ML20196J497
| ML20196J497 | |
| Person / Time | |
|---|---|
| Site: | Byron, Braidwood, Zion File:ZionSolutions icon.png |
| Issue date: | 07/30/1997 |
| From: | Hosmer J COMMONWEALTH EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GL-97-01, GL-97-1, NUDOCS 9708040212 | |
| Download: ML20196J497 (9) | |
Text
I l
Commonwealth Edison Comhuny 1400 Opus Place Downers Grove. II. 60515-570 t July 30,1997 U.S. Wclear Regulatory Commission W.ashington, DC 20555 Attention:
Document Control Desk
Subject:
Braidwood Station Units I and 2 Byron Station Units 1 and 2 Zion Station Units 1 and 2 NRC Dockets Numbers: 50-456 and 50-457 NRC Dockets Numbers: 50-454 and 50-455 NRC Dockets Numbers: 50-295 and 50-304 Commonwealth Edison Company (Comed) Response to NRC Generic Letter 97-01, " Degradation of Control Rod Drive Mcchanism Nonje and Other Vessel Closure Head Penetrations" dated April 1,1997
References:
1.
NRC Generic Letter 97-01, " Degradation of Control Rod Drive Mechanism Non.le and Other Vessel Closure Head Penetrations," dated April 1,1997.
2.
J. Hosmer letter to the Nuc! car Regulatory Commission dated April 29,1997, transmitting Comed's 30 Day Response to GL 97-01 In Reference 1, the Nuclear Regulatory Commission (NRC) transmitted Generic Letter 97-01,
" Degradation of Control Rod Drive Mechanism Nov.le and Other Vessel Closure Head Penetrations."
Reference 2 transmitted the Commonwealth Edison Company's (Comed's) 30 day respoasc to the Generic Letter. In that letter, Comed stated that they would provide the Staff with an appropriate response and the requested information to the extent practicable within 120 days of the date of the Generic Letter. Comed's response is provided as an enclosure to this letter.
To the best of my knowledge and belief, the statements contained in this document are true and correct.
If you have any questions concerning this correspondence, please contact this office.
IACEEE l Sincerciv, MARY JO YACK
/$$NsN((PR',
John B. Hosuer b-^^^^~~~~~'''''''''''''"'~~~~
Engineering Vice President 7 9 ') m m R M y.do -97 tw,
Comed Response to Generic Letter 97-01 for Braidwood,/
Byron and Ziot)
Enclosure:
Attachments'.
8, (A through H) cc:
G. Dick, Byron /Braidwood Project Manager - NRR C. Shiraki, Zion Project Manager - NRR C. Phillips, Senior Resident inspector - Braidwood S. Burgess, Senior Resident Inspector - Byron iI A. Vogel, Senior Resident inspector - Zion A.B. Beach, Regional Administrator - Region 111 7h OfTsce of Nuclear Safety -IDNS M
k:nla:bybwd stmgen:gl9701 As i
{ll(({\\.\\(({\\\\{\\}l((\\!(({\\k,kh 2 970 2 PDR ADOCK 05000295 P
l PDR A tJnicom company
Enclosure Comed Response to Generic Letter 97-01 for Braidwood. Byron. and Zion Introduction Generic Letter 97-01 (GL 97-01), " Degradation of Control Rod Drive Mechanism Nozzle and Other l
Vessel Closure Head Penetrations," was issued to request licensees to describe their program for insuring timely inspection of pressurized water reactor (PWR) control rod drive mechanism (CRDM) nozzles and I
other vessel closure head penetrations (VHPs). This response provides information relative to the Generic Letter request for Braidwood Units 1 and 2, Byron Units 1 and 2, and Zion Units 1 and 2.
Prior toissuanc
- he Generic Letter, Comed worked with the Westinghouse Owners Group (WOG), the t
Electric Power Research Institute (EFRI), and the Nuclear Energy Institute (NEI) to understand the applicable operational experience and identify appropriate technical issues, causal factors, their relative importance, and potential solutions. Attachment A provides a thorough summary of the industry approach to management of the issue. Comed participated in the development of this NEI white paper, which provides a listing of all WOG reports which have been submitted for resiew by NRC on this issue.
One of the most important WOG deliverables was the development of a safety evaluation that characterized the initiation of damage, crack growth, and the structural consequences of cracking (Attachment B). This safety evaluation, including an evaluation in accordance with 10 CFR 50.59 which concludes that no unresiewed safety question exists, is applicable to Braidwood Units I and 2, Byron Units I and 2, and Zion Units 1 and 2. The NRC resiewed this safety evaluation and issued a Safety Evaluation Report (SER) to NEl on November 19,1993 (Attachment C).
j Note that the documents referenced in this enclosure are listed in a separate Reference section near the end of the enclosure.
Comed RCsDonse In Generic Letter 97-01, the NRC requested addressecs to proside a written rer - hat includes the following information:
1.
Regarding inspection activities:
1.1 A description of all inspections of CRDM no :le and other 171Ps performed to the date of this generic letter, including the results ofthese inspections.
F:>otnote:
Those licensees that have previously submitted the requested information need not resubmit it, but may instead reference the appropriate correspondence in their response to this Generic Letter.
l Restens i All Comed PWR units perform visual examinations for boric acid leakage in accordance with the Comed response (May 31,1988 W.E. Morgan to A.B. Davis letter) to Generic Letter 88-05 (GL 88-05) at every refueling outage in accordance with applicable station procedures. No esidence ofleakage attributable to cracking of CRDM nozzles or other VHPs has been found to date, although other sources ofleakage from canopy seal welds and mechanical connections have been observed and appropriately repaired. 1 l
l l
l t
b
l l
Enclosure All Comed PWR units, except Braidwood Umt 2, have performed penetrant (PT) examinations from the outside diameter of the CRDM nonle-to-CRDM head adapter dissimilar metal weld region in accordance with the applicable station Insenice Inspection (ISI) Program. No indications have been obsen'ed.
Braidwood Unit 2 will perform this examination in the Fall of 1997.
Braidword Units 1 and 2 and Byron Units I and 2 have elected to perform visual examinations of accessible areas of the closure head interior, including CRDM nozzle and other VHP attachment welds, and have incorporated tais examination into the applicable station ISI Program. No indications have been observed.
1.2 Ifa plan has been developed to periodically inspect the CRDAf nonle and other T7IPs:
a.
rovide the scheduleforfirst, andsubsequent, inspections ofthe CRDAf no=le and other 171Ps, including the technical basisfor this schedule.
Response
j All Comed PWRs will continue to perform boric acid leakage examinations as described in tne Comed response to GL 88-05 at every refueling outage in accordance with applicable station procedures as well as all other examinations requirst by the applicable station ISI Program. As noted in Attachment C, the performance of GL 88-05 visual examinations leads to the conclusion that cracking of the CRDM penetrations is "no immediate safety concern."
Comed has determined, through a probabilistic assessment of the potential for primary water stress corrosion cracking (PWSCC) in Comed CRDM nozzles as a function of plant life (described in the response to Question 1.4, below), that, to assure long term safe and economic operation of the Comed PWRs, the most appropriate course of action is to perform volumetric examinations of CRDM nozzles and other VHPs at Zion Unit I during refueling outage 17 (ZlR17), currently scheduled for the year 2001. If indications are found, they will be dispositioned and followup examinations performed in accordance with American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code Section XI and the flaw acceptance criteria accepted by the NRC in Attachment C. If no indications are found, Comed has further determined through a probabilistic assessment that followup examinations will be performed every 5 refueling cycles.
Since the probability of flaw initiation and growth in the Zion Unit 2, Braidwood Units I and 2, and Byron Units 1 and 2 CRDM nozzles and other VHPs has been determined to follow behind Zion Unit I by a significant margin, examinations of these units will not be scheduled until a periodic engineering assessment of all industry data, in the context of an integrated industry inspectian program with the participation of the three PWR Owners Groups, EPRI, and NEI, indicates that an examination is appropriate. When it is determined that examinations of CRDM nozzles and other VHPs at Zion Unit 2, Braidwood Units I and 2, and Byron Units 1 and 2 are appropriate, Comed will provide the schedule for these examinations in a supplement to this Generic Letter response The technical basis for this conclusion is described in the Comed resDonse to Question 1.4.
b.
Provide the scopefor the CRDAf nonle and other 171P inspections, including the total number ofpenetrations (and how many will be inspected), which penetrations have thermalsleeves, which are spares, and which are instrument or otherpenetrations.
The Zion Unit I closure head contains 53 sleeved CRDM nozzles,12 unsleeved spare nozzles,8 nozzles designated for part length CRDMs,5 unsleeved thermocouple nozzles, and I smaller unsleeved head vent nozzl:. The presence or absence of thermal sleeves in the nozzles for designated part length CRDMs will be determined prior to finalization of the exan ination scope. The CRDM nozzle and other VHP examinations are planned to be performed on the alloy 600 volume in the vicinity of the nozzle-to-closure l
head partial penetration weld, 2-
/
Enclosure t
as well as in the vicinity of the nozzle bimetallic weld, using eddy current (ECT) and ultrasonic examination (UT) as appropriate, from the bottom of the head. The number and location of nozzles to be-inspected will be determined no later than six months prior to the beginning of ZlR17. This scope of examination will be determined using a combination of engineering assessment and industry examination data available at the time, and will be provided to the NRC in a supplement to this Generic Letter response.
1.3 Ifa plan has np.t been developed to periodically inspect the CRDAf norte and other 17/Ps, provide the analysis that supports why no augmented inspeciton is necessary.
l l
See the response to Question 1.2a and 1.4.
i 1.4 In light ofthe degradation ofCRDAf noale and other 171Ps described above, provide the analysis that supports the selected course ofaction as listedin either 1.2 or 1.3, above. In j
particular, provide a description ofall relevant data and:or tests used to develop crack initiation and crack growth models, the methods and data used to validate these models, the plant-specific l
inputs to these models, and how these models substantiate the susceptibility evaluation. Also, if l
an integrated industry inspection pwgram is being relied on, provide a detailed description of this program.
l Comed has contracted with Dominion Engineering, Inc. (DEI) through EPRI to perform a probabilistic l
assessment of the potential for primary water stress corrosion cracking (PWSCC) in Comed CRDM j
nozzles as a function of plant hfe, as well as to perform economic strategic planning to develop an l
economical approach to dealing with the issue. The software package developed by DEI, referred to as l
CIRSE (CRDM Nozzle PWSCC Inspection and Repair Strategic Evaluation), uses a methodology which includes:
Weibull distribution ofindustry CRDM cracking data corrected for temperature, stress, material, and fabrication to predict crack initiation.
Power law stress intensity equation corrected for temperature with i
distnbuted growth rates to predict crack growth.
Monte Carlo analysis to handic variable-cracking susceptibility e
i l
population and distributed input parameters and to calculate l
probability of a crack or leak.
Calculate life-cycle cost for alternative strategic scenarios for l
inspection, repah, and remediation.
The results of this analysis (documented in Reference 1, which contains extensive Comed proprietary l
cconomic data, and summarized in Attachment D) indicate that a low (i c., less than or equal to 5%)
probability of a 75% throughwall flaw (a 75% throughwall flaw bei tg the commoviy utilized maximum allowable end-of-period depth-to-thickness ratio of AS.ME Section XI IWB-3600) 1er all CRDM nozzles will not be exceeded before the ZlR17 refueling outage. Further, the results of t'te pobabilistic analysis indicate that a subsequent five cycle examination period maintains a similarly low flaw probability.
1
(
In light of the available industry inspection data for CRDM nozzles described in Attachment H, Section i
1.3, and the structural margin assessment of the Attachment B safety evaluation, Comed considers the 5%
l probability of a 75% throughwall flaw to be a sufficiently conservative criteria to assure compliance with tne 10 CFR 50 Appendix A Criterion 14 requirement that there be "an extremely low probability of abnormal leakage, of rapidly propagating failure, and of gross rupture." The time to develop a 5%
probability of a 75% throughwall flaw for all CRDM nozzles in Zion Unit 2, Braidwood Units I and 2 l l
7
Enclosure 4
and Byron Units I and 2 is substantially beyond the year 2001, and examinations of these units will not bc xheduled until a periodic engineering assessment of all industry data, in the context of an integrated LAustry inspection program with the participation of the three PWR Owners Gror.ps, EPRI, and NEl, indicsks that an examination is appropriate.
Further support for this approach to managing the issue can be found in Reference 2. This NRC-funded report concluded that either axial or circumferential CRDM nozzle cracking is unlikely to propagate through the nozzle wall to cause rupture, and that even this unlikely c.cenario is within the design basis of the plant. Reference 2 further concluded that throughwall circumferential flaws are unlikely, that leakage from a throughwall axial flaw is unlikely, and that if such leakage was to occur, it is very unlikely that it would remain undetected prior to boric acid corrosion challenging the structural integrity of the low alloy steel closure head.
l l
Attachment E provides a description of the relevant data and testing used to develop the DEI crack initiation and crack growth models, the methods and data used to validate these models, the plant-specific inputs to these models, and how these models substantiate the susceptibility evaluation. Section 2.0 of Attachment H provides additional crack growth data which is consistent with that presented in Attachment E.
Comed is a participant in the WOG RPV head penetration integrated inspection program. See Attachment H, Section 1.3, for a description ofinspections that have been performed to date by WOG members and other utilities. In addition to the WOG integrated inspection program, all three PWR Owners Groups, EPRI, and NEl are cooperatively working to compile and compare information on the probabilistically estimated operating time from January 1,1997 needed to initiate and grow a crack 75%
throughwall in all PWR vessel penetrations. This information will be evaluated to determine whether an l
adequate number of plants have examined or are planning to examine. This PWR industry integrated evaluation will be completed by the end of 1997 and will be prmided to the NRC in a supplement to this Generic Letter response.
l More detailed evaluations for other VHPs, which in the Comed PWRs consist of the head vent, will also l
be provided by the end of 1997 in a supplement to ti,is Generic Letter response. The head vents are l
currently considered to be bounded by the CRDM nozzle evaluations on the basis of their location near the I
center of the closure head, similar yield strength, and postulated smaller weld size.
l l
l 2.
Provide a description ofany resin beadintrusions, as described in IN 96-11, that have l
exceeded the current EPRI PIFR Primary Water Chemistry Guidelines recommendationsfor j
primary water sulfate levels, including thefollowing information:
To summarize, primary coolant system chemistry and conductivity data were reviewed from the beginning of plant life to the present for Braidwood Units I and 2 Byron Units I and 2, and Zion Units 1 and 2, and no evidence of resin intrusion events was identified.
i l
i i
2.1 Were the intrusions cation, anion, or mix. d bed?
I l
l No resin intrusions were identified.
, l i
l t
i
l
- Enclosure 2.2 li' hat were the durations ofthese intrusions?
No resin intrusions were identified.
2.3 Does the plant's RCS water chemistry Technical Specification: follow she EPR1 guidelines?
l Current Comed PWR reactor coolant system water chemistry Technical Specifications are consistent with EPRI "PWR Primary Water Chemistry Guidelines, Revision 2," (Reference 3) limits for chloride, fluoride and dissolved oxygen. In the near future, Zion will transition to the Improved Tecimical Specifications, I
which are based upon NUREG 1431, " Standard Technical Specifications for Westinghouse Plants."
