-3).
As discussed in Letters PY-CEI/NRR-1374L dated October 18, 1991 and PY-CEI/NRR-1463L dated March 4, 1992, this Summary Technical Report serves as the case-specific evaluation
-required per Regulatory Guide 1.84 when Codt case N-43'.-1 (alternative damping values-for response spectra piping analysis: reference condition #5) is utilized on piping where IGSCC has occurred. This case-specific evaluation is being submitted for NRC review even though the presence of intergranular stress corrosion cracking could not be definitively confirmed based upon the recults'of the inspections. 'The case-specific evaluation by CEI has
. considered the RF0-3 incpection results and MSIP mitigation effectiveness, as
- ta snubber d'scussed above, as well as' revised nozzle loaaingt of this
'cptimization_re-analyses for the feedvater system.
The
-~
evaluation find'the use of Code Case N-411-1 to be fully acceptable.
Due to
-the above, no additional actions' pursuant to Rc<ulatory Guide 1,84/ Code Case s
N-413' 1 are considered-necessary.
f-s
7 PY-CEI/WRR-1491-L
%ttachment-1 5ASME XI IWA 3200: REPORTED ' SIZE"****' ' '
~
FEEDWATER.N0ZZLE.TO SAFE ENDl WELD:
INDICATION SIZING-(PERRY UNIT 1'l RF0 #3
'RFD #3'.
. RFO..# 3 PRE-M91P-PRE-MSIP-
-POST-MSIP RFD #2 MANUAL UT.(IN)
- AUTOMATED UT '(IN)
AUTOMATED UT (IN)
MANUAL UT. (IN)
N4E
- 1
'0.15 x 1.8 0.15 x i.7 0.15 x 1.6 0.15 x 1.6 t
N4E #2 0.10 x
0.B 0.10 x
0.8 0.10 x
0.55 N/A N4C #1 0.15 x 2.9 0.15 x 2.8 0.15 x 2.4
'0.15 x 2.9 NOTES:
I.D. CIRCUMFERENCE -.38.45 (IN) j WALL THICKNESS.
- 1. 2- (IN)
TABLE i-p
PY-CEI/NRR-1491 L Page 9 of 9 FLAW ACCEPTANCE CRITERIA CIRCUMFERE" TIAL FLAW IN SMAW
^
^
^
.6
~
l h
.4 -
n 3-a/t
.3
$N d
sa
.2 -
to we cLE RF02 FLAW
{$
RF03 FLAW
/
-sax 1-TEY'0 Bas 14-aASEOET0t
*" ' ^
XI NO EVALUATION REQUIRED.
O g
4 g
i g
i 0
.1
.2
.3
.4
.5
.6
.7
.8
.9 1.0 L/ (2 Tr r)
N0ZZLE N4C FIGURE i
.x PY-CEI/NRR-1491 L Page 1 of 1 RFO-3 SIZING OF N4C AND N4E FLAVS UTILIL NG THE P-SCAN SYSTEM Historically, automated inspection systems have proven to yield more accurate and. repeatable results for planar flav sizing when compared to manual sizing.
In-RFO-2 time and procedure contraints prohibited the use of automated sizing on the detected N4C and N4E nozzle to safe-end flaws.
In RFO-3, automated sizing utilizing the P-SCAN (Projection Imaging Scanning) systen was used to supplement the pre-MSIP manual sizing, and was s'oTeTy used for the post-MSIP sizing '
After the; nozzles had been inspected with the P-SCAN system in the imaging mode, the flav locations and areas'of maximum amplitude vere determined during data analysis. Data was then collected from each area with two refracted longitudinal vave probes, ADEPT-60 and 45SEL, and two shear wave probes, 48ET-3 and MVB60-2, with the system in the A-SCAN mode.
Using standard sizing analysis techniques, each individual A-SCAN in the area of interest was analyzed for flav tip signals end flav opening signals.
In addition, the SuperSAFT (Synthetic Aperture Focusing Technique) sosevare was also used for A-SCAN reconstruction and image enhancen. ant. The purpose of the SAFT process-is to_ provide a more accurate and detailed image of the area of interest. Utilizing these different techniques ensures a high level of confidence in the' reported flav depth and location.
To determine the length of the flav, a sub-volume scan was performed on each area of interest. This provides higher resolution of the flav area. During analysis the attenuation is adjusted until the material noise level is i
attained. :By observing where-the flaw extremities recede into the material l.
noise' level,.the' flav length is determined.
As discussed;with-.the NRC Staff in the April 29, 1992 teleconference, the pre-MSIP manual and' automated depth sizing results of the N4C and N4E flaws
- vere tolbe within
- 2% (of vall) of that reported in RFO-2.
The manual and
-automated depth sizing data correlated well..Both techniques exhibited-
[
favorable flav signal to noise ratio of 15db or greater and. indicated
= essentially the same depth. Due to the good correlation between the pre-MSIP
. automatic and manual sizing data, the improved accuracy and repeatabilit, af the automatic sizing technique, and the significant ALARA benefits obtained vith the_ automatic sizing technique during the pre-MSIP. examinations, only l
automated sizing vas performed following application of MSIP. The data l
analysis determined there to be no appreciable difference between the ore-and
' post-MSIP sizing.
It-I.
l l
I L
-