ML020850200
| ML020850200 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 02/26/2002 |
| From: | Thadani A Office of Nuclear Regulatory Research |
| To: | Collins S Office of Nuclear Reactor Regulation |
| References | |
| FOIA/PA-2001-0256 | |
| Download: ML020850200 (8) | |
Text
-,
NV
"'I 4"W*
MEMORANDUM TO: Samuel J. Collins, Director Office of Nuclear Reactor Regulatio FROM:
Ashok C. Thadani, Director /
of u ear Regulator
SUBJECT:
RE EST FORW IR.EPENDEN EVALUATION REGARDING ST INTERVAL AND FEBRUARY.13, REGRADINGF* RP IRCRCI IA Thismemoran urn is in r
- onse to your inde'andent review of safety evaluations
!ik*
for tha Indian Point Station, Unit*
J..
'ie.ted q review of these issues bas i
ed our review to &5 YoU stated thai the purpose of ndepe conclusions*Are technicallyb and thi ayed ir 0robak
/a F MAY 26,*IMAFETY
,TOR TUBE INSPECTION "Y EVALUATION ra um otjebruary U000, requesting an
- arding It "
iffiherato. ube inspection and repair Staff in ise ni",gineering Technology, REJ d oni rbal r stfrom your staff on February 18, e the
- criteria, sed on your memorandum.
revie as to =determine if the staff's la a. resented by the licensee provided p nd the use of the F* repair criteria would not tube failure prior to the next scheduled taffs'sfy Evaluation of May 26, 1999, and other written it ev a*tion. in performing our review, we addressed the specific
§$ inspection interval with the assumption that the original 0and then evaluated the technical basis for the original inteival.
provided in the attachment to this memorandum.
us 4 the F* repair criteria, we did not identify any issues related to the staff's 4ofmation submitted by the licensee. The evaluation and the information Densee do provide reasonable assurance that the use of the F* repair criteria an appreciably increased probability of tube failure prior to the next inspection With regard to tht extendedinspection interval, working from the assumption that the original inspection interval was justified, we concur that the licensee's lay-up procedures for the steam generators were appropriate, and granting the requested 48 day extension of the inspection terval would nct have appreciably increased the probability of tube failure.
U
.P 14,I:'
v int
Samuel J. Collins However, in our review of the original inspection interval for cycle 14, we cannot reconcile several statements and conclusions in the SE with the RAI and the information we reviewed, particularly with respect to the operational assessments conducted for stress corrosion cracking in the second row U-bend region and at the top of the tubesheet under the sludge. In its review of the licensee request, the NRR staff recognized the. importance for maintaining required tube structural and leakage integrity for the entire cycle 14, and in a requestfor addi information, posed the following question (question 1):.LF]or each degraddat echanism, please provide a general description of the operational assessment Aim-eth ogy used to ensure that SG tube integrity will be maintained for the entire fue.?
(cyc&1 )
edescription should include an explanation of the predictive methodo
,flIaw grov t
ad NDE uncertainty used to determine structural and accident Iage inte.t1 2
We find the lioneee's response to the staff's question incomplete.F le the licensee provided ony a-very so-sSn-rega ion"alassessment for stress corrosion cracking at the row 2 U-bend. No predictiv'
'I f a discussed nor were growth rates or NDE uncertainty applied in their evalui i n...
e s e1 simply stated that the indication was below the in-situ screening thresholdQ smal n !]sthis represented the first detected U-bend indication after approximateor f
yfr any growtl'ratW l.
associated with this indication would be consided i(mijN While)'e detailed discussions regarding the weakness of the analyses c&te, i~nsee-Ae included in the attachment, we disagree with the licens*"s onte oscaus, inconsistent with the evolution of stress corrosion crackin go with otr i-xperience. Contrary to our findings, the SE indicates that thep
- 4"pee dcontwcted mo thorough operational assessments than we have identified and conc that t*h ubes w d meet structural and leakage integrity through the end of operating cyc
..,4 Based ontei 3in eiwd Based o
n d
thegr nfo mavationi ave.re.;view
.ieve the licensee's assessment of two forms of degrai%,frd in enertors was I adequate: (1) ODSCC above the top of the -
tubesheel oJ*udge pil
- d; 2) PWSCC at a row 2 U-bend. We believe that a proper operational a r thesi-e f degradation might have revealed an increased Th-&..
probability oo tu 1r rupt"ure
'the end of cycle 14 based on the findings and actions Z
taken at the If yo our staff rther discuss our findings please let us know. For additional tec cal information re rdipýg this review, please contact Dr. Joseph Muscara of my staff.
2
ATTACHMENT REVIEW OF SAFETY EVALUATIONS REGARDING STEAM GENERATOR TUBE INSPECTION INTERVAL AND F* CRITERIA FOR INDIAN POINT STATION 2 Inspection Interval Evaluation The RES evaluation is based on review of the following documentation (1)
The May 26, 1999 Safety Evaluation; (2)
The original licensee submittal dated December 795 (3)
The licensee response dated May 12, 1999 to thAJRR requeian 99 gto atorI i
nfr (RAI);
(4)
The licensee report dated July 29, 1997, of theIsteam g rator tube i examination conducted during the 1997 refueli The licensee was effectively requesting a one time exten-i-s eanm generator inspection interval from June.1999 to June 2000. Upon return to irvice -.
e 1997 refueling outage, Indian Point 2 (!P2) was shut down on October 25, 19 i
nscheduled maintenance outage that lasted 304 days. in
'6leuse of tthe plant w ut down, the licensee was requesting an exten eoi48 da Because the licensee followed industry guidelines aintai4.6 t layp chemistry to minimize corrosion of the generators during the o4 ge, anegrt'i a
would have occurred during this period would have been negligible1urther e licen a conducted an extensive inspection program during the 1 u e' e ng age. T hfore, if the issue is reduced to an assessment of whether the additUona148 dag f*foperatiU would significantly adversely affect the integrity of the steam gener r, given-at the ered integrity is maintained during the 24-month ccIle of operatio.'::
'would co a no appreciable increase in the probability of tube fail e&ould res*
In its revieWk f theensee t
NRR staff recognized the importance for maintaining required tub ct Al ndlage in enty for the entire fuel cycle 14. In this context, a request for addi n
atio wasssued with two of four questions relating to tube structural integrity.
i l]orea degradation mechanism, please provide a general A
descriplin of the ope tiiat sessment methodology used to ensure that SG tube integrity Sr will beaWaintained for t&At
)uel cycle (cycle 14). The description should include an explaNation of the predict=j viethodology, flaw growth rates, and NDE uncertainty used to aPmine structural a accident leakage integrity.- In discussing the licensee's steam f
egerator tube integr 3ssessment for the eight forms of degradation that were detected at the Eh~dl.fjuel cycle 1 ~~he SE states that "[Tlhe licensee's evaluation determined that the forms of dation istebove did not present a challenge to the 3&P structural margin criteria for the
ý99cte it'g cycle length of 21.4 effective full power months (EFPM). Based on a review o
spoftion of the licensee's assessment the staff expects the steam generator tubes will continue to satisfy structural and leakage integrity requirements under normal and accident conditions through the end of the current operating cycle (cycle14).7 4t,
ýegarding the licensee's operational assessment in general, RES found it to be incomplete and the arguments presented to be weak. For most of the degradation mechanisms addressed, the operational assessment was more of a condition monitoring evaluation. The condition at the d0 a 0A,
2 end of cycle 14 was assumed to be similar to the condition at the end of cylcle 13.Since the structural and leak integrity were met at the end of cycle 13, the licensee concluded they would also be met at the end of cycle 14.