Based on this transition, Zion Station has relocated the Zion chemistry requirements, as they exist today in the current Technical Specifications, to the UFSAR.
2A Identify any RCS chemistry excursions that exceed the plant administrative limitsfor l
thefollowing species: sulfates, chlorides orfluorides, oxygen, boron, and lithium.
l Comed has reviewed plant historical chemistry and conductivity data to determine if any incident of resin l
ingress similar to those which occurred in 1980 and 1981 at the Jose Cabrera Zorita plant has occurred at Zion, Braidwood, or Byron. The data review was structured to provide the data requested, and, to be consistent with Attachment F, to identify any resin intrusion events into the primary coolant system with a
(
magnitude greater than one cubic foot (30 liters). The threshold of one cubic foot was chosen as a conservative lower bound, since it represents less than 15% of the estimated volume of resin released into the reactor coolant system during the two events at Zorita.
The data review was performed for all operating modes using the following conservative screening criteria:
sulfate
>50 ppb o
>150 ppb a
fluoride
>150 ppb e
>100 ppb specific cor>Aenm
>28 S/cmelevation o
a lithium
>2 pprn over baseline Braiducod and Byron have routinely analyzed for sulfate in the primary coolant system since 1991 and 1993, respectwely. Zion has routinely analyzed for sulfate in the primary coolant system since February 1997 for Unit I and sina September 19% for Unit 2. However, since there is no naminietive limit for sulfate, the screening criteria limit is based on the guidana of Attachment F.
i l
The screening criteria limits for chloride, fluoride, and oxygen are based on current plant Technical l
Specification limits for steady state operation.
r i
The screening criteria for specific conductance is based on the Attachment F criteria for resin intrusion events greater than one cubic foot.
There is no administrative limit for boron, and lithium is only indirectly controlled through its relationship with boron in controlling pH. Since an increase in coolant lithium concentration could be an 1
l l
l l
l l
l Enclosure i
inc. ; ion oflithium reicased from degraded mixed bed or cation resin, the screening criteria above are l
based on a 2 ppm increase in the lithium concentration from the baseline trend, as specified in Attachment F.
Attachment G provides a description, technical assessment, and disposition of each primary coolant system chemistry excursion beyond the screening cliteria. No primary coolant system chemistry excursion observed was attributed to a resin intrusion.
2.5 Identify any conductivity excursions which may be indicative ofresin intrusions. Provide a technical assessment ofeach excursion and anyfollowup actions.
See the answer to Question 2.4 above for a description of the screening criteria utilized for the review of conductisity data.
Attachment G provides a deription, technical assessment, and disposition of each primary coolant system conductisity excursion. No primary coolant system conductivity excursion obsened was attributed to a resin l
intrusion.
l l
2.6 Provide an assessment of the potentialfor any ofthese intrusions to result in a sigmficant increase in the probabilityfor IGA of171Ps and any associatedplanfor inspections.
Based on the primary coolant system chemistry and conductivity data reviewed for Braidwood Units 1 and 2, Byron Units 1 and 2, and Zion Units 1 and 2, no resin intmsions hate occurred.
Further, of the sulfate excursions identified beyond the 50 ppb screening criteria and attributed to other sources, none exceeded the 1.7 ppm criteria, which was conservatively establisixxl in Attachment F as being indicative of a small (<0. I cubic feet) resin intrusion, in the primary coolant system.
On these bases, Comed concludes that I
1)
No CRDM nozzle or other VHP IGA has occurred due to primary coolant system resin intrusion, 2)
No CRDM noule or other VHP IGA has occurred due to other p'imary coolant system l
chemistry or conductivity excursions, and l
3)
No associated plan for inspections is necessary.
References:
t 1) del-480, "CRDM Nonle PWSCC Strategic Planning: Zion Units 1 and 2, Braidwood Units 1 I
and 2, Byron Units 1 and 2," March 1997.
2)
NUREG/CR-6245, EGG-2715, " Assessment of Pressurized Water Reactor Control Rod Drive i
Mechanism Nozzle Cracking," October 1994.
j 3)
EPRI "PWR Primary Water Chemistry Guidelines: Revision 2," NP-7077, November 1990.
l l
1 l
Attachments:
l l
A)
NEl White Paper," Alloy 600 RPV Head Penetration Primary Water Stress Corrosion Cracking,"
l March 5,1996.
l l
B)
WCAP-13565, Revision 1," Alloy 600 Reactor Vessel Head Adaptor Tube Cracking Safety Evaluation," February 1993.
C)
NRC " Safety Evaluation for Potential Reactor Vessel Head Adaptor Tube Cracking,"
November 19,1993.
D)
DEI letter L-5061-01-03, Attachment 1, July 10,1997.
l l
E)
DEI Letter L-5057-00-1, July 22,1997.
F)
" Screening Criteria for Evaluation of Reactor Coolant Chemistry Relative to RCS Resin Ingress During Power Operation," and " Rationale for Selection of Screening Criteria Parameters Relative l
to Resin Ingress to the RCS," both excerpted from Westinghouse Owners Group " Transmittal of Fint Resin Intrusion Guidelines and Suggested Response for Item 2 of GL 97-01 (MUHP-i l
50: e24)," June 30,1997.
l G)
Chemistry and Conductivity Excursion Summary, Braidwood Units 1 and 2, Byron Units 1 and 2, and Zion Units 1 and 2.
H)
WCAP-14902, " Background Material for Response to NRC Generic Letter 97-01: Reactor Vessel Closure Head Penetration Integrity for the Westinghouse Owners Group," June 1997 l
1 7
i i
l I
i
L 9
2p.,,
a-
-4 4A-
-,e'.M-l t
1 i
P l
ATTACHMENT A P
i i
I I
i I
I I
l t
P 1
- _.;; 7
_..q I
l l
l L
l-l AT.T.OY 600 RPV HEAD PENETRATION PRIMARY WATER STRFAS CORROSION CRACKING
/
1 l
1 i
l i
Nuclear Energy Institute 1776 I Street, N.W., Suite 400 Washington, D.C. 20006 3708 l
MARCH 5,1996 i
4 l
,.3 fn l 3
- -m
{ h O d l 6 " ' 'o (d'
=
i 4
Af.T Oy 800 RPV HPAn PFvrvn ATrnN 1
i 1
PnfurWyWATFW ETRF44 CORROMTON CRArinNG i
j MCUTIVE Stmurny:
1 The purpose of this paper is to review the sigmficance of primary water stress i
3 corrosion cracking (PWSCC) in pressurned water reactor (PWR) vessel head penetrations and to describe how industry is managmg the issue. This is achieved i
by:
j Reviewing worldwide PWSCC history in head penetrations:
e Summanung safety evaluation conclusions reached by industry and e
j.
approved by the Nuclear Regulatory Comnussion (NRC);
[
Summarizing supporting tasks performed by the Owners Groups (OGs) t and the Electric Power Research Institute (EPRD; Discussing the penetration inspection acceptance criteria and the n9n-j destructive erammation (NDE) capabilities demonstration:
Discussing licensee decision considerations: and i
Clarifying the proactive approach licensees will use to manage PWSCC of e
{
Alloy 600 head penetrations.
I
!J Reactor pressure vessel head penetration cr=aMagis caused by PWSCC and is not an immediate safety concern. Using conservative evaluations, the NRC and industry concluded that internally initiated cracking will be in the smal orientation and will take at least six years to propagate through wall. Additional leak-before-i break evaluations have determmed that the mmimum critical axxal crack length ranges from 8.5 inches to 20 inches, depending on plant design. External circumferential cracking is possible, but only in the presence of an above the weld i
through wall crack with active leakage. If coolant is present on the outer diameter i
of the penetration, one conservative analysis estunated that it would take in excess 2
of 90 years before failure would occur. In the presence ofreactor coolant, cormaion of the alloy steel head is possible. Conservative evaluations estimate it would take longer than six years after a through wall crack occurs before the code structural l
integrity margin for the head would be impacted by corrosion. It was concluded i
that visualinspection of the head in accordance with Generic Letter 88-05 is sufficient to detect PWSCC leakage prior to sinnma=nt cr=aW=e and head corrosion.
1 a
Worldwide, approximately 5,150 Alloy 600 reactor head penetrations have been I
inspected since the first cr=aMag was observed in 1991. Approximately 2% of these 1
penetrations were found to be cracked. Moet of the cracks were observed in French j
reactors. As noted in NUREG/CR-6245, " Assessment of Pressurized Water Reactor i
Control Rod Drive Maahaai m Nozzle Cr=eM= ", their nozzles were fabricated from 1
- Alloy 600 forged bars. If the French inspection results are removed from the j
inspected population, the percentage of head penetrations with inriirwtions is 4
i 1
T 1
q m
~,
c r
i 4
i decreased to about 0.5%. Only one plant worldwide has experienced PWSCC head i
penetration leakage and that was identified during hydrostatic testing, not i
operatioP3.
.t Specialized NDE methods have been developed and veri 5ed using mock.ups. The 4
ability of these NDE methods to detect the size and type ofindications expected to be present in the vessel head penetrations were demonstrated. Inspection criteria i
developed by the industry were also approved by the NRC staff.
i 1
Owners Groups have developed predictive methodologies and teols to assist utilities i
in evaluating various inspecuan. repair. and replacement. options. Each of the i
i Owners Groups has methods to evaluate the probability of a penetration developing a crack or a through wallleak during a plant's lifatime. This information is then used to evaluate a utdity's need for in parrion of the reactor vessel head penetrations. In addition. some tools may be used to determme the economic consequences of a i
j utility's actions.
i 4
Industry is t= Mar a pmactive approach to address RPV head penetrations PWSCC.
3 This approach is based on the conclusion that it is not an immediate safety concern and that leak-before break will occur. Industry's management approach has four i
elements: (1) a commitment to share new information as it becomes available; (2) j use of management / evaluation tools; (3) inspections; and (4) development and
[
evaluation of mitigation and repair alternatives.
l i
i i
4 1
1 i
=
- ~
l At i ny 800 RPV Hr An ?ENETRATION ParMARvWATrn MTRret COMRnginN CW APINNG I. PURPOSE The purpose of this paper is to review the signibance of primary water stress corrosion cracking (PWSCC) in pressurized water reactor (PWR) vessel head penetrations and to describe how the industry is managmg the issue. This is achieved by:
Reviewing worldwide PWSCC history in head penetrations:
Su===en:ing safety evaluation conclusions reached by industry and approved by the Nuclear Regulatory Commi==ian (NRC);
Su===rizing supporting tasks performed by the Owners Groups (OGs) and the Electric Power Research Institute (EPRI);
j
^
Discussmg the penetration inspection acceptance criteria and the non-destructive er==ination (NDE) capabilities demonstration:
j Discussing licensee decision considerations; and Clarifying the proactive approach licensees will use to manage PWSCC of i
Alloy 600 head penetrations.
i IL BACKGROUND a
h
' PWR vessel heads are fabricated with Alloy 600 control rod drive maA=ni==1 i
(CRDM) penetrations (Figure 1). In 1991, the first CRDM penetration cracking was i
observed at the Bugey 3 plant in France, when a small amount ofleakage was i
detected durmg a primary system hydrostatic test. Subsequent inspections
{
identified several additional axial cracks located in the penetration tube at an elevation near the partial penetration weld between the penetration and the vessel
}
head.
l l
Root cause evaluations concluded that the cracks were caused by PWSCC of the j
Alloy 600 material. Furthermore, Electricits de France (EdF) concluded that the j
followmg factors con *ributed to the Bussy PWSCC:
i
. a susceptible nuarostructure produced durmg manufacturing;
. surface finish on the inside of the nenetration: and
(
. - stresses induced during welding, w'aich caused ovalization of the pumsa 4
EdF conducted additional CRDM penetration inMens at its nuclear plants, i
. using eddy current techniques for indication detection and ultrasonic methods for
'j i also, Connel Element Drive Mach==== (CEDM) 1
- =.
. _.. _.__ _ ~__.___ _ _. _ _ _ _.. _ _ _
sizing. Inspection results indicated PWSCC in CRDM penetrations at several other EdF plants. This was a concern to French auuorities and to other PWR owners and regulatory authorities around the world. By the begmnmg of 1996.
{
approximately 5,150 penetrations had been inspected worldwide. Table 1
{
. summarizes the worldwide inspection results. Indications were found in approximately 2% of the inspected penetrations.
Table 1: Worldwide Vessel Head Penetration PW5CC Inspection itesults =
Country Plants Total Penetrations Penetrations Penetrations inspected inspected With indications France 47 l
3225 l
3213 l
105 Sweden l
3 l
195 l
190 l
7 Switzerland l
2 l
72 l
72 l
2 Jooan l
17 l
960 l
634 l
0 Belgium l
7 l
435 l
435 l
0 Spain l
5 l
325 l
102 l
0 Brazd l
1 l
40 l
40 l
0 i
South Africo 1
l 63 l
63 l
0 i
i UnitedStates l l
249 l
197 l
2 i
4 TOTALS l
47 l
5565 l
5144 l
116 j
- Availante inspecuon results ss of January 1996.
In 1994, three U.S. PWR plants voluntarily performed inspections of CRDM penetrations. No hdications were identified at Point Beach 1. One penetration at Oconee-2 was identified with numerous very shallow indications and one penetration at D. C. Cook-2 showed three confirmed axial cracks considerably smaller than the NRC approved acceptance limit. In 1995, Palisades inspected i
eight peripheralin-core instrument penetrations that were found to be free of indications. It appears that the frenuency ofidentifying indications at U.S. plants is less than at EdF nuits. Ifinspection results from France are excluded, the percentage of penetratices with indications in the remamder of the world is approximately 0.5%. A possible explanation for the differing results between the EdF units and the U.S. nuclear plants may be due to the different material fabrication methods used by EdF.
2
=
J t
,i Domestically, the Babcock and Wilcox Owners Group (B&WOG), Combustion Engineering Owners Group (CEOG), Westinghouse Owners Group (WOG) and the i
Electric Power Research Institute (EPRD agreed to combine their efforts as part of the Nuclear Energy Institute's Alloy 600 CRDM Head Penetration Cracking Tash l
Force. The purpose of the task force was to evaluate the issue and to recommend appropriate generic actions. Through this effort, the Owners Groups (OGs) and EPRI have performed safety analyses of CRDM PWSCC, standardized flaw i
evaluation methods, developed flaw acceptance criteria, developed inspection methodologies to detect and size indications, evaluated remedial measures and
{
created decision tools. The NRC has evaluated the submitted safety analyses and
]
concluded that PWSCC of Alloy 600 head penetration is not an immediate safety j
concern.
i
$3
/
/
i
/
/
/
//
4
//
1 d
taesman of Adel d
/
/
(
\\
p w.w
/
d
~
cau thenom soon
""4 FIGURE la GEOMETRY OF TYPICAL RPV Hrrn PENETRATION 3
=
i
- m. SAFETY EVALUATIONS The PWR Owners Groups developed extensive safety evaluations that were submitted to and accepted by the NRC stas. The purpose of these analyses was to evaluate the safety signi&am of PWSCC initiation and growth on internal and
)
external penetration surfaces. The NRC staf reviewed these safety evaluations J
and provided a safety evaluation report (SER) that concluded "there is no immediate safety concern for cracking of the CRDM/CEDM penetrations"(1).
i The PWR Owners Groups' safety evaluations are documented in References 2 through 12.
The Owners Groups' safety evaluations and NUREG/CR-6245 (13) conclude that CRDM penetration cracking is not an immediate safety issue. In addition to the overall conclusion, the safety evaluations document that:
)
a)
CRDM penetration crackmg is caused by PWSCC, which occurs when l
i high tensile stresses, high operating temperatures and a susceptible material are present simultaneously, i)
Residual fabrication stresses have a greater influence on PWSCC crack initiation than stresses created durmg plant operations. Typically, the outer-most head penetrations have the greatest amount of residual stresses.
ii)
Crack initiation is sensitive to head operating temperatures.
Inwer temperatures sici8a=atly reduce the probability of PWSCC inination.
i iii)
Typically, lower yield strength and greater carbide coverage on the grain boundaries increase resistance to PWSCC.
b)
CRDM penetration cracking from the internal diameter (ID) surface, if it occurs, initiates at the highest stressed locations and propagates in the axial orientation.
i)
Assummg crack initiation, it has been conservatively estimated that it will take at least six years for the indication to propagate 3
through-wall.
ii)
In a leak-before-break type evaluation, the time for a through.
waRindication to grow to criticallength was determined for each NSSS design. Using conservative evaluations, the i
maimum critical crack length ranges from 8.5 inches to 20 inches.
iii)
Residual stresses located in the penetration above the vessel head di=iini=h rapidly. Therefore, a propagating crack is j
expe::ted to terminate growth prior in machia the critical e
r s
4 1
l length and leakage idenufication is likely before CRDM penetration failure occurs.
c)
.CRDM penetration cracking from the external surface is possible only in the presence of reactor coolant. To have coolant present externally above the weld, a leaking through. wall crack unst be present above the weld.
f i)
Penetration failure due to external circumferential cracking would take significantly more time than the licensed 40. year operating period.
l ii)
One conservative evaluation estimated that it would take in l
excess of 90 years before this type of failure would occur.
l iii)
External circumferential cracking and penetration failure, prior i
to visual detection ofleakage, is considered a highly improbable event.
s d)
An assessment of boric acid wastage (i.e., pittag and wall.thinnin-by general corrosion) of the alloy steel reactor vessel head was performed.
Such corresion can occur if a leak exists. It is a concern because excessive loss of the li-===nt material between penetrations could t
challenge the head's structuralintegrity. Assuming the existence of a through. wall crack, conservative evaluations conclude that wastage could occur for at least six more years prior to challengmg the head's
.6 i.umiintegrity.
e) 14akage rates were evaluated considering the nature of PWSCC geometry, crack size, and the gap between the head and penetration.
While the rate of coolant leakage is predicted to be small, boric acid buildup would be sinnineant prior to the rupture of the penetration or
=i-nine==t wastage of the head.
i)
Many cubic feet of boric acid crystal buildup would occur during the years of service prior to a penetration failure or a challenge to the strik.i.umiintegrity of the reactor head.
iD Visualinspection walkdowns of the head in accordance with licensee commitments to Generic Letter 88 05, " Bone Acid C=.veien of Carbon Steel Reactor Pressure Boundary i
l Ce==anents in PWR Plants," are aufBcient to detect leakage l
well before challenging the structuralintegrity of the head.
f)
Evaluations using 10 CFR 50.59 requirements conduded that head penetration cracking is not an unreviewed safety question.
I 5
l m
.. ~
)
Tasks identified in Table 2 reflect the activities undertaken or under consideration by the PWR Owners Groups and EPRI to support utility management of Alloy 600 CRDM PWSCC.
I Table 2: Summary of Taska Performee by the OGs ans EPRI ITEM TASK B&WOG CEOG WOG EPRI 1
Root Cause of Craciong X
X X
X 2
Key Matena & Operanon Parameters X
X X
X 3
Os,oc Finas Element Analysis:
X (20 only)
X Resu$uallooeraconal X
4 EissectPlasoc Finne Element Analysis:
X X
X
.X Rossdual/Opersbonal: 3 Locenons 5
Cr s Fi-y-y IAcceptacle Flaw Saza X
X X
i Analysis 6
Penetracon Leakage & Vesset Heao Wastage X
X X
i Assessment i
7 Safety Evaluenon X
X X
l 8
Plant SJ
.ym
- ai Cntena X/O X
X 9
Matensi Microstructure Charactensocs O
X/O 10 t.askoge Detocoon Meinoes Survey X
X 11
"..-.;.asson of W..I,C Millganon Melnoes O
X 0
X/O 12 G.
- ., ~.M on Ressual Stresses P
X 13 t.ow Alloy Steel Womage Data X
X X
14 NDEIndussy Stanomros W.
X X
X 15 NDE. -.'...
C. -
3; --
X 16 andustry Concis.
of NDE Si-X X
X 17 Cr;----. e.
n of Reperso X
X X
X ConAgurations 16 OD Crack Asseeement X
X l
X 1g Cract Growm Dets ano 1 sang O
O l
0 0
20 inspecnon Timing and L. Deceson Toots X/O X
NO 21 Ranses PWR Water Chemmary Guidennes X
22 inaustry Alloy 600 Dassesse I
X/O 23 FK-- Attacnment Weed Safety Evaluanon O
X Report 24 Cract insbeson Characienzanon Studes O
O O
O 25 C;. - _, _.; of Tootng for Top of-Hese O
inspeelinn & Cleenmg I
26 industry Wonanops on PWSCC of Ahoy 600 X/O X/O X/O X/O Penessalons 27 ReessumiStress Measurements X
X X
X KEY: "X" = Complete. "O" = Ongomg. "P" = Planned, Blank = no plannes actmbes 4
9 6
m
IV. Demonstration ofInnnection Canabilities To ensure that inspection methods can accurately detect indications, the Owners Groups working with the EPRI NDE Center developed head penetration mock ups i
I to demonstrate inspection capabilities. Mock ups of the various head penetration i
configurations were fabricated and made available to utilities and their inspection contractors to demonstrate the procedures and techniques.
I I
Mock ups representing the B&W, CE, and Westinghouse vessel head penetration i
]
designs were constructed. For CE designs two mock up penetrations were constructed, representing in core instrumentation and the control element drive i
mechanisms. N mock ups were constructed using materials of the same alloy, product form, fabrication methods, and heat treatment used in production. N
{
penetrations were welded to a mock up of the vessel head usmg production welding j
procedures. The mock ups were required to have :he same thMn-ss as the j
production heads, but flat plates were used inste ad of the curved head material.
h mock-ups were designed to simulate the outarmost penetracions, which represent the largest offset (difference in height of the weld from one side to the l
other caused by the slope of the head).
]
Prior to welding, the penetrations were fitted into the head mock-up using typical production tolerances. In most cases this involved a shrink fit, where the 1
i penetration is cooled by liquid nitrogen and then inserted into the head. h final weld profile was required to be typical of production practice. and the configuration i
1 of the weld preparation was required to be documented.
1 i
A substantial number of flaws were introduced into the penetrations, all designed to be typical of those which might be found in service. N flaws were alllocated on j
the inner surface of the penetration, and the orientations varied from purely axial i
to purely circumferential Flaws were introduced into the penetrations by areW 1 i
means to provide more accurate knowledge of the true flaw sizes and locations than j
would be possible forinduced PWSCC.
i Great care was taken to ensure that the eddy current and ultrasonic responses of the aremei=1 flaws were realistic and that the flaw accurately simulated the response of actual PWSCC. Personnel from the EPRI NDE Center visited the EdF I
laboratories and compared the responses directly with those from indications in a l
penetration sample removed from an operating plant. The signal responses were 1
similar in both amplitude and noise levet h mock ups are available at the EPRI NDE Center for use by utilities and their l
inspection contractors.
i 4
7 4
1 i
V. PENETRATION INSPECTION CRrrPWTA I
Table 3 summarizes the acceptance criteria developed for indications that may be j
found dunng inspection of reactor vessel head penetrations. The acceptance criteria i
for axial flaws have been reviewed and accepted by the NRC (1).
4 4
)
Table 3: P.eactor Head Penetration indication Acceptance Critens j
inoicaten i.ocanon Indication Orientation and Maximum Acceptable Size
]
in Penetration Amal Cwm....i.ntial "
j 1
I j
Below Weid t
no twrut" t
.75c At/ADove Wald 0.75t no imt 0.75t 0.10c i
i s, = Flaw depth as defined in IWB 3600, Secten XI i
C = Circurnference I = Flaw length t = Wallthickness
- Must not touch weld.
1
{
" Circurnferentialindmatens will be reviewed by the NRC on a case-by-case bass. The Table 3 circurnferential acceptance entens is a screening entens that the NRC has accepted for some i
licensees in advance of their plant inspecnons.
i 1
The approach used to develoo the acceptance criteria applicable to all PWR plant designs followed that used by the ASME Code,Section XI. Because the safety j
j evaluations demonstrated that the penetrations are tolerant oflarge Saws, the goal j
is to protect against leakage during service. The flaw acceptance criteria are more conservative than those used in Section XI applications.
The maximum allowable depth (at) for flaws at or above the weld is 75 percent of 3
the penetration wall tMahass regardless of the flaw orientation: this value was j
selected to be consistent with the maximum acceptable flaw depth in Section XI and i
to provide an additional margin against through wall penetration.
Axtal flaws found below the weld are acceptable regardless of depth as long as their 3
upper extremity does not reach the bottom of the weld during the period of service i
until the next =Wn. Axial flaws which extend through and/or above the weld i
2 are not limited in length but are limited to a depth of 75 percent of the wall.
Circumferential flaws will be accepted by the NRC on a case by-case bases. Some licensees have received NRC authorization, in advance of their inspection, to leave in service circumferentialindications with lengths up to 10% of the circumference if j
their depth is less than 75% of the tMehass. Calculations have been completed that demonstrate all penetration geometries can perform their function in the i
presence of a continuous (360*) circumferential flaw above the weld with a wall j
depth of 75 percent.
4 i
8 j
_.. ~. _. _ _ _ _ _
i
)
i i
1 The need for follow.up penetration inspections will be made on a plant. specific j
basis. Flaws that exceed, or are expected to exceed, the acceptance criteria must be j
repaired unless analytically justified for further service and accepted by the NRC.
i Such justifications wi11 require accurate crack growth predictions and indication I
sizing information.
l VI. Insnection Timine and Economic Decision Tools i
1 The Owners Groups have develop predictive methodologies and tools to assist utdities evaluate various ia===acian. repair, and replacement options. Each Owners Group i
has methods to evaluate the probability of a penetration developing a crack or a i
through wallleak durmg a plant's lifetime. This mformation is then used to evaluate a utility's need for reactor vessel head inspectaons. In addition. some of these methods are able to predict the economic consequences of a utility's actions.
The susceptibility to cracking of a penetration materialis based on industry. accepted parameters such as material yield strength, operating temperature, and operating time. The predicted stress on a penetration is based upon the same type of elastic-plastic three dimensional finite-element models used in the Owners Groups Safety l
Analyses (References 2 thmugh 12). The probability of developing cracks is also a=1~1= tad using industry accepted values for crack growth and actrvation energies.
Life-cycle costs are then a=1-1=ted using probabilistic techniques or developed i
separately by the utilities based on the types ofinspections performed and the actions j
taken (e.g., penetration repair, mitigerian or replacement).
i Utility management of PWSCC head penetration cr=aking concerns ensures nuclear j
safety and plant operation, while mmimiting potential economic consequences. In the absence of a safety issue, these types of economic planmng methods are used to evaluate the impact of PWSCC head penetration
-H on plant operation and maintenance costs.
Utility ia===a+i-d--i=ians are deteramed on an individual basis, since reactor heads differ in design, materials, fabrication and operation history, Utilities have and will continue to evaluate their need for reactor vessel head in a--+iaa=. After careful
==i% tion, they have developed programs appropriate for their plant (s) with varied scopes such as 100% penetration inspection, representative sample insp+Iei and ongoing========nt in lieu of planned inspections. It should be recognized that all utdities perform visual ia===arians of the reactor vessel head for potentia 11eakage.
This is done in acordance with licensee responses to Generic Iatter 88 05 for boric acid corrosion of pressure bonpriary components.
9 M
+4
VII. MANAGNFNT OF RPV HEAD PENETRATION CRACKING Industry ir '=Ma-a prometive approach to mange RPV head penetrations PWSCC.
This approach is based on the conclusion that it is not an immediate safety concern and that leak-before-break will occur. Industry's management approach has four elements; these are:
Information Sharine Owners Groups, EPRI, and others continue to interact and share inspection results and other information as it becomes available.
This mformation is constantly evaluated to determine if changes to industry's previous safety analyses, acceptance criteria, or evaluation methods are warranted.
Predictive Methodolonies and Tools Industry has continued to develop tools (see Table 2) that enable licensees to determine when inspections are appropriate. The decision whether to inspect, or when to iaw. is based on ia-i-h+=
gained from applying these methods, results from other penetration inspections, and careful consideration of the plant-specific economics.
Inanections i
Visualinspections are performed per Generic Letter 88 05. These
{
inspect. ions continue to demonstrate that a challenge to safety does i
not exist by confirmmg that =i-nih=nt leakage is not present.
j Other inspections (e.g., eddy current testing, ultrasonic testing, liquid penetrant inspection, or augmented visualinspections) are p '.med when determmed to be appropriate by licensees.
1 Mitiration and Renmir Industry continues to develop and evaluate mitigation strategies and repair alternatives (see Table 2). Close cooperation between utilities and vendors assures that utilities will have the ability to safely address and, if necessary, repair any confirmed cracks which may be found.
1 If and when licensees decide to perform additionalinspections, the plant's NRC Project Manager will be notified and provided information descrzbing the j
mspections to be performed.
10
=
1 vm. REFERENCES
)
t
- 1. - NRC letter from William T. Russell to W. H. Rasin of NUMARC (now NEI),
Nov.13,1993.
i i
- 2. BAW-10190P, " Safety Evaluation for B&W-Designed Reactor Vessel Head i
Control Rod Drive Mechsmmm Nozzle Cracking," May 1993 (Proprietary).
- 3. BAW-10190, " Safety Evaluation for B&W. Designed Reactor Vessel Head Control Rod Drive Mechani=m Nozzle Cracking," June 1993 (Non-proprietary).
- 4. BAW-10190P, Addendum 1 " External Circumferential Analysis for B&W Design Reactor Vessel Head Control Rod Drive Maah==i m Nozzle,"
December 1993 (Proprietary).
- 5. BAW-10190, Addendum 1, " External Circumferential Analysis for B&(
Design R==~~ Vessel had Control Rod Drive Maah==i=== Nozzle," January 1994 (Non-proprietary).
- 6. CEN-607, " Safety Evaluation of the Potential for and Consequences of Reactor Vessel &nd Penetration Alloy 600 ID Initiated Penetration Cr=aMar," May 1993.
~
- 7. CEN-614, " Safety Evaluation of the Potential for and Consequences of Reactor Vessel Head Penetration Alloy 600 OD-Instinted Penetration Cracking," December 1993.
- 8. WCAP 13565, Rev.1, "Adoy 600 D--~~ Vessel Head Adapter Tube Cracking Safety Evaluation," February 1993 (Proprietary)..
\\
- 9. WCAP-13525, Appendix 1, Addandum 1,"RV Closure Head Penetration Alloy j
600 PWSCC (Phase 2)," December 1993 (Non Proprietary).
10.WCAP-13603, Addendum 1, "RV Closure Ea.. Penetration Alloy 6000 i
PWSCC (Phase 2)," December 1993 (Non Proprietary).
11.W. CAP 14219, Rev.1, "RV Closure Head Penetration Supplemental Assessment of NRC SER Issues," March 1995 (Proprietary).
12.WCAP-14432, "RV Closure Head Penetration Supplemental Assessment of NRC SER Issues," March 1995 (Non-Proprietary).
13.NUREGICR 6245, " Assessment of Pressurized Water Reactor Control Rod Drive Maah==im Penetration Cr=awar," October 1994 11
.--.-m
.-r m
.m e,,-,-m
t l
t l
ATTACHMENT B l
f r
I
i WCAP-13565 m.1 WESTINGHOUSE CLASS 3 l
l f
ALLOY 600 REACTOR VESSEL HEAD i
l ADAPTOR TUBE CRACKING l
SAFETY EVALUATION r
l l
February 1993 J
1 i
l l
1 l
\\
l WESTINGROUSE ELECTRIC CORPORATIODI Nuclear and Advanced Technology Division P. O. Box 355 Pittsburgh, Pennsylvania 15230 1993 Westinghouse Electric Corporation All Rights Reserved M % h'/fv33.-
- i s~
1
GENCL NO.92-011 Rev.1
(
Customer Reference No(s).
L Westinghouse Reference No(s).
MUHP-5016
?
WElfrINGHOUSE NUCLEAR SAFETY GENERIC SAFETY EVALUATION CHECK LIlif (GENCL) t 1.) NUCLEAR PLANT (S):
W= tin hnuse NSSS Pl==te 2.) SUBJECT (TITLE):
Allov 600 Dametar V===1 Head ki=ntor Tube Crackina j
3.) The written safety evaluation of the revised procedure, design change or modification required by 10CFR50.59(b) has been prepared to the stant required and is attached. If a safety evaluation is not required or is incomplete for any reason, explain on Page 2.
l Parts A and B of this Safety Evaluation Check List are to be completed only on the basis of the safety evaluation performed.
l l
CHECK LIST - PART A - 10CFR50.59(a)(1) l l
3.1) Yes _ No _X A change to the plant as described in the FSAR7 3.2) Yes _ No X. A change to procedures as described in the FSAR7 3.3) Yes _ No.X. A test or experiment not described in the FSAR7
[
3.4) Yes _ No.X_ A change to the plant eachnical specifications?
(See Note on Page 2.)
i l
4.) CHECK LIST - PART B - 10CFR50.59(a)(2) (Justification for Part B answers must be included i
on page 2.)
i 4.1) Yes _ No X_ Will the probability of an accident previously evaluated in the FSAR be increased?
4.2) Yes _ No X. Will the consequences of an accident previously evaluated in the FSAR be increased?
l 4.3) Yes _ No X. May the possibility of an accident which is different than any already i
evaluated in the FSAR be created?
4.4) Yes _ No X. Will the probability of a malfunction of equipment important to safety previously evaluated in the FSAR be increased?
4.5) Yes _ No _X. Will the consequences of a malfunction of equipment important to safety previously evaluated in the FSAR be increased?
4.6) Yes _ No _X. May the possibility of a malfunction of equipment important to safety l
different than any already evaluated in the FSAR be created?
4.7) Yes _ No.X_ Will the margin of safety as described in the bases to any technical j
specification be reduced?
Paes I of 35
- _ _. = -
NOTES:
If the answer to any of the above questions is unitnown, indicate under 5.) REMARKS and explain below.
l If the answer to any of the above questions in Part A (3.4) or Part B cannot be answered in the I
negative, based on written safety evaluation, the change review would require an application for i
license amendment as required by 10CFR50.59(c) and submitted to the NRC pursuant to 10CFR50.90.
l i
5.) REMARKS:
The answers given in Secticn 3, Part A, and Section 4, Part 3, of the Safety Evaluation Checklist, are based on the attached Safety Eva;uation.
\\
FOR FSAR UPDATE 1
Section:
Pages:
Tables:
Figures:
No FSAR Update Required SAFETY EVALUATION APPROVAL LADDER:
9grt J Nuclear Safety Preparer:
J. S. G=1a=hiith Date:
J 9J b
- 6. & a a t Nuclear Safety Verifier:
_ _ G. O.
Date: 2/9/ff Nuclear Safety Group Manager:
Date: 2 93
/ /
i I
i
)
i
?
Pese 2 of 35
3 TABLE OF CONTENTS Section East
1.0 Background
4 2.0 Licensing Basis 5
3.0 Evaluation 3.1 Penetration Stress Analysis 5
3.2 Crack Growth Analysis: Flaw Tolerance 10 3.3 Assessment of WOG Plants 17 3.4 Leak Rate Calculations 20 for the Reactor Vessel Head Penetration 3.5 Reactor Vessel Head Wastage Assessments 30 4.0 Determination of Unreviewed Safety Question 33 5.0 Conclusions 34 6.0 References 35 l
l l
Page 3 of 35 l
SAFETY EVALUATION POTENTIAL REACTOR VESSEL HEAD ADAPTOR TUBE CRACKING WESTINGHOUSE NSSS PLANTS
1.0 BACKGROUND
In late September 1991 Westinghouse was informed by Electricite de France (EdF) of the discovery of a leaking reactor vessel head adapter penetration (Figure 3.4-1) at the Bugey 3 plant in France. Bugey 3 has been in commercial operation since 1979. De leak was found during a hydrotest associated with a 10 year in-service-inspection (ISI). The hydrotest was performed at approximmely 3000 psi and 194 degrees F. De leak was discovered at core location H-14 (Penetration # 54), a peripheral full length CRDM location. The leak was located by using.
microphones attached to both the top and bottom heads of the reactor vessel. EdF determined that the leak rate was approximmaly 0.701/hr (0.003 gpm).
A visual examination p fvi-ed at that time indicated the presence of longitudinal (axial) cracks i
in the I.D. of the head adaptor tube. De head adaptor tubes are manufactured from Alloy 600 l
material. The use of Alloy 600 material for the head adaptor tubes is common to both Frammame and Westinghouse plants.
A subsequent inspection of all 65 head adaptor tubes at Bugey 3 revealed axial cracks at two peripheral head adaptor locations. After finding the leak at Bugey 3, EdF performed examinations at two additional plants. Framinmian of 24 penetrations at Bugey 4 revealed axial cracks at eight peripheral head adaptor locations. Twenty-six penetrations were inspected at Fessenheim 1. His examinmian revealed axial creckihg in one head adaptor. Based upon these subsequent inspections, EdF undertook an inspection program which ecompassed all of their operating plants. To date, over 500 reactor vessel head adaptor penetrations have been inspected encompassing thirteen (13) European plants. Of these irrpia=,26 penetrations have exhibited crack indications. Plant inspections are continuing in Europe.
At Bugey 3, EdF removed the penetration corresponding to core location H-14 for hot cell examination / root cause determmation. De rnachanism of the degradation (root c.tuse) was identified by EdF as primary water stress corrosion cracking (PWSCC). Westinghouse has reviewed the available metallographic records and concurs with this conclusion. De weld-induced bending and ovality of the peripheral penetrations appears to be the initiating source of the stress which is promoting the degradation.
De purpose of this safety evaluation is to assess the continued safe operation of Westinghouse designed NSSS plants focusing on the likelihood of cracking, the cherec; iunion of any such potential cracking, the pa**i=I for leakage, and finally, the disposition of low alloy carbon steel 1
i wastage issdes; in the knowledge that similarities do exist in the various plant designs between Westinghouse and European manufacturers.
This safety evaluation will provide the following elements:
1.
A summary of the stress analysis focusing on the type of cracking that may be expected in the Alloy 600 material, and the stresses necessary for crack propagation.
Page 4 d 35 m
2.
A summary of the crack propagati:n analysis will be provided along<with ths background of the crack prediction method.
3.
An assessment will be made of the Westinghouse Owners Group (WOG) plants with respect to penetration crack indication data from plant inspections at Ringhals Beznau and various EdF plants. De key parameters for cracking will be compared against WOG plants.
4.
A leakage assessment will be provided summarizing leak rate vs. crack size, and in postulating leaks for those few WOG plants for which leakage considerations may apply.
5.
A vessel head wastage assessment will assess the process by which wastage may potentially occur and an estimate of allowable wastage will be provided.
r 2.0 LICENSING BASIS The situation regarding the potennal cracking of reactor vessel head adaptor tubes at Westinghouse designed NSSS plants represents a change to the normal plant configuration. Title 10 of the Code of Federal Regulations, Section 50.59 (10 CFR 50.59) allows the holder of a license authorizing ope ation of a nuclear power facility the capacity to evaluate these types of situations. Prior Nuclear Regulatory Comminnion (I 3C) approval is not required to return the plant to power as long as the situation does not involve an unreviewed safety question or result in a change to the plant technical specifications ir-sposated in the license. It is, however, the obligation of the licensee to maintain a reco 1 of the change or modification to the facility, as a result of any given situation, to the extent that sixh a change impacts the FSAR. While this situation does not represent a change to the FSAR, 10CFR 50.59 further stipulates that these records shall include a written safety evaluation which provides the basis for the determination that the situation does not involve an unreviewed safety question. It is the purpose of this document to support the requiramant for a written safety evaluation.
The scope of this document is limited to an evaluation of the potential cracking in reactor vessel head adaptor tubes centermg on any effects this situation may have on existing plant equipment or any unreviewed safety questions that may be identified.
3.0 EVALUATION 1
3.1 PENETRATION STRESS ANALYSIS
Background
l Initially, several 3D-elastic fin'te elemsat analyses were performed to establish penetration stress magnitude and distribution. Dese analyses demonstrated that stresses caused by operational l
pressure and temperature (2250 psi and 6007.) loads are not large enough to cause penetration tube ovality of the =t-nimde which has been measured in a number of plants (based on penetration I.D. diametral and profile measurements taken from irradiated and non-irradiated vessels). It was further determined, qualitatively, that the residual stresses in and near the weld region due to welding are significantly higher than those caused by operational loads. It was also Paps 5 of 35 1
i determined that stresses experienced dus to the welding fabricati:n processes of httaching thm penetration to the vessel head exceed the yield strength of the Alloy 600 weld and penetration l
material at some locations.
i i
It therefore laame essential to perform stress analyses considering the inelastic mechanical
]
{
properties of the penetrations, to more r,uantitatively define the stress field in the penetration.
l The additional analyses not only have to be detailed enough to provide quantitative stress l
distributions, but also must envelope all WOG plant penetrations. To make sure that the 4-loop i
i models are enveloping, a parametric study was performed to study the effect of a) vessel size and b) penetration location. De results indicated that the outermost penetrations of the 4-loop plant,
{
having the largest weld offset angle among the 2, 3 and 4 loop plants, are the highest stressed I
penetrations under cioning loads as well as having the largest residual stresses. Therefore, it
)
was concluded that the outermost Wnions of the 4-loop plants are the enveloping a
1 penetrations of all WOG plants.
}
Having determined that the 4-loop plants are appropriate to represent all WOG plants, three j
penetration models were built using km,-. hue haaadant elastic-plastic material properties.
l Three different radial locations were modeled as described below:
i 1.
the center location (#0) i n
l j
2.
outermost location (e.g. penetration M8), and i
i I
3.
next to the outermost location (e.g. penetration #65).
i i
In the Westir.ghouse 4-loop plant, penetrations #65 and #78 are located radially from the vessel j
centerline r.t 59.8 and 64.5 inches respectively.
i The mo Jeis utilized 3-dimensional isoparametric brick and wedge elements. Taking advantage of j
symme ry through the vessel and penetration centerlines only half the penetration geometry plus j
the sur:ounding vessel are modeled. Dese medels are shown in Figures 2-1,2-2 and 2-3 of j
l Reference 6.
I J
De penetration tube, weld metal and buttermg were modeled as Alloy 600 and the vessel head shell as carbon steel. Elements with elastic-plastic capabilities were incorporated in the weld region and surrounding elements in both the penetration tube and vessel head shell. De stress-strain material propemes of the elastic-plastic elamanta representing Alloy 600 were derived from j
test data obtamed using an actual Alloy 600 penetration material sample taken from the outermost 3
penetration of an ir.arradiated plant. At this elevation in the reactor vessel, material irradiation effects are considered to be negligible. De curve used was a half-life cycle stress-strain curve (Cyclic stress-strain curve at one-half of the life of the penetration). Use of the cyclic stress-i strain curve is expe::ted to provide conservative upper bound levels for stress estimates in the penetration. Existing monotonic stress strain curves were used for the carbon steel elements.
Page 6 of 35 i
l To simulate the stress history cf the penetrati:n tube the following leading sequence was applied j
to each of the three models described. The stresses caused by each of the load cycles are stored and maintained in the model before the next load cycle is applied to simulate the effect of residual stress. His provides for the accumulation of plastic stresses.
i 1.
Dermal load from first weld pass.
4 2.
Thermal load from second weld pass.
3.
Fabrication Shop Cold Hydrotest (@ 3170 psi)
(a) Cold hydro-test loading (b) Cold hydro-test unloading 4.
Field / Site Hydrotest (@ 3170 psi)
(a) Cold hydro-test loading (b) Cold hydro-test unloading l
5.
Steady state operational loading i
i Imdings i
It was found from the analysis that the welding process introduces high residud stresses in the penetration tube near the partial penetration weld. The welding process was simulated by adding 1
the weld material to the model in layers and subsequently specifying the stress-free reference j
l temperamres for the weld and surrounding elements so as to provide shrinkage in the weld (due J
to cooldown). The reference temperatures were " benchmarked" or adjusted to generate ovality j
levels in the penetration tube approxunating those measured in actual penetrations while still 3
maintaming welding temperatures within reasonable limits. Two methods were tried, the first used two consecutive layers of welding, the second used three. De difference in the results i
between the two methods was insignificant and the two layer method adopted as the approach to be used.
He stress developed in the penetration model after applying the first weld pass (residual stress)
{
was maintained as the initial stress as the elements of the second weld pass were applied. He j
stresses induced by the welding simulation were large enough to cause plastic deformation in weld region of the model.
i jl The cold hydro-test loading required applying an internal pressure to the model of 3107 psi and
. performing the analysis at a temperature of 150'F. His combined with the cold hydro-test unloading (the unloading step is designed to return the model to an unloaded condition) step I
simulates the conditions during the fabrication shop hydro-test. De combination of the two weld 4
passes and the shop hydro conditions produced permanent set ovalities in the penetration i
comparable to field measured values. A second cold hydro-test loading and unloading was used to model the field hydro-test.
i 1
f
)
Page 7 of 35 3
i
i l
Adding the steady state operational loadings (i.e. pressure of 2250 psi and temperatura of 600'F.)
i brings the model to its final stress condition at operation. It has been determined through analysis that adding additional operational steady state pressure and temperature loadings and unloadings would not significantly alter the rtress state of the penetration tube. The stresses calculated at this point of the analysis provide an upper envelope of the steady state operating conditions in all the operational Westinghouse plants, i
i j
Series Analysis Results - Outernmost Penetration l~
De outward displacement of the vessel and penetration are shown in Figure 2-4 of Reference 6 l
for steady state operation. Note that the vessel displaces radially, and the tube displaces toward a i
position perpendicular to the head. Also, the gap opens slightly between the penetration and the j
vessel, on the vessel center line side of the iw ion, as shown in Figure 2-4 of Reference 6.
These results were obtained for an initial interference fit of 0.0 inches as fabricated, which is the I
J
)
minimum case for Westinghouse plants. A higher interference value would add a compressive stress to the steady state stresses, therefore, the 0.0 inch interference is conservative. As input to the potential leakage and wastage assessment, an average annular radial gap of 0.003 in. during plant operation represents a conservative estimate.
i a
The haop stresses at the inside surface of the penetrations are shown in Figure 2-5 and 2-6 of J
Reference 6. De highest stresses are found la a zone around the weld. De peak stresses fall along lines nearest the center of the vessel (centerside), and the side furthest away from the
)
center (180* away) (hillside). Note that these two locations correspond to the locations where i
axial cracks nave been found in service.
7 i
De axial s. tress distribution at the inside surface is shown in Figure 2-5 and 2 6 of Reference 6, l
which also shows that the highest axial stresses are along the weld. Note that the magnitude of i
axial axial stresses are less than the hoop component of stress. No axial cracks have been found j
in any penetrations in service.
The degree of plastic deformation in the penetration can be seen in the color contour stress plots j
shown in Figure 2-6 of Reference 6. Here we see that the most significant plastic deformation occurs at and below the centerline of the weld, extending down the sides of the tube at the upper j
and lower hillside locations, similar to the pattern of hoop stress in Figure 2-5 (Ref. 6).
he stresses decrease significantly above the weld, as may be seen qualitatively in Figure 2-5 (Ref. 6). Since the extent of crack propagation above the weld is of interest, the stresses at the inner and outer surfaces of both the inner and outer hillside location have been evaluated. Plots of both the axial and hoop stresses as a function of :listance from the bottom of the penetration are also shown in Figure 2-5 of Reference 6. Note that the weld location is shown on each plot, and also note that the weld location is different for the hillside and center side locations.
The highest hoop stress at the I.D. surface avemak the axial stress at the same locatica by a factor of appronmately 1.4, which in turn corresponds approximately to the ratio of hoop to axial stress ratio of 1.6 reported to be obtained by field measurements of an actual penetration.
This finding strongly supports the contention that axial cracks are the most likely orientation to be expected, which is consistent with all reported inspection findings.
Page s of 35
I Intermediate Penetration ne intermediate penetration analyzed is on a radius of 59.8 in, from the vessel centerline (i.e.
penetration No. 65 in a four loop plant). De stress analyses were carried out in precisely the i
same manner as the outermost penetration.
Figure 2-8 and 2-9 of Reference 6 provide plots of the hoop, axial and Von Mises stresses for the steady state operating conditions.
Center Penetration The stress analysis carried out for the center penetration was performed in the same manner as j
the outermost penetration. He maximum hoop stresses,34.6 ksi, are higher than the corresponding axial stresses,21.4 ksi. De overall stress magnitudes are lower than those of the outer penetrations. De stress contours along the inside and outside surfaces are shown in Figures 2-10 and 2-11 of Reference 6.
Bolt-Up Effects
)
An additional 2-dimensional, axisymmetric model was developed and used to determine the stress contribution in the closure head shell, due to bolt-up loads. Bolt-up loads are generated in the closure head shell by the rotation of the flanges caused by the tightening of the stud ten:*one.c.
De analysis evaluated the change in the stresses at a radial distance of 64.5 inches (penetration location #78) from the vessel centerline through the thickr.ess of the shell. De results of the analysis indicate that the bolt-up stresses are negligible compared to the operational stresses. This analysis also demonstrates that the simple boundary conditions (symmetry) used in the 3-dimensional, elastic-plastic analysis provide suitable constraints to the p oblem.
Ih*Hhaad of Circumfesuntial Cracks The hoop stress is the dominant stress in all the head penetration locations attached by an angular weld, which leads to the conclusion that cracks will be oriented axially. His is consistent with the inspection findings from several plants. He hoop stress at the inside surface of the outermost penetrations ar=& he axial stress by a factor of approximately 1.4, which is t
approximately the same ratio reported from actual measurements on head penetrations in service.
In all of these cases there is little or no likelihood of circumferentially oriented flaws in any of the head penetrations.
j Summary and Conclusions De stress analyses have shown that the stresses in the head penetrations are a strong function of the weld offset, or the angle of intersection between the penetration and the head. In Westinghouse plants, the four loop units have the largest weld offset, and therefore the outermost penetration location was chosen for a complets stress analysis. Dese results will be conservstive for all WOG two and three loop plants, which have the same penetration dimensions, but smaller offsets.
Page 9 of 35
4 mpAe J-e_m.Af=44-
-4mb--MaAbedp'"a**-C
+ - + *4hM
-AM
- M'"
i Fcr penetrations attached by an angular weld, th2 hoop stresses are greater than thz axial stresses, which is consistent with the inspection findmgs in that the flaws are axially orientei Furthermore, the locations of the maximum hoop stress correspond with the locations where cracks have been observed.
3.2 CRACK GROWTH ANALYSIS: FLAW TOLERANCE Introduction The goal of this work was to provide a quantitative measure of the tolerance of the head penetrations for the presence of a flaw. De mode of crack extension is primary water stress corrosion cracking, and therefore the loading of interest is the steady state operating condition.
1 The crack growth law used in this study was obtamad from a survey of the available literature on this material and environment, with the temperature effect on the growth rates based on a l
collection of crack growth information from both laboratory and field data.
Approach and Results The results of the three dimensional stress analysis of the outermost head penetrations were used directly in the flaw tolerance evaluation. The maximum stress is the hoop stress, but there is a component of axial stress which makes the maximum principle stress at a small angle to the hoop direction. Cracks would be PM to be oriented perpendicular to the manmum principle stress, which, based on the stress results would be at a slight angle to the axial direction. He flaws which have been found inservice are nearly all longitudinally oriented, thus the hoop stress component was used in the crack growth calculations. Stress analyses have shown that the penetration location with the highest stress is the 4-loop vessel head outermost penetrations, so this location was chosen for analysis.
j De crack growth evaluation for the "part-through" flaws was based on the stress distribution through the penetration wall at the location which corresponds to the highest stress along the inner surface of the penetration. The highest stressed location was found to be in the immediate vicinity of the weld.
The results of the calculated growth through the wall for surface flaws postulated in the highest stress location of all the penetrations are summarized in Figure 3.2-1. (De highest stress location was found to be just below the weld on the center side of the outermost -4.iion.)
Figure 3.2-1 applies to surface crack locations below and near the weld region, while cracks above the weld would grow more slowly because of the lower stresses. Note that the predicted extension through the penetration thickname requires a number of years for the entire range of operating temperaturne, regardless of the location.
Figure 3.2-1 presents the predicted crack growth for a "t' rough-wall" flaw postulated to exist h
below the weld region on the center side location. Although there are various levels of ovality (and therefore residual stress) in the various penetrations, in the vicinity of the weld and below it on the center side location the total stresses exceed the yield stress of the material. His was found to be the highest stress location on any of the penetrations, so the crack growth calculated will be conservative for all other locations.
t-Pass 10 of 35
To make these crack growth calculations, the average hoop stress through the call of the penetration was used. As the flaw propagates past the weld, the stresses in the penetration decrease and become compressive in some cases, so the crack will slow down and nearly stop once it reaches this location. He bottom of the centerside portion of the weld is located 7.2 inches from the bottom of the penetration, and is 2.2 inches in width at this location, so a crack length exceeding 9.4 inches will have the upper tip of the crack above the weld, he results for propagation above the weld are also shown in Figure 3.2-2, and show the growth slowing considerably above the weld, nearly stopping before reaching 11 inches.
On the lower hillside location the average hoop stresses are only slightly lower than those on the i
center side, and therefore the predicted crack growth is slightly lower. Dese results are shown in Figure 3.2-3. Note that the weld is located 3 inches from the bottom of the penetration, as shown on $e figure. Figure 3.2-3 shows the results for a postulated crack growing from below the weld, and again shows crack growth slowing considerably above the weld, this time nearly stopping at a height of about six inches,1.5 inches above the weld.
Sanmary and Conclusions An extensive evaluation has been carried out to characterize the loadings and stresses which exist in the head penetrations of Westinghouse plants. Three-dimensional finite element models were constructed, and all known loalings on the penetrations were analyzed. Dese loadings included internal pressure and thermal expansion, and embodied the conservative assumption that there is zero interference between the penetration and the head. In addition, residual stresses due to the welding of the penetrations to the vessel head (as evidenced by the observed ovality) were considered, using an elastic-plastic finite element analysis.
Results of the analyses reported here are consistent with the axial orientation and location of flaws which have been found in service in a number of plants, in that the largest stress component is the hoop stress, and the peak stresses were found to exist in the circumferential locations farthest away from the center of the vessel. The most important loading conditions were found to be those which exist on the penetration for the majority of the time, which are the steady state loading and the residual stresses.
Dese stresses are important because the cracking which has t
. observed to date in operating plants has been determined to result from primary water stress corrosion cracking (PWSCC).
These stresses were used in fracture calculations to predict the future growth of flaws postulated to exist in the head penetrations. Crack growth laws were developed specifically for the range of j
operating temperatures of the head for Westinghouse plants, based on information from the 1
l literature as well as a compilation of crack growth results for operating plants.
1 De crack growth predictions discussed in previous sections show that the future growth of
]
cracks which might be found in the penetrations will in general be very slow, in that a number of l
years will be required for any significant extensions, it is concluded, therefore, that it is i
conservative to assume that no "through-wall" crack will grow to a length longer than 2 in.
above the penetration to vessel weld.
i 5
i Page i1 of 35
- ~~
It is appropriate to avamina the safety consequences of an indicati:n which might be found below the weld. The indication, even ifit were to propagate through the penetration wall, would have no consequence at all, since the pressure boundary would not be broken, unless it were to propagate above the weld.
Further propagation of the indication would not change its orientation, since the hoop stresses in the penetration remain larger than the axial stresses as a crack might move up the penetration.
Therefore, it is extremely unlikely that the head penetration would be reveied as a result of any indications.
Any indication is unliksly to propagate very far up the penetration above the weld, because the l
hoop stresses decrease in this direction, and this will cause it to slow down, and perhaps even to stop before it reaches the outside surface of the head. This result from the stress analysis supports the conclusion that it is extremely unlikely that leakage of any magnitude will occur.
l The high likelih,cd that the indication will not propagate beyond the head ensures that no l
hhuyhic failure of the head penetration will occur. This is because the indication will be l
enveloped in the head itself, which precludes the opening of the crack and limits leakage. In l
order to produce a failure of the head waison, the flaw would have to extend over 13 inches above the head, an extremely unlikely event.
i i
l 1
l l
I i
I
{
l
(
Page t2 of 35 l
l
4 l
I l
I i
1 320*C(608'F) 310*C(590*F) fo 0.8 l
300*C(572*F) t 0.7 g
l
.9 0.6 5 0.5 290*C(554*F)
CL g 0.4 x00.3 0
O 2
4 6
8 10 12 14 16 18 20 Time (Years)
FIGURE 3.2-1 CRACK GROWTH PREDICTIONS FOR SURFACE FLAWS BELOW AND AT THE WELD REGION IN THE HEAD PENETRATIONS FOR A RANGE OF TEMPERATURES l
Page 13 of 35 l
t l
t b
12 t
I r e ampn,
^ 10
)
g su.s.fm.sm [
.c 4
~
.... ne
. y;~
- ,w - ;Ma,.
m a.> n ;. &
&a ht gesyj[jb..e,,..$3 4 SN.... ; Weld Region.
,:4._
Q gj$
Ecn c6 B M ira m 3
x0 4
a wQ t
t 2
0 O
30 60 90 120 150 180 210 240 270 300 Time (Months)
FIGURE 3.2-2 I
CRACK GROWTH PREDICITONS FOR THROUGH-WALL FLAWS LOCATED AT THE CENTER SIDE OF THE OUTERMOST HEAD PENETRATIONS, j
FOR A RANGE OF TEMPERATURES 1
Peas 14 of 35
t
(
(i 4
2 4
7 4
6 i
mro
\\
e 4
r5 T'EPPSEFFn N
O b< =if5isyt3ESWd7df 8@ild#
winen s:m gan e x w :"
' ~f.I@NF d
.a
- ~
n r
w~
hlase$m$m$$$$Y$N8ib$dNMN N5NN$
y.se
~:
hhif e: Weld Reg.
h ion
$@@b "4M Us 3 a
x O
L~jllEPCW<ts.J CD 2 3
astremnen 1
0 O
20 40 80 80 100 120 140 160 180 200 220 240 260 Time (Months)
FIGURE 3.2-3 CRACK GROWTH PREDICT 10NS FOR THROUGH-WALL FLAWS LOCATED t
ON THE LOWER HILLSIDE OF THE OUTERMOST HEAD PENETRATIONS, FOR A RANGE OF TEMPERATURES Pae 15 or35
____~
I
/N
/
/
//
/
/;
/
/
//',
's
\\
/
\\
/
CENTERSIDE
/
/
LOWER HTT TSIDE
/
/
1
/
i
/
/ ----
/
/
\\Ns
,/
d -- -
2 Nounter Bore V
Differential Weld Heig%
ht (Weld Offset)
Weld Length-FIGURE 3.2-4 WELD LENGTH ON HEAD PENETRATION TUBE
.i Page 16 of 35
j
3.3 ASSESSMENT
OF WOG PLANTS i
Using the stress and crack propagation analysis data and methodology provided above, an assessment of the potential for cracks in the 54 WOG plants was performed. This asscument is summarized i
below. Indication data from inspections performed at the hghals and Beznau plants is used to estimate the condition of the reactor vessel head penetrations in the WOG plants.
i f
Review of Penetration Indication Data j
A revice of the inspection results from the Ringhals and Beznau plants ha hwn performed. A
{
summary of the inspection results is provided below.
L i
l Plant Total No. of Penetrations Penetrations l
Penetrations Inspected W/ Indications l
bghals 2 65 65 6
l Ringhals 3 65 60 0
Ringhals 4 65 65 2
Beznau 1 36 22 2
h l
Beznau 2 36 27 0
l In the case of hghals 2, six penetrations were found to be cracked. The maximum depth of I
these cracks were reported to be 4 mm with a length of 16 mm. This plant had operated for approximately 108,400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br /> at the time the measurements were taken. Two of the hghals 4 penetrations were found to be cracked. Rese cracks were judged to be very shallow with a maximum length of 7 mm. Ringhals 4 had operated for approximately 75,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> when the i
measurement was taken.
)
Indications were found on two of the penetrations at Beznau Unit 1. Depth was estimated to be i
less than 2 mm and the longest indication was estimatad to be 28 mm. These indications were I
found after at least 157,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of operation.
Applicability to WOG Plants Westinghouse has completed a review of the cause of the cracking in French plants, and the key parameters associated with this situation for the Ringhals and Beznau plants, as well as selected French plants, and made a comparison of these key parameters to the WOG plants. Westinghouse has concluded that the mechanism of degradation in the reactor vessel head penetrations is primary water stress corrosion cracking (PWSCC). This conclusion was reached after an engineering review (Reference 1) of the data from the French plants. His review included consideration of metallographic records, material, environment, and penetre. tion bending and ovality.
Page 17 of 35 i
Several p-wid key factors which could impact susceptibility to PWSCC cf the reactor vess:1 i
head penetrations were investigated (Reference 2). Dese included residual and operating stresses in the penetrauon, environment, material condition, temperature, and time of operation at temperature and pressure. De review of material condition included such parameters as microstructure, heat treatment, welding procedures, chemistry, yield strength, and machining / grinding. It was concluded that a relative susceptibility could be developed. The comparison of the key parameters between the Ringhals, Beznau, and French plants, to the WOG plants resulted in the conclusions that the hghals 2 data is representative of the WOG plants.
Further, an assessment of the WOG plants can be made by comparing the relative susceptibility of those plants to Ringhals 2. De condition of the Beznau units fall within the boundaries of this model.
Estimate of Condition of WOG Plant Penetrations An estimate of the condition of the WOG plants, relative to hghals 2, was developed. This estimate was made for each of these plants by comparison of the key parametus discussed above and application of the stress and crack propagation analyses discussed previously.
Ringhals 2 was reported to have a maximum 4 mm deep crack after approximately 108,400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br /> (12.4 effective years) of operation. The head 6.+.uie during all of this time, except for an intermediate period of approximataly 11,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />, was 605.6T. As a result of plant operstmg conditions, the head 6-y e during the 11,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> was 579.91. An equivalent operation time of 100,130 hours0.0015 days <br />0.0361 hours <br />2.149471e-4 weeks <br />4.9465e-5 months <br /> (11.4 effective years) was calculated assuming the head temperature was a constant 605.6T during the entire time of operation. His equivalent operation time was determined using the relationship:
1/t oc Ave *'
(Reference 2) where:
A = A constant related to the material microstructure characteristics
- = A constant related to the material stress Q = 50,000 cal / mole R = 1.987 cal / mole *K T = head temperature (K)
It is possible to make comparisons and draw conclusions relative to the WOG plants on the basis of the Ringhals 2 experience, by determining an equivalent operation time for the WOG plants, which has been corrected for the key parameters. By comparing this equivalent operation time for each plant, to the operation time of Ringhals 2 at the time the cracks were found, estimates can be made about the condition of the head penetrations in the WOG plants, relative to the condition of the Ringhals 2 head penetrations. He equivalent operation times were developed using the expression provided abcve to account for the key pcrameters, as described below.
Page 18 of 35
Experimental evidence based cn the tests conducted ovIr the past several years suggests that PWSCC susceptibility of Alloy 600 is strongly sensitive to, among other factors, ths '
microstructure of the matarial. It was further established that the extent of grain boundary coverage by carbides plays a very significant role in controlling the cracking behavior of this alloy. It is generally established that increasing the grain boundary coverage increases the resistance to PWSCC. In a recent systematic study of microstructure versus crack initiation time en Alloy 600, it has been shown that the crack initiation time can increase by a factor of five when the grain boundary carbide coverage is increased from zero to 100%. This observation is also found to be very consistent with the recent cracking experience of reactor vessel head penetrations in French plants where the microstructural examinations conducted on the material from ninmaan penetrations which exhibited cracking, showed very little or no grain boundary carbide coverage in every case.
A microstructural study of eight samples of material repmenting typical WOG plant head penetration material was conducted to ernmina the grain boundary carbide coverage. This population of samples included two heats of matrial which are also representative of the Ringhals 2 material. De results showed that at least five of the eight samples exhibited good carbide coverage. Of the three samples which did not exhibit good carbide coverage, two were samples of the Ringhals 2.w.
ive material. The results of these examinations were compared against data representative of the head p
. ion material in the French plants. A quantitative conaparison of the grain boundary carbide coverage suggested a factor ranging from 3 to 5 increase in crack initiation time relative to the worst case microstructure for the French plants. Additionally, the data indicates that a factor of at least 3 could be applied to most of the l
WOG plants for crack initiation time, relative to Ringhals 2. However, in vien of the relatively small sample size, a factor of I was conservatively used in the assessment of the WOG plants relative to Ringhals 2 and therefore the superior structure was not taken credit for in this evaluation.
Two key parameters were ce --- J for with the stress constant, o. First, a review was made of material certifications in the Quality Assurance data packages for all of the plants to determine the yield stress of the head penetration material applicable for each plant. It is generally accepted that material yield stress is a factor in the s==911ity of a material to PWSCC, higher yield j
stress material being more susceptible. A constant determined by the ratio of the WOG plant yield stress to the Ringhals 2 yield stress, raised to the fourth power, was then determined for each WOG plant. A second constant was determined for each plant to account for the level of I
operational and residual stresses applicable to each WOG plant, as compared to the dinghals 2 plant. Stress analysis and ovality measuramante of the head penetrations have shown that the level of stress in the outer penetration (highest stressed) of a four loop plant is higher than that of a two or three loop plant. Since Ringhals 2 is a three loop plant, this second stress constant was determined for the four loop plants by taking the ratio of the ovality measu' red at four loop plants l
to the ovality measured at a three loop plant, raised to the fourth power.
Finally, a correction was made for the head ep..iug. ym.nire by determming the WOG plant " equivalent" operating time, if it were operating at the Ringhals 2 head temperature.
With all of the key parameters =-:-x =-i for as des ribed above, an equivalent operating time was determined for each of the WOG plants. Rese operating times were then compared to the operating time of Ringhals 2, at the time the 4 mm deep crack was discovered, corrected for the i
Page 19 of 35 l
,m
11,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> cf operation at the Inwer temperature as described abova. Those plants with a j
lower number of equivalent operating hours are considered to be less susceptibts to PWSCC than 1
Rhighals 2. Likewise, those plants with a higher number of equivalent operating hours than Ringhals 2 may be more susceptible to PWSCC.
I Results and' Conclusions It was concluded that in terms of relative susceptibility to PWSCC, the Ringhals 2 condition envelops 45 of the 54 WOG plants. It is therefore expected that none of these 45 WOG plants l
would have cracks deeper than 4 mm, if there are cracks present at all. Funber, it is estimated based on the crack propagation analysis described in Section 3.2, a 4 mm crack would not l
propagate to become a through wall crack in any of these 45 plants within an additional time l
period of 14,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> (2 years of operation at 8% availability).
Conservatively, it was concluded from the assessment of relative susceptibility to PWSCC, that 9 1.
{
of the WOG plants may be more susceptible than Ringhals 2. These results suggest that these plants may have partial psion cracks deeper than the 4 mm crack found at Ringhals 2. In i
j addition, conservative predictions indicate that through-wall cracks at or below the weld may exist in some of these plants. In these cases, however, crack extensions above the weld would be i
less than 1.0 inch at the end of an additional operational period of 24 months. It should be noted, as can be seen in igure 3.2-3, crack growth rates decrease signifk:antly above the weld. Section
)
3.5 of this evaluation summarizes the wastage assessments. Conclusions concermng plants with W. dons cracked above the weld are made in that section.
i i
3.4 LEAK RATE CALCULATIONS FOR THE REACTOR VFAML HEAD PENET j
i i
De purpose of this analysis is to estimate the leak rate between the reactor vessel head pene
}
and the reactor vessel head when an axial through-wall crack is postulated in the former. A
=eh=atic of the vessel head penetration and vessel head portion associated with this region is shown l
in Figure 3.4-1. To determine the leak rate through this annulus, the leak rate through the head l
penetration crack is first estimated. His value is then compared to the leak rate through the based on a choked flow rate. De lower of these two rates is taken as the leak through the annulus.
2 De clearance between the head penetration and the vessel is effected by the different thermal expansion coefficients of the base metal and the weld metal, internal pressure, and "as-fabr interference fit of the penetration.
De pressure inside the head penetration was 2250 psia. De calculations were performed tha to the entire temperature range of operation for the WOG plants (550'F to 620'F). The outside ra and thickness of the tube are 2 in, and 0.622 in. respectively. In this analysis, the clearance between the head penetration and vessel is considered to be a variable parameter, and it is assumed to b uniform through the direction of the head penetration axis.
The leak rate through the ares of concern is estimawl using two systems models. The first lesk rat is estimated through a postulated axial through wall crack in the head psnetration. Second, a leak is calculated through the tight annulus corresponding to the clearance 1etween the head penetra and the vessel. De water inside the head penetration is in the subcooled liquid phase. The leak ra through the postulated crack in the head penetration is estimated using a two phase flow mo Page 20 of 35 6
j i
(Reference 1). The leak rate through the clearance between th2 tube and the vessel is est a single phase flow model (superheated steam). If subcooled liquid leaks from thm postulat the head penetration, the pressure will drop radically due to crack friction effects. He temperature, however, of the fluid would not drop as quickly as the pressure, since the crack is surrounded by a large heat source containM in a structure such as the vessel. Hence, the flow in the annulu single phase. This fluid injected through the postulated crack of the head penetration would ch steam and would leak through the tight annulus clearance of the vessel. The leak rate of steam is estimatM using the choked flow rate of steam.
Mass conservation is assumed at the exit of the head penetration crack and at the exit of the vessel.
Since the interaction of fluid flow resistance between the head penetration crack and the tight annul If the flow rate through the clearance of the vessel is not known, an engineering assumption is made:
head penetration is smaller than that through the annulus, the flow rate is considered to be the rate through the head penetration. Alternately, if the flow rate through the head penetration is lar than that through the annulus, then the flow rate is considered to be the flow rate at the annulus.
implies that flow resistance imposed on the flow rate through the head penetration crack is the very tight clearance between the penetration and the vessel. His results in the reduction o ft:w rate from the besl penetration crack.
Using the engineering assumption described above, a numerical calculation is performed. De stagnation pressure in the clearance between the head penetration and the vessel is consi equal to the choked pressure at the exit of the head p idion crack. De stagnation temperature in the annulus is assumed to be the same as the vessel e + nore (600*F). De ratios of heat capacity C,/C, are obtained from the Mollier diagram s.t the choked pressure and temperature pressure is a function of the crack geometry.
The crack apaning area of the head penertation and leak rates are calculated for various cirw=J-alal crack lengths. Dese values are tabulated using both British and SI units. Mass flux l
for various axial crack lengths and heat capacity ratios are also given in both British and SI units.
The clearances in the axial direction are obtainM using a finite element nW 1. The leakage throug these clearances are tabulated and provided along with tables showing calemated leak rates, it can be concluded that generous margins exist for all the plotted leakage size flaws. The annular clearance between the head penetration and the reactor vessel during operation is estimmtM to be l
0.003 inches. Therefore, since the crack extensions above the weld are==ead o be less than 2.0 t
j inch, as indicated in Section 3.3 of this evaluation, a review of Figure 3.4-2 provides that the maximum leakrate which can be -aaaad is 0.70 gpm. Laage in excess of 1.0 gpm is detectable l
WOG plants.
For the case where the "as-fabricated" penetration ini f= vise fit is at the high end of the tolerance, the annulus between the head penetration O.D. and the corresponding vessel hole I.D. is reduc l
very small levels during plant operating conditions. As such, the leakage will be smaller above value for penetration cracks lengths equal to that of the vessel head thickness. It is co even less likely that cracks would grow to this length because of the compressive stress in the penetration. If crack growth beyond the outside surface of the head is considered, leak detected prior to the crack reaching the critical flaw size of 13 inches in the penetration.
1 i
)
Page 21 of 35 I
i
Table 3.41 hak Rates Through a Range of Crack Emths: Head Penetration Crack Leak Rate Choking Pressure Crack Length Opening Area Ib/sec psi 2
kg/sec (kg/cm )
CL in2 (cm )
2 1.0 in 0.0004117 0.0103 381 (2.54) cm (0.002656)
(0.004672)
(26.7874) 2.0 0.00246 0.0968 5%
(5.08)
(0.01587)
(0.04391)
(41.9037) 3.0 0.00922 0.5253 846 (7.62)
(0.05948)
(0.2383)
(59.4807) f 3.5 0.01622 1.1115 979 (8.89)
(0.1046)
(0.5042)
(68.8317) 4.0 0.02721 2.3772 1138 (10.16)
(0.1755)
(1.0783)
(80.0107) i j
Page 22 of 35
Table 34-2 Crack 19: Mass Mux and Heat Capacity Ratio in the Annulus Clearance Between Penetration and Vessel Cl gA in.
Iblin:
K=Cp/Cv 2
1.2885 1.0 4.7677 lb/sec in 2
(2.54) em (0.3352) kg/sec. cm L
2.0 7.4409 1.2800 (5.08)
(0.5232) 3.0 10.5354 1.2709 (7.62)
(0.7407) 3.5 12.1750 1.2%)
(8.89)
(0.85600) t i
4.0 14.1285 1.2600 (10.16)
(0.9933) 1 I
l l
l l
l l
i
)
i i
Page 23 of 35
Table 3.4-3 Imk Rate Thuwagh the Annular Clearance Between the Head Penetration and Vessel '
Crack Length e
8~
A Cf = 1.0 in.
2.0 3.0
- 3.5 4.0 (2.54) em (5.08).
(7.62)
(8.89)
(10.16) 0.0001 in 0.0012567 in' O.0060 lb/sec 0.009400 0.01300 0.0153 0.01780 (0.000254) cm (0.008108)cm (0.002722) kg/sec (0.004264)
= (0.005897)
(0.006940)
(0.008074)
O.0013 0.01634 0.07790 0.1216 0.1722 0.1990 0.2309 (0.003302)
(O.19542)
(0.03534)
(0.05516)
(0.078II)
(0.09027)
(O.1047) i 0.0019 0.02389 0.1139 0.1777 0.2517 0.2908 0.3375 (0.004826)
(0.15413)
(0.05167)
(0.08060)
(O.I142)
(0.1319)
(0.1531) f 0.0032 0.04024 0.1919 0.2995 0.4240 0.4900 0.5686 l
(0.008128)
(0.25%I)
(0.08705)
(0.1359)
(0.1923)
(0.2222)
(0.2579) l O.0033 0.04150 0.1979 0.3088 0.4373 0.5053 0.5864 (0.008382)
(0.26774)
(0.08977)
(0.14007)
(0.1984)
(0.2292)
(0.2660) i 0.0054 0.06795 0.3240 0.5056-0.7159 0.8273 0.9600 i
(0.013716)
(0.43839)
(0.1470)
(0.2293)
(0.3247)
(0.3753)
(0.4355) i k
Numbers in the parentheses are SI units I
h I
f f
1
- 5 1
Pos 24 or35 i
~
TaWe 3.4 4 Final Imk Rat 2 at else Exit of Ammahms Clearanee Between Head Penetraties and Vessel
}
Clearance Cf = 1.0 in.
2.0 in.
3.0in.
-3.5in.
l 6(in)'
Q, IWsec GPM*
Q, IWsec GPM*'
- Q, IWsee GPM*
-Q,IWsee GPM*
l 0.0001 0.00600 0.043 0.0094 0.06700 0.013-0.095 0.0153 0.11 (0.0025) mm (0.002722)*
(0.1633)***
(0.004264)
(0.2558)
(0.005897)
(0.3538)
(0.006940)
(0.4164)
?
0.0013 0.010261 0.074 0.09676 0.700 0.1722 1.24
-0.1990 1.43 (0.0330) mm (0.004654)
(0.2793)
(0.04389):
(2.6334)
(0.07811)
(4.6866)
(0.09027)
(5.4160) 0.0019
'O.010261 0.074 0.09676
.0.700 0.2517 f.81 0.2908 2.1 l
l (0.0483)
(0.004654)
(0.2793)
(0.04389)
(2.6334)
(0.1142)
(6.8503)
(0.1319)
(7.9144) j 0.0032
'0.010261.
0.074 0.09676 0.700 0.424 3.05
.0.4900 3.5 (0.0813)
(0.004654)
(0.2793)
(0.04389)
(2.6334)
(0.1923)
(i1.5396)
(0.2222)
(13.3358) f f
0.0033 0.010261 0.074 0.09676 0.700 0.4373 3.14 0.5053-3.6 (0.0838)
(0.004654)
(0.2793)
(0.04389)
(2.6334)
(0.1984)
(11.9016)-
(0.2292)
(13.7522) i 0.0054 0.010261 0.074 0.09676 0.700 0.52531 3.78 0.8273 5.9 I
(0.1372)
(0.004654)
(0.2793)
(0.04389)
(2.6334)
(0.2383)
(14.2968)
(0.3753)
(22.5158)
{
i
- at room temperature condit~mns I
- kg/sec
[
l
- liter / min I
.t
)
f i
1 I
i res. 25 of35 i
t 5
F
Table 3A-5 Final 14ak Rate at tIse Exit of Annahms Clearance Between Head Penetration and Vessel
-t Clearance Cf = 4.0 in.
8(in)
GPM*
Q,Ib/sec 0.0001 0.0178 0.13 (0.0025)
(0.008074)
(0.4844) i 0.0013 0.2309 1.7 (0.0330)
(0.1047)
(6.2842) 0.0019 0.3375 2.4 (0.0483)
(0.1531)
(9.1854) 0.0032 0.5686 4.1 (0.0813)
(0.2579)
(15.4750) 0.0033 0.5864 4.2 (0.0838)
(0.2660)
(15.9595) l
)
0.0054 0.9600 6.9 (0.1372)
(0.4355)
(26.1274)
- at room temperature conditions i
^
[
I L
P s 26 et 35 r
a k
I
l
./
/
/
/#:
/e
/
/
)
/
/
'N Location of Axiol Cro
/
/
/
Penetfelien Wald
/
t CADhl Thermal Sleeve i
t i
Head Penetradon and Vessel Assembly Figure 3.41 l
l I
Page 27 of 35
4 30 jn,j,,nn,,,uolniliingna opunnneinitusunnnnini n i un-null i
i I
i l iiij o
"N'
' O I
U e
6=
li l !I I
i l l
n 2nR(rii/A)
I i i
i j 4.0 in.
(10.16 cm) !
'"' Flaw Orientation in CRDM:
I Longitudinal Flaw-T = 600*
l c
l Critical Flaw Size: 13 in.
3.5 in.
3 l
l lH ll l l l
(8.89 cm)
, nimii,.m l
C I
I o
i N 20 e
6 e
,1i iii s..u, m,,
a i
f e
o o
I i
l m
3.o in.(7.az em) e g-f o
o i
3 o
s w
o
/
y g
e niiii..a
($
l l
g
.g
,i i
g 10 j
j j
g
/ /
v tg o
/
e x
J
)
f I
2.0 in.(5.08 cm) o usi m
i in i
l 6 "
1.0 in.(2.54 cm)
,,,,,mm,
--m i
O i
,,,,,m,,,,,,,,m.m..,m,m....,,m, i
0 0.02 0.04
, 0.06 0.08 0.1 0.12 0.14 clearance (mm) i 1
i Iask Rate as a Function of Clearance and Crack 19 in Terms of Liters per Minute at the Exit of Annulus Clearance Between Head Penetration and Vessel Figure 3.4-2 Pye 28 of 35
...._ _ _. _ _ _. _ _.... _.. _ _ _ _.. _ _. _ _ -.... _. _ _.. _ _ _. _. _.. _ _.... _ _. _.. _ ~.
i 1
i J
}
i 4
i i
I f
j 8
- 1 y
f o
s.
e d
Mm/A) 7 cs
- d'0 I"l 4
+
Flaw Orientation in CRDM:
I f
j z Longitudinal Flaw-T = 600*
/;
_a Cdtical Flaw Size: 13 in.
,.m.m.
E i
- L L y
6
- c,, 3 s ini 7
1
+,.
E J
F I t
F A :
[
I I
I I I I E I E y
4 1
P I
I F
I 1y
,?'
. uv Q
- E :
J i1
- 11 l I
i 1
g I
1 1
]
f y I 1
[2 l
1
[
I I
1 FI E
i G
- 4 O
Cf = 3.0 inI a
1
]
J F
l 2
V J
P Y I I
I 171 P
E I I
A -
I F
CU 3 F
1
! f 1
]
A E
F J
I 1
l I
r j
/
Mmum:
I I
E 2
2 A
I J
J J
I I T
I i
2 El F
]
I l I 1
I J
F 1
I V
I q
4 I
r 7
3 i
1
/
, :cs. 2a ini
_m 1
-J W.'A AV
' cr. 1.o ini 3
0 1
2 3
4 5
6 x 10 1n.
Clearance (6)
Lenk Rate as a Function of Clearance and Crack Lengths in Terms of GPM at the Exit Annulus Clearance Between Head Penetration and Vessel Figure 3.4 3 Page 29 of 35 l
3c5 REACTOR VESSEL HEAD WASTAGE ASSESSMENTS General Technical Dimen==lan v
The purpose to this section is to conduct an assessment of the potential wastage (i.e. pitting, and wall thinning by general corrosion) of the reactor vessel head due to the leakage of the boric acid coolant through a postulated axial through wall crack in the Alloy 600 head penetration. The wastage assessment considered here it based on the existing wastage data obtained from the laboratory test programs conducted at Westinghouse and the results of a penetration mockup test conducted under a Combustion Engineering Owners Group (CEOG) program, Reference [15].
For the current wastage considerations, under steady state operating conditions, it is assumed that the top of the reactor head is maintained at approximately 500*F while the top of the insulation above the vessel head is maintained at approximately 150*F. Under these conditions, the coolant leak through the penetration will leave the penetration and the counter bore annulus in the form of (superheated) flashing steam leaving behind a " snow" of boric acid crystals in the crevice and at the top of the vessel head around the penetration. The majority of the boric acid crystals formed in the crevice are expected to be slowly pushed out to the top of the vessel head by the exiting steam. Westinghouse laboratory test results showed that boric acid crystals heated to 500*F contributed to no or negligible wastage of carbon steel. Due to the high temperature environment in the crevice region, the wastage in the crevice region due to steam moisture is a' mad to be minimal. Any occurrence of wastage at the vessel head would require a re-wetting machaniam of the dry boric acid crystals deposited on the top surface of the vesse!
liend. His re-wetting mechanism would require a condition whereby as the steam escapes tirough and above the insulation, it encounters lower temperatures in the range of 150*F to 212*F where it starts to condense into moisture, a fraction of which would potentially find its way back to the vessel top surface through the available flow paths in the insulation. This, of course, could create a wetting condition of the boric acid crystals deposited on the top of the vr.ssel head. Since the vessel head is maintained at near 500'F, any wetting is armad o be t
i mimmal if at all. Conservatively, a continuous leak could establish a wetting and dryout condition of the boric acid crystals on the top of the vessel head resulting in some wastage at the crevice mouth region.
Laboratory tests conducted at Westinghouse showed that aqueous boric acid solutions caused
)
carbon steel to corrode at rates dependent on the concentration of boric acid in solution at any given temperature. Low concentrations (approximataly 1500 ppm boron) produced corrosion rates on the order of 5 to 10 mils per month, whereas concentrations of 25% by weight of boric acid removed carbon steel from a specimen at a rate of approximatal) 400 mils per month at l
200*F. A concentration of 25% by weight of boric acid is saturated at about 200*F. Galvanic l
corrosion between carbon steel and Inconel400 appeared to contribute little to the carbon steel l
attack in aqueous boric acid solutions. The boric acid crystals, when heated, dehydrated to form l
B,0, a glass-like substance which coated the specimens and apparently protected them from 3
I atmospheric oxidation. De glass-like material is, however, a potentially corrosive agent because it hydrolyzes in hot water to form boric acid. A 25% by weight of boric acid solution, de-aerated for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> with a nitrogen sparge, produced an average corrosion rate of 250 mils per month. A 25% by weight of boric acid solution containing dissolved oxygen yielded a corrosion rate of about 400 mils per month. De presence of insulation was found to inhibit the corrosion of carbon steel at 200*F.
i Page 30 of 35
De space between the vessel head top surface and the insulation is Gxpected to be in the temperature range of 500*F to 350*F (on a conservative basis) so that active steam condensation l
is not likely to occur here and the majority of steam is expected to escape through the clearance i
at the head penetration. His creates the situation where by the boric acid crystals are left below the insulation at the penetration while the majority of steam condensation is occurring above the insulation. Under this condition, only a fraction of condensed steam can reach the boric acid l
crystals located at the annulus of the affected penetration to re-wet the crystals and boil away, creating an intermittent wetting and dry out conditions. The above conditions are expected o significantly moderate the maximum observed wastage rate of 400 mils per month achievable in the laboratory under complete aqueous, concentrated, and full oxygenated condition of the boric acid at 212'F.
As indicated in the previous section, the leakage rate to be considered through the largest axial "through-wall" flaw expected in a WOG plant penetration, estimated at two inches in length, is i
apprortmataly 0.7 gpm at an average annular gap of 0.003 inch. In comparison, the maximum i
leak rate expected for a "through-wall" flaw of approximately 1 inch is 0.074 gpm for the same l
annular gap. His value is consistent with the leak rates experienced in the GOG test program for comparably sized cracks.
j 1
An assessment of realistic wastage rates achievable at a head penetration can be made from the results of the recent mockup test conducted by the GOG to establish wastage rates due to the leakage at the bottom of a pressurizer. Key parameters of the GOG test are; a) leak rates 1
ranged from 0.026 gpm to 0.119 gpm, b) the diametral crevice clearances ranged from 0.0015 inch to 0.0099 inch, and c) the block (head) 4&res ranged from 351 *F to 566*F.
J Dese parameters are consistent with the conditions found at the WOG reactor vessel closure penetrations. De relevant results of the test can be summarized as follows:
1.
Although the maximum penetration rate (at the deepest pit) observed was 2.15 inch / year at a localized region, the maximum average penetration rate achieved was 0.0835 inch /
year.
2.
The maximum total metal loss rate (wastage volume) observed was 1.07 in'/ year.
3.
De greatest damage occurred almost entirely where the leakage left the annulus.
He &OG mockup test results are judged to represent a conservative estimate of the wastage rates which could be expected due to the leakage at the vessel head penetration. High condensation in the GOG test is postulated, as compared to the head penetrations; 1) due to the i
close proximity of insulation in the test, and 2) it is postulated that exiting steam would have been expected to rise back to the pressurizer head surface rather *han escaping away from the surface, due to the inverted geometry of the simslated pressurizer test configuration. The i
anticipated higher condensation would serve to maintain a relatively moist environment, i.e.,
rewetting; thus resulting in conservative wastage rates as compared to the head penetration geometry. Recall that leak rates range from 0.074 gym to 0.7 gpm for maximum expected crack lengths of one to two inches depending on the circumferential position about the weld. Existing Page 31 of 35
in plant leak detection capabilities are limited to 1.0 pm er bl her. Wus a flaw which results in 3
J leakage equal to or greater than 1.0 gpm can be detected and addremed appropriately. Dus, it is only necessary to associate a wastage rate with leak rates ranging from 0 to 1.0 gpm. Based on the conservatisms jMged to be in the crack length determination, the leak rate assessment, and the CEOG test data, the 1.07 in'/ year metal loss rate was selected as an appropriate value for use over the 0.0 to 1.0 gpm leak rate.
Analysis of Reactor Vessel Head T o dimensional finite element analyses of 2,3, and 4 loop reactor vessel heads were performed to assess the impact to the structural integrity of the reactor vesse.1 head of the wastage discussed 5
above. As discussed above, the wastage rate considered was 1.07 in / year. Six years was chosen as the time period which the wastage at this rate would occur undar tad. This would result in a total loss of approximataly 6,4 in' of vessel head material. The CEOG test data indicates that the wastage would be very localized, in a very small area where the leak exits from the annulus between the reactor vessel head and the head,u
. son. As indicated by the head i
penetration stress analysis discussed in Section 2.0, and experience in the French plant at which leakage actually occurred, the leakage would exit the annulus on the up-hill side of the pu. ion. His is because the annulus between the reactor vessel head and the penetration tends to open on this side of the penetration as the reactor vessel is pressurized during plant operation.
Based on the CEOG test data, two defect shapes were postulated to umbrella the various possible wastage defect shapes that could occur. One shape considered was a~ defect approximately 2.0" wide racially X 1.0" wide circumferentially X 3.2" deep (into the thickness of the head). The second shape considered was a defect approximately 1.07" wide radially X 1.0" wide circumferentially and extends through the vessel head thickness to the head nemetration to vessel head weld, ne two dimensional finite elamant analysis was performed to assure the structural integrity of the ve:::! head is maintainad. His calcala6on is an allowable alternative to the ASME Code Section III conservative sizing rules brud on nozzle reinforemmant and ligamant efficiency calculations. For each plant size, analyses were performed on the reactor vessel head in the normal or as manufactured condition, as well as with the above described defects introduced, and comparisons of the results were made to draw conclusions on the impact of the wastage. He wastage defects were introduced into the models as described below.
Each model developed was a two dimansional axief esic model representing the reactor vessel head and including the effects of the head adapters, especially near the expected location of possible leakage; and an approxunation of the effect of the vessel head flange. The input parameters for these taodels were taken from the vessel design reports representative of 2, 3 and 4 loop plants. De information required inc!udes the inside radius of the head to the base metal, the thickness of the vessel head (base metal), the locations of the p,
. ions, the diameter of the pa. ions, (all 4.0 inches) and informatica describing the configuration of the vessel head flange. We code used for evaluation of this model is the WECAN-Plus computer code which is proprietary to the Westinghouse Electric Corporation. Various finite elements are available for evaluation, with element 53, the "Two Dimensional Isoparametric", being selected. His is a 2-D Quad element and is considcred with axisy.
esic properties. Furthermore, midside nodes were also considered to refine the displacement function to consider quadratic edges (of the quads). A minimum of five elements were selected through the thickness of the hevi, providing for eleven integration points through the thickness.
Pes. 32 or 35
De pitch between adjacent adapters was considered as 11.035 for the 2 loop plant and 11.973 mehes for the 3 and 4 loop plants. The elements represenung the base metal of the head were identified as matarial property #1. At each location of a penetration, starting with the one at the centerline of the vessel, and spaced according to the pitch dimension mentioned above, the material properties were identified as property #2. The material properties for the adapter region were modified to consider the hole penetrating the base material. The hole properties were adjusted based upon a ratio of the adaptor diameter (4 inches for all p' ants) to the adapter pitch (pitch of holes in head).
The wastage is introduced at the inside of an outer most adapte. starting at the outside of the vessel head surface. De material property for the wastage is identified as material #3. For the 4 loop size vessel head, several different defect geometries were evalus'ed to envelop the two t
defect sizes described above. These included a 2X2 element defect, a IX4 element defect, a 2X4 element defect, a IX5 element defect and a 2X5 element defect. It was concluded from the review of the results of these cases that the 2X2 element case and the 2X4 element case conservatively envelop the two representative defects described above. Dese two cases, along with the normal head configuration case (no defect) were also run for the 2 loop and 3 loop size heads.
The !=ad=y condition was applied to this model at a node near the gasket seal of the head (junction with the vessel). Since the axisy wic option was selected, the boundary condition at the centerlin need not be specified. A uniform pressure loading of 2500 psi was applied to the inside surface of the head.
The stresses in the cantar of the ligament between the -p ion hole with the defect and the adjacent penetration hole were w y.4 for each of the three cases, for each size head. De largest increase in stress intensity between the normal head configuration and the head j
configuration with the postulated wastage defect is only 6.5%. With this minor incasse, the general primary membrane stress intensities remain below the corresponding ASME Section III allowable stress intensity limit (S,).
In the unlikely event that a leak would develop in a WOG plant reactor vessel head penetration, and continue uMat~*M for a period of time of up to six years, the wastage that would result on the vessel head is a====8 to be local to the immediate. area of the penetration. It is conservatively estimarM that the low alloy steel in the vessel head would waste at an approximate rate of 1.07 in' per year or 6.4 in' after six years of n=tarareM leakage. Analysis of the vessel head in this degraded condition conclusies that the stresses remain within the ASME code allowables and therefore the structural integrity of the reactor vessel head would not be jeopardized. His conclusion is applicable to 2 loop,3 loop, and 4 loop size reactor vessel heads.
4.0 DETERhENATION OF UNREVIEWED SAFETY QUESTION 1.
Continued plant operation with the siti:ation as described in this evaluation does not increase the probability of an recident previously evaluated in the FSAR. Inasmuch as catastrophic failures of vessel head penetrations are not a==*~i, and any postulated through wall crack l
Pese 33 of 35
-^-
would caly lead to a minimal amount of leakage, the accident scenarios a pr.a. sed in the FSAR are not impacted. Concerning the question of wastage, this evaluation has shown that over six years of operation is possible without impacting plant safety even with undetects leaking penetrations.
2.
De consequences of an accident previously evaluated in the FSAR are not increased due to l
continued plant operation. The preceding safety evaluation has shown that the reactor coolant system is not challenged in such a way as to deleteriously affect continued operation.
As described above, catastrophic failures of vessel head penetrations are not av;w'M, and any postulated through wall crack would only lead to a minimal amount of leakage.
i Wastage issues, for the plants most susceptible to postulated through wall cracks, have been shown not to affect plant operability for over six years. Herefore, the conclusions presented in the FSAR remain valid such that no more severe consequences will result from an accident condition.
i 3.
Continued plant operation will not create the possibility of an accident which is different i
than any already evaluated in the FSAR. No new failure modes have been defined for any j
system or component important to safety nor has any new limiting single failure been identified. Derefore, the possibility of an accident different than any already evaluated is not created. The postulated leaks and wastage issues, evaluated herein, will not create an accident different than any previously evaluated in the FSAR.
{
4.
Cominued plant operation will not ircrease the probability of a malfunction of equipment important to safety. Potential cracking, postulated leaks and wastage issues, as presenied in j
this evaluction, will not cause the malfunction of equipment important to safety.
5.
Continued plant operation will not increase the consequences of a malfunction of equipment important to safety previously evaluated in the FSAR. De preceding safety evaluation has concluded that this situation will not adversely affect the reactor coolant system in such a way as to affect the au-*M consequences of the malfunction of any equipment important to
(
safety.
6.
Continued plant a ration will not create the possibility of a malfunction of equipment important to safety different than any already evaluated in the FSAR. Catastrophic failures of vessel head, a. ions are not avW, and any postulated through wall crack would i
only lead to 1. minimal amount of leakage. As such, the malfunction of equipment important to safety is r.ot avW 7.
The evaluation for tne effects of continued plant operation with potentially cracked reactor vessel head adaptors has taken into account the applicable Technical Specifications. De preceding safety evaluation has concluded that design and safety information, as presented in the FSAR, remams bounding for all plant operational conditions. As such, the margin of safety, as defined in the bases to the Technical Specifications and demonstrated by the safety analyses, will not be reduced.
5.0 CONCLUSION
S Based upon this evaluation and the engineering analyses and assessments performed pursuant to the Westinghouse Owners Group program regarding Reactor Vessel Head Adaptor Cracking, it is Page '>4 or 33 l
f concluded etct catastrophic failures cf reactor vessel head adaptor tubes will not occur inasmuch as circumferential cracking is not arpae*ari to occur and any potential axial flaw will not propagate to the point at which it reaches a critical flaw size. Additionally, it is considered extremely unlikely that vesal wastage, as described in this evaluation, could continue undetected for a six year period.
j Further, the supplamantal plant operating requiramant< stated in NRC Generic Letter 88-05
. (R.L --.10) requiring walkdown inspections looking for visible boric acid deposits reduce the likelihood that any such situation would remain n=taractari. Accordingly, it is concluded that this i
situation does not represent an unreviewed safety question per the definitions and requirements I
l delineated in 10 CFR 50.59 (a)(2).
3 t
j As determined from this evaluation, the wastage issues contim= to be the most limiting factor.
Catastrophic failures of head adaptors are not - ;==4. Potential cracking is not aW to have l
l progressed through wall above the anachmant weld, but if such is postulated, the calculated leak rates i-are minimal. Though considered unlikely due to the reduced stress levels in the penetration tube above the weld region, if a through wall crack were to propagate to a significant level above the weld, laak rates would increase to a riaearrahle level. Potential cracks are not a=W to reach the i
j postulated critical flaw size of 13 inches.
i t
6.0 RFTERENCES t
i i
j 1.
" Root Cause of French Reactor Vessel Head Penetration Cracking," WOG Program MUHP-j 5016, MED-PCE-12224, June,1992.
I 2.
Reactor Vessel Closure Head PenetTWi9n Kpy Paramaser Comparison. WCAP-13493, j
September,1992.
d 3.
Fauske, H. K., " Critical Two-Phase, Steam Water Flows," Proceedings of the Hear i
Transfer and Fluid Machania lastitute, Stanford, California, Stanford University Press, i
j 1%1.
j 4.
Shapiro, A. H., 'Ibe Dv===ia==d 'Iber=ariv===in of r%=n==ihle mid Flow. Ronald j
Press.
i 5.
Eiber, Pressure Vessels and Piping. Volume 94, pp. 99,100.
6.
WCAP-13525, Alloy 600 Reactor Vessel Head Penetration Cracking 7.
WCAP-13426, "RV Closure Head Penetration Finite Element Stress Analysis" l
8.
Scott, P.M., "An Analysis of Primary Water Stress Corrosion Cracking in PWR Steam l
Generators," in Proceedings, Specialists Meeting on Operatmg Experience With Steam Generators, Brussels Belgium, Sept.1991, pages 5, 6.
9.
" Corrosion and Corrosion / Erosion Tesdag of Pressurizer Shell Material Exposed to Borated Water", CE-NPSD-648-P, April 1991.
- 10. NRC Generic letter No. 88 05, " Boric Acid Corrosion ci Carbon Steel Reactor Pressure Boundary Components in PWR Plants", dated March 17, 1988.
Pes.35 of 35 w
-,,n.
v.,.
1 I
l ATTACHMENT C t
i i
i
umTso STATES 5
NUCLEAR REEULATORY COMMISSION
(*****
waswineron, o.c. msess.essi
)
l November 19, 1993 l
William Rasin, Vice President l
Director of the Technical Division Nuclear Management and Resources Council i
1776 Eye Street, N.W.
I Suite 300 l
Washington, D.C. 20006-3706
Dear Mr. Rasin:
The attached safety evaluation was prepared by the Materials and Chemical Engineering Branch, Division of Engineering, Office of Nuclear Reactor Regulation, on the NUMARC submitta' of June 16, 19g3, addressing the Alloy 600 i
Control Rod Drive Mechanism (CRDM)/ Control Element Drivo Mechanism (CEDM) i presscrired water reactor vesse head penetration cracting issue. This j
subwittel addressed stress analyses, crack growth analysus, leakage i
assesseauts, and wastage assessments for potential cracking of the inside l
diameter of CRDM/CEDM nozzles. Basad on the overseas inspection findings and f
l the review of your analyses, tiin staff has concluded tha,t there is no immediate safety concem fer cracking of the CROM/CEDM penetrations. This j
finding is predicated on the performance cf the visual inspection activities requested in Generic Letter 88-05. Also, special nondestructive examinations are scheduled to commence in the Spring of 1994 to confira your safety analyses for each PWR owners group.
Your submittals for tach PWR type did hot add:sst the 'egey-3 flaw that was criented approximately 30* cff the vertical axis nor a circumferential, J-groove flaw discovered st Ringhals.
Preliminary information supplied to the staff by Swedish authorities indicates that the J-groave flaw may be associated with a fabrication defect. We are ceni.ineing to work with the Swedish authoritiss to contim this.
From the infomation available to us tcday, neither of these f1sws would pose a threat to tr.e integrity of the CRDM lienetrations.
It is our understanding that you are also reviewing these flaws and you will provide your assessment as to their significance and origin.
NRC will issue t supplemental safety evaluation after reviewing your supplemental assessment.
The staff agrees that there are no unreviewed rafety questions associated with CRDM/CEDM penetration cracking. The staff agreos that the flaw predictions based upon penetration stress analyses are in qualitstive agreement with inspection findings.
However, the stress analyscs ao not address stresses These stresses, if large, g of CRDM penetration t !bos during fabrication.
from possible straightenin could result in circumferential flaw orientations.
The staff requests that you also address this issue !n your supplemental assessment. Based upon infomation received from on:rseas regulatory authorities, your analyses, and staff reviews, the.itaff believes that catastrophic failure of a penetration is extremely unlikely.
Rather, a flaw would leak before it reached the critical flaw size and would be detected during periodic surveillance walkdowns for boric acid leakage pursuant to Generic Letter 88-05.
However, the staff recommends that you consider C M f (9 Q
Willian 9asin
- enhanced leakage detection by visually examining the reactor vessel head until either inspections have been completed showing absence of cracking or on-line i
leakage detection is installed in the head area. The staff requests that you also address _the issue of enhanced leakage detection in your supplemental assessment.
4 The NRC staff has reviewed your July 30, 1993 submittal, which proposed flaw acceptance criteria to be used in dispositioning any flaws found during l
CRDM/CEDM inspections. The staff finds the proposed flaw acceptance criteria acceptable for axial cracks because the criteria conform to the American d
Society of Mechanical Engineers (ASME)Section XI criteria.
The staff 1
i determined that flaws that are primarily axial (less ti.an 45' from the axial
~
j direction) should be treated as axial cracks as indicated in Figure 1(b), (d),
j and (f) of your July 30, 1993 letter.
Flaws more than 45' from the axial j
direction should be treated as circumferential flaws. However, based upon j
information submitted to date and the more serious safety consequences of circumferential flaws, the staff does not agree with your proposed criteria l
for circumferential flaws.
Circumferential flaws which a licensee proposes to j
leave in service without repair, should be reviewed by the staff on a case-by-case basis.
i Sincerely, 1
MUD William T. Russell, Associate Director i
for Inspection & Technical Assessment j
Office of Nuclear Reactor Regulation
Enclosure:
j; As Stated Distribution:
J
' Central File JStrosnider WRussell EMCB RF BDLiaw RHermann JDavis JWiggins WKoo PDR j
- SEE PREVIOUS CONCURRENCE
- EMCB:DE
- EMCB:DE
- EMCB:DE
- EMCB:DE
- DE:D NRJt;ADJ rp sv6 JDavis WKoo RHermann JStrosnider JWiggins WRussell i
09/23/93 10/25/93 10/27/93 11/18/93 11/19/93 il / /i/93 0FFICIAL RECORD COPY G:\\ DAVIS \\WOGSER.JAD (s:\\ DAVIS) 1 i
SAFETY EVALUATION i
f.QB POTENTIAL REACTOR VESSEL HEAD ADAPTOR TUBE CRACKING i
i 1.0 INTRODWGTION i
Primary water stress co'resion cracking (PWSCC) of Alloy 600 was identified as an emerging issue by the NRC staff to the HRC Commission following a 1989 leakage from an Alloy 600 pressurizer heater sleeve penetration at Calvert Cliffs Unit 2, a Combustion Engineering designed pressurized water reactor (PWR).
Several instances of PWSCC of Alley 600 pressurizer instrument nozzles had been reported to the NRC between the time period of 1986 to the present on domestic and foreign i
pressurized water reactors (PWR). The licensee at Arkansas Nuclear Operations, Unit 1, a Babcock & Wilcox (B&W) designed PWR, reported a j
leaking pressurizer instrument nozzle in 1990, after 16 years of i
operation.
Westinghouse PWR's do not use Alloy 600 for penetrations or nozzles in the pressurizers.
According to the information provided to the staff by NUMARC at a public meeting held on July 5,1993, a leak was discovered in an Alloy 600 control rod drive mechanism (CRDM) adaptor tube penetration during a hydrostatic test at the Bugey 3 plant in France in 1991 after 12 years of operation. A visual examinatton of the CRDM adaptor tube penetration indicated the presence of axial flaws in the inside diameter (ID) of the CROM adaptor tube penetration.
The remaining 65 CROM adaptor tube penetrations were camined at Bugey 3 and 2 additional CRDM adaptor tube penetrations contained axial cracks on the ID of the CRDM adