ML19317H006
| ML19317H006 | |
| Person / Time | |
|---|---|
| Site: | Oconee, Rancho Seco  |
| Issue date: | 01/16/1978 |
| From: | Mattimoe J SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| To: | Goller K Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8004020683 | |
| Download: ML19317H006 (11) | |
Text
O 3
,e
+,o REGULATORY INFORMATION DISTRIBUTION SYSTEM (RIDS)
DISTRIBUTION FOR INCOMING MATERIAL 50-312 REC: GOLLER K R ORG: MATTIMOE J J DOCDATE: 01/16/78 NRC SACRAMENTO MUN UTILITY DISTRIC DATE RCVD: 01/25/7@
DOCTYPE: LETTER NOTARIZED: NO COPIES RECEIVED
SUBJECT:
LTR 1 ENCL 1 FORWARDING EVALUATION OF DISTORTED EDDY CURRENT SIGNALS OBSERVED AT SUBJECT FACILITY.
PLANT NAME.RANCHC SECO (SMUD)
REVIEWER INITI AL:
XJM DISTRIBUTOR INITIAL:
co*o****,********
DISTRIBUTION OF THIS MATERIAL IS AS FOLLOWS ******************
GENERAL DISTRIBUTION FOR AFTER ISSUANCE OF OPERATING LICENSE.
(DISTRIBUTION CODE AOO1)
FOR ACTION:
BRANCH CHIEF REID**W/7 ENCL INTERNAL:
(REG FILE **W/AI NRC PDR**W/ ENCL T & E**W/2 ENCL OELD**LTR ONLY HANAUER**W/ ENCL CHECK **W/ ENCL EISENHUT**W/ ENCL SHAO**W/ ENCL BAER**W/ ENCL BUTLER **W/ ENCL GRIMES **W/ ENCL J.
COLLINS **W/ ENCL J.
MCGOUGH**W/ ENCL EXTERNAL:
LPDR'S SACRAMENTO, CA -- REGION V**W/ ENCL TIC **W/ ENCL NSIC**W/ ENCL ACRS CAT P**W/16 ENCL DISTRIBUTION:
LTR 40 ENCL 39 CONTROL NBR:
-?o^26^1'1 SIZE: 1P+9P U
occoc******************************
THE END
- 4 NA 8004 0gg 0O f
(
r; esuuo
-~
SACRAMENTO MUNICIPAL UTILITY DISTRICT O 6201 S Street, Box 15830, Sacramento, California 95913; (916) 452-3211 Janua ry 16, 1978 l
'2 D' m',NED %
UpE84 Director of Nuclear Reactor Regulation c,3, W
U. S. Nuclear Regulatory Commission Washington, D. C. 20555
)
ATTENTION:
Karl R. Gol le -
\\.
Assistant Director for Operating Reactors
't'g Division of Operating Reactors RE:
Docket No. 50-312 Rancho Seco Nuclear Generating Station Unit No. 1 Gen tlemen:
The letter to you on October 8,1977 concerning the telephone conversation with L. Olshan, J. Strosnider, L. Shao and R. Steward stated that the District would forward to you the results of the Once Through Steam Generator tube examinations. The metallurgical examinations of the tubesheet samples from the Duke Power Company's Oconee Units I and 11 has been completed, and the summary report is included.
This should close out all items of disct.ssion pertaining to our commitment letter.
I f you have any questions, please contact me.
Sincer y yours, k.
Tor
'1 J. J. Matti e
Assistant General Manager and Chief Engineer JJM:RWC:sc
Attachment:
Evaluation of Distorted Eddy Current Figures at Rancho Seco W%
Acel S*
,. n c,, ~ ~_ _,
O n,
- s. a AN EVALUATION OF DISTORTED EDDY CURRENT SIGNALS OBSERVED AT RANCHO SECO Summary A recent recomendation to stabilize OTSG outer peripheral tubes near the open tube lane with distorted eddy current signals at the 15th support plate and/or upper tubesheet locations is reviewed in light of results from metallurgical examinations of two tubes with similar eddy current signals.