However, the behavior of stress corrosion cracks is expected to differ from one op;eating cycle to the next especially when the cracks first initiate or are detected. The appear e of a 'first' stress corrosion crack typically indicates that an incubation phase has
- i nd that more cracks are likely. Further, in the relatively early stages of *c!k growt*!wth rate is dependent on crack size and loading. For the relatively 9 t loa di.fgol:m generator tubes, this means that as the crack size increases, the Wth rate w1 ere will transition from thisincreasing growth rate to a more co nt a
ro9ae a, s
larger. However, given the first indication of stress co ion crking i stenal\\,
gor
- bes, the physics of the process and service experience sug s both the numberýfý cks and their rate of growth will increase. Thus it cannot be exp the number and sizes of Cracks, for the degradation mechanisms first 3, would be the same at the edof cycle 14.
ietfe 4~
,9 "There were two aspects of the licensee's May 121749W. response I related tothe operational assessment that RES considers d iW antedf e
nav4!
er evaluation. These were stress corrosion cracking above the fkf tubS r t ludge pile, and primary water stress corrosion cracking at the r
.U-be (1) ODSCC Above Top of Tubesh l
9"'
The licensee reported that ODS -
in the u e pile s detected for the first time in the 1997
,4$
inspection, and that 22 ind'd ion of this weected. The licensee contended that the 0
bounding g rat acks wa s
' 40 to 50 percent throughwall cracks that
)o
\\r*_*,,(
i might n e
e M
the in cion would still meet the integrity requirements at "the end of a S o owing discussion, RES concludes that this contentio*)js not credible.
- MXOM
.vu j
j A
The limiting-jd~
o s p10 wa identified as having a maximum depth of 69%, average depth of 48o, anda f.55 inch. The tube With this indication was inspected in 1995 with laecco-5 probe ndication was detected at that time. The licensee reports that the growth in average dtir cycle 13 is bounded by about 18% to 28% for sludge pile ORSCC indications. T jp was determined by assuming that the indication was 20% to 30%
trough wall at the be fnning of cycle 13. But the tube with this indication was inspected at the MOD cycle 12 and indications were detected. Therefore, another plausible assumption is crack~stae to grow in cycle 13, either at the beginning of the cycle or even later in the CI Ina~o, the licensee assumed that the +Point depth profile was accurate, i.e., no NDE sizing ainty was applied to the detected crack size even after the licensee has stated that "Recent +Point depth sizing evaluations performed by Westinghouse for axial ODSCC indicate that flaw average depth standard deviation measurement error is about 10% through-wall."
Certainly, assuming that the crack was 20 or 30% through-wall at the beginning of the cycle and not allowing for inspection sizing error, did not provide a bounding estimate of the crack growth rate. If the crack had started to grow at the beginning of cycle 13 and a one standard deviation
ivy Ddo d_________
L~b~nLD(AIDAA.
VAP." 4 ti s4ing errorh befen applied to the detected crack, then the growth in average depth would have been 58% for the cycle. The licensee did not discuss the growth for the maximum depth of the crack which was 69% at the end of the cycle. The licensee stated that "the modest growth would lead to acceptable end-of-cycle (EOC) structural integrity even if 40% to 50% average depth indications were not detected." However, if one applies the higher growth rae (58% for one cycle) that is obtained assuming that the crack had initiated at the
" "innin
-pi. cycle 13 and.
making some adjustment for sizing error, then the-undetected cracks v
ge depth indications of 40 to 50% would penetrate through-wall* durinone op.
ra cle, and potentially not meet the structural integrity requirements a end o
if these cracks with average depths of 40 to 50% have si Irn m holo Y "Mchk found during the inspection, Le., the maximum depth is 21% p later thai ave a
e growth during the cycle is added to the maximum dep hen th acks wou ;r3 h
wall during the cycle and the tubes would leak even if',
rate of 28% is apI'ias estimated by the licensee. Further, if the cracks maintai nSpect ratio similar to the limiting crack analyzed by the licensee, the crack length whe i t
rates the wall could be close to the critical length under steam line break pressure reasing the probability of tube rupture under these conditions.
The licensee stated that "[W]hile ODSCC in), ;tluidgile region ishew mechanism at Indian Point 2, the 22 indications detected repre1V e
to170 ffý ittal tub, population. Therefore, based upon the observed sludge fla fiddy c nt en
&)ii at IP-2, and in-situ testin*
results from more limiting flaws at iipOants n
ued that this corrosion Smechanism would not re Ibreensq represent burs steam break leakage potential at EOC 1.1 This implies that the condition of igenerat*
vith resj to this cracking phenomenon will be*
similar at the end of cycle 14 to atthen of cycl
- 3. The fact that the licensee detected 22 ODSCCsfn the sludge i
cated tatti tion period for this phenomenon had been rea!'Ve.
d that in numbe s could now initiate and grow during subseqenoi R6,nsee I
conduct a thorough operational assessment with respeCt!oes gliUstribution at the beginning of cycle 14, i.e., the cracks left in the generatl bse they Weodtected by NDE. They did not determine the number of new cracks th*w'dia ngtne cycle; this number would likely be greater than was expen6enc.
gh*
ou since the phenomenon was still relatively new at IP-2.
They did:..ot apply cr g
rates to the undetected cracks and the newly initiated cracks so that th could estimate thecc distribution at the end of cycle 14. Therefore there was not a Wbasis f.o a
goo bs for estimatin tructural and leak integrity at the end of cycle 14.
SCCatrow 2 nd -The stress corrosion cracking process involves two separate an initiation
.ncubation period, and a.growth period. Once cracks initiate, the growth ts.cracks in tubes that take either a short time or long time to initiate. The crac grCowt s can be quite high for U-bend regions because of high residual stresses and dya Nsn caused by either or both fabrication and the tube denting process during operation.
The licensee cites that PWSCC at the row two U-bend was detected for the first time in the June 1997 inspection. The licensee further states that "[A]s this represents the first detected U-bend indication after 23 years of operation, any growth rates associated with this indication would be
.5
considered minimal." Based on the stress corrosion cracking process, this conclusion is not Scredible.
GJ Aa t
AP 6"
NA C%&
A AA4 The detection of the first row 2 U-bend crack at IP2 was an important finding in that it indicated that the incubation period for crack initiation had been reached, and now the cracscould begin to appear and grow. Further, in addition to the residual stresses preseL t
from t h#&brication of the tube, inspection results for IP2 have shown the tubes to be locked ni-e pport plates by the denting that has occurred at this plant. The 1997 inspecion sho "als everal tubes at the upper support plate, including row 2 tubes, were lock*
"he su I a as evidenced h the 610 mril or 640 mil diameter probe not being able to) a throughe eiher, o both, the cold leg side or the hot leg side at the uppers port ptevati he U-bend tubes are locked in the upper support plate, t s o tube begin o4,
&6.er
- together as the denting process continues, the supo ndbcrack Jte flow slots begin to hourglass. The motion of the U-bend tut M ses ovalization and operation induced straining of the upper portion of the tube at the di."
straining leaves the tube egion highy susceptible to stress corrosion cracking The 1997 inspection also found evidence that ten eforred by the
'denting process due to the inability of the 61 pas-to econdary side inspection (as reported in ' cen s
tio port) of the upper support.
plate in 1997 also found so mall cr s inthe m rten a reviousli observed.
Leakage from s e corrosion cracki ight end lo shas occurred in operating plants, inclu ding two cases of tube
-r U-be
. Some licensees have preventively plugged rows of tight-radius U-beu rbes i ir stea enerators before placing the generators in.service, during sece, or u detecti f the first crack(s) to avoid stress on crackin inden ring serv ocations.
""Te an serva as s d ppea to be in on the licensee's assessn taf's on.
In eval iting the F* crit iýp ved for IP-2 in 1995, RES reviewed the 1995 SE and the Dec i'ber.24, 1994 lice response to an NRR RAI. F* is a repair criterion that allows defects toTe specified n ce (the F* distance) below the end of the roll transition region in the t eet of the SG.
r proper implementation, the F* distance must be shown to be sufficient J
st operational.
d transient pul-out forces on the tube; and maintain primary to secondary aiJn ac o nce with the plant technical specifications. The minimum F* distance is Dseo-on consideration of the shear stress developed at the tube-tubesheet interface, thý'
ntact, and the coefficient of friction between the tube and tubesheet. The licensee provided calculations, and results of tests on mock-up tube-tubesheet assemblies to validate the calculations. The mock-up test conditions reasonably simulated the conditions that would be expected in the SGs (e.g., variations in tube yield strengths, variations in tubesheet bore surface roughness and diameter). The minimum calculated F* distance was increased to account for the limited sample size in the testing, statistical scatter in the data, and for NDE uncertainty.
The evaluation and the information submitted by the licensee do provide reasonable assurance
5 9
that the use of the F* repair criteria would not result in an appreciably increased probability of tube failure prior to the next inspection interval.
V.,
4
has not ACCentuated a PWSCC concern, and may be attributed In pan to the relatively low Tom value of thd Unit.
ODSMAI at Dented Tube 1=U 11c*
PlAte nereetipnI totl of eleven cold leg dented tube support plate Intersections were Identified with the Cecco 5* po be, "00.. 161i Ind.ctionj, all were +Point lnspecew*o0. wereound to contain ODSCC dc n
onlefts rip+wld o n a volumetric lnditios baid on the +Polnt b eiul e
~ 9~esanes A wt te otle ltesetonj 0,10b6spor lae would be expected to remain 4a"c+en tofthe W g
I
~~~~~~~~N ajantote dcaonthreby p Wcuing burst durlnit posrtulated steam line breaki event.
y onefteseentdnt to estricted passage th 0,640 Inch diameter Cecco.S po be 'du iitoar tr i ctio a bt a tower elevatio-n n pls' te.1 "Non ofthe Indlcations at dented
"++
e support pliat (ei I o r Inated)
IT e 14 the +Polnt wexended out of the plate 0
'whileelongest Indicaion was repod at 0.37 inch.
PWSCC at Row 2 Ut.he For the first time, a Row 2 U-bend PWSCC Indication was found. The dimension ofthe I1ndIcion by *Polnt ehaiatiwertla°n was below the Inifltu _creenins threshold for Row 2 U.
ben,..aw.. All Row I tubes wer prevenively pluiged ir to operaion for I
notube
.at Cw-~~~~~~~-~~_ [sue Athsrrenste rtdectodIjbd Oldctt.n aftir aproimately 23yerof op
- ation, ay growh rw a usloclated with thi andisato w160od ee onidereminimal.
Please discussthe results of!you condition monitoring ssessment conducted during our most "recnt" Insp*e bci.
- nlue, Ih eaton Mec se mvled using the Westng u
I arlSW cMsw evaluated using the r i criia? What assuranc is provided that dhe structural Integritj wouldke malntaned?
B-All of the above listed mechanislms were evaluated to the WestInghouise screening parameters.
with the exceptin s,.ludge pile pining (pinun above top o1 tube et) The pit indications were no asssedagaitthe West scening criterla nee icreening criteria s
o uded. Ai discussed aboi v
.ep Itndications generally do not rpresent structura"or ileake rtissues
.ad as lniiti testing ofsuch Indications wil. not provide addbloliat, W 0 o
t for lue g
ement. Pit Indications during the ouag as ot n.s.u tesIngb osd io laxmuml bobbin coil depth of 50% and voltage of3volta No ind1catln met this ri"te Disusionsfl the.arahe used Ibf each mechanism. Including s lud le ple piIng, ArM provided As pr of the Question I response.
Indmlc
.I dte4