The examination results revealed some small loss of tube wall which was not predicted by our earlier understanding of eddy current analysis.
Nevertheless, it is shown that the recomendation to stabilize tubes, based on the recognition at Rancho Seco of a characteristic eddy current signal distortion already associated with tube failures at Oconee, is still appropriate.
Introduction During the September,1977, eddy current examination of Rancho Seco's Unit No.1, four distorted eddy current signals were observed in the "A" 0TSG.
These signals were termed " ding-like" eddy current indications since computer analysis of the raw signal yielded a resultant signal simi-lar to that produced in the laboratory by an intentionally introduced " ding".
The signals were distorted in the sense that although they were not similar i
to eddy current signals indicative of tube thinning or loss of wall, they were also not the signal which has been associated with a normal, cndamaged tube.
B&W recommended that three of the tubes with distorted signals be stabilized based on correlation between the shape of the distorted signal, the location of the signal in the OTSG, and experience with similar eddy current signals at Oconee.
_1
n?
r B&W concluded that this characteristic eddy current signal distortion, while not indicative of tube thinning or degradation of structural inte-grity, may be a precursor to tube failure by some as yet unknown failure mechanism. Thus the recomendation to stabilize tubes was based on a conservative oosition with reprd to future plant availability.
In an effort to determine the cause of distorted eddy current signals, two tubes removed from Oconee OTSG's were subjected to examination at B&W's Lynchburg Research Center.
Tube 77-25, removed from Oconee Unit I-B, had a highly distorted signal at the upper tubesheet (UTS) location, and tube 77-27, removed from Oconee Unit II-8, had a slightly distorted UTS signal.
Radiographic examination of tube 77-25 revealed a circumferentially oriented shallow serpentine deoression on the outside tube surface, slightly below the lower surface of the upper tubesheet.
No such defect was indicated by radiographic examination of tube 77-27.
Detailed results of a metallur-
)
gical examination of the tubes, which included scanning electron microscopy i
and metallography, are presented in the attached report.
In what follows, the recommendation to stabilize OTSG tubes with characteristic signal a,stortions is reviewed in light of the metallurgical i
examination results.
First, a refined explanation of distorted eddy current
.i signals is presented to help explain the basis for the recommendation.
This explanation, in which four types of distorted signals are delineaud, pro-vides a means of categorizing various distorted signals which in the past have all been referred to as " ding-like" indications.
The criterion used to select tubes to be stabilized is then developed, followed by a discussion '
O O
of the recomendation made to SMUD.
Results of the metallurgical examination are then sumarized and analyzed with respect to expected eddy current findings.
The last two paragraphs contain con.cluding remarks.
_D_istorted and " Ding-like" Eddy Current Sinnals Distorted eddy current signals have been detected at upper tubesheet locations during in-service eddy current inspections of the Oconee steam generators.
Sketches of typical normal and distorted signatures (for a 400 KHz examination) are shown in Figure 1.
Theraw(unreduced) eddy current signals are affected by material surrounding the tube in the vicinity of the probe; e.g., tubesheets and support plates.
The normal, undistorted signal for an undamaged tube, as illustrated in Figure 1(a), is produced as the probe passes the lower surface of the upper tubesheet in the direc-tion shown.
A similar signal would be generated at the lower surface of.
a support plate;at the upper surface of a support plate the signal would be similar but to the left of the vertical axis.
In a similar fashion, all the sketches in Figure 1 apply to both upper tubesheet and support plate signals, although they are labeled UTS signals.
To subtract out the influence of a tubesheet or support plate on an eddy current signal, distorted raw-signals are reduced by computer analysis such that only a resultant signal representative of the distortion is dis-played. The four types of distorted signals shown in Figure 1(b), from the slightly distorted Type 1 signal through the highly distorted Type 4 signal, all reduce to the same characteristic resultant signal lying along the hori-zontal axis.
T'his computer analyzed resultant signal indicates no loss of tube wall since such a signal would fall within the sector defined by the dashed lines of Figure 1(b).
It does, however, resemble the signal I
obtained in the laboratory from a " ding" or " crimp" which is defined as I _
p v -
q
\\
l a minor displacement of the tube wall toward the tube axis without a reduc-tion of wall thickness. This distortion of the tube wall is termed a " ding" if it is localized on a small part of the circumference and a " crimp" if it is mostly or all the way around the circumference.
The analyzed resultant signal in Figure 1(b) is not identical to the
" ding" or "cririip" signal, but is more like it than any other identifiable signal.
Therefore it has been tenned " ding-like".
It is believed that l
material cold working, and perhaps other factors, influence the distorted eddy current signals.
It is emphasized that the resultant " ding-like" signal does not indicate the presence of loss of metal and that it has not been used as a basis for stabilizing tubes.
Distorted Eddy Current Sicnals as a Basis for Stabilizina Tubes Prior to the inspection at Rancho Seco, Type 4 raw eddy current signals had been noted at Oconee at the upper tubesheet and 15th support pl. ate loca-l tions along the open tube lane.
It is convenient to further classify these Type 4 distorted signals according to their location along the open lane.
Lane row tubes 10-30 define a reference zone of tubes where failures at i
Oconee have been co'.icentrated.
It is not unusual to obtain Type 4 signals in these tubes, and in fact, three such tubes had these signals before they failed (75-12 and 77-15 in Oconee I-B, and 77-19 in Oconee III-B).
It is also significant that Type 1, 2 and 3 signals were noted in lane
~
tubes in this zone at the time of the above failures.
The second class of upper tubesheet/15th support plate, lane row, Type 4 raw eddy current signals are those found near the periphery (tubes 1 or 2).
Type 4 signals in this region are not comon, as evidenced by Oconee's operating history. They also seem to develop more rapidly here l-p-
q than in the reference zone of lane row tubes.
Four Type 4 signals have been noted in the outer periphery of the Oconee III-B generator, in tubes 77-2, 78-1, 75-2, and 78-2. Two of these tubes failed, 77-2 and 78-1.
(All four were eventually removed from service.)
Furthermore, a scattering of less distorted signals in the region at the time of tube failures, as noted in the reference zone, was not noticed in the periphery.
It, therefore, seems that a Type 4 raw signal distortion in the upper span region of a peripheral lane tube is indicative of an increased probability of tube failure, more so than in any other region of the generator, and for this reason is used as a basis for stabilizing tubes.
Although it is not yet fully understood what causes the signal distor-tion, it appears that the consequences of having a Type 4 distorted signal causing flaw depends on environmental factors which vary throughout the gen-erator.
Two such factors could be radial steam velocities and moisture in the upper span (between the 15th support plate and upper tubesheet).
If steam velocity does influence the growth of flaws, it is reasonable to assume that lane tubes would be more severely affected by a flaw than bundle tubes due to higher radial velocitier.
'arthermore, as steam velocity is proportional to radial position along the lane, peripheral lane tubes would be the most f
likely candidates for failure.
The accumulation of moisture in the open lane would also create adverse environmental conditions for lane row tubes.
The combined effects of moisture and increased velocities in the lane could be i
particularly damaging to flawed peripheral lane row tubes.
Recommendation from B&W In September of 1977, four distorted eddy-current signals were reported at SMUD, in tubes 75-2, 70-2, 74-2, and 66-130. All signals were obtained at
(",
r, o
the 15th support plate and all were found to be Type 4 distorted signals.
Based on the increased probability of failure noted for Oconee tubes in the peripheral lane area having Type 4 signal distortions, tubes 75-2, 70-2, and 74-2 were recomended for stabilizing.
Tube 66-130, located in a region where no tubes have failed with Type 4 signal distortions, was not included in the recommendation.
_ Metallurgical Examination At the time of the recommendation to SMUD concerning distorted eddy-current signals, it was decided to metallurgically examine via destructive testing two available Oconee tube samples in an effort to characterize these signal distor-tions. Tube 77-25 (0conee I-B) had been removed because of a Type 4 UTS signal distortion. Tube 77-27 (0conee II-B) had been removed because of visually indi-cated wear on the outer surface of the tube at the 15th support plate.
Ic was discovered during a reexamination of the eddy-current inspection' records, after the tube had been removed, that tube 77-27 had had a Type 1 UTS signal distortion.
Attached to this letter is the laboratory report of the metallurgical examination.
Since tube 77-27 revealed a Type 1 distorted UTS signal, the results of the examination of this tube sample are not particularly rt.ievant to SMUD's problem concerning distorted eddy-current signals.
It appears that superficial surface damage possibly associated with localized plastic strain was responsible for the distorted signal.
Examination of the tube 77-25 sample revealed a shallow (2 mils deep) depres-sion in the outside surface that would have been located slightly below the lower surface of the upper tubesheet.
This depression took the shape of a serpentine curve and extended about 120 around the tube in the circumferential direction.
Within the depressed surface were found transgranular circumferential microcracks t
4 J
' penetrating 6 mils of t tube wall. Th2 individual crb.s were approximately 5 mils long and were at the tube surface.
These microcracks resemble what would be expected for Stage 1 fatigue.
It was anticipated that examination of the tube samples would show no reduction of wall thickness, based on the original interpretation of distorted signals.
But the Type 4 distorted signal for tube 77-25 was found in an area with a shallow OD depression (and no inward displacement at the ID).
It is felt that the microcracks in the 77-25 tube sample are too small to be detected by the eddy-current probe. Assuming such to be the case, this leaves only a 2 mil defect to be picked up by eddy-current analysis, which cannot accurately indi-cate such a small flaw.
It is concluded that some effect other than microcrack-ing or tube thinning produced the distorted eddy-current signal, such as localized cold-working.
Conclusions Results of metallurgical examination of a tube sample exhibiting a Type 4 distorted eddy-current signal do not conclusively reveal the cause of the signal.
On the contrary, they suggest that changes in the material introduced by such effects as cold-working can produce signals similar to those caused by tube wall distortions without metal loss.
In fact, from the results of the metallurgical examination, it cannot be determined, using what we now know about eddy-current signals, how the Type 4 signal distortion was generated.
Microcracking of the tube wall is thought to be more a possible indication of an eventual failure mechanism (fatigue) than a source of the original distorted eddy-current signal.
Whereas results of the metallurgical examination tend to weaken our under-standing of distorted eddy-current signals in that a multiplicity of causes now seems probable, they do not force us to alter our previous recommendation to.
(
n stabilize outer peripheral lane tubes with Typ] 4 signal distortions at UTS and/or 15th support plate locations.
The tube stabilizing criterion is based on correlation between Type 4 raw eddy-current signal distortions and Oconee tube failures.
It is postulated that the tube wall distortions and/or material changes which may be responsible for distorted signals are more damaging if they occur in the upper span of the outer peripheral lane tubes because of en -
vironmental conditions.
Based on Oconee experience, the presence of distorted eddy-current signals in other regions of the OTSG is not considered to be significant at this time.
1 l
l l
l AO
~
)
NORMAL UTS
^
SIGtML (400 KHz) h-l h,I, / I I ' l / fl
/
/
UTS
( /
- ll
/fI /
l 7
EDDY CURRENT TUBE L
PROBE
~
FIGURE 1(a) fl0R!!AL UTS EDDY CURREtlT SIGNAT'JRE
~
~~~
RAW UTS SIGNALS ICOMPUTER ANALYZED RESULTA!!T WITil DISTORTIONS (i.e.SIGNALCAUSIt!GDISTORTIOi!)
A'NGLE SECTOR FOR TYPE 1 MILD ID AND OD FLAWS N.
~
]
QUE "-
~
x- [
7 p
f
~
\\
TYPE 2
)
s
/
- t: ;:
AA-f
\\
~
N TYPE 3
%)
TYPE 4 UORST CASE w.
J FIGURE 1(b) - REPRESEllTATIVE UTS SIGilAL DISTORTIONS (0C0f!EE)
.g.
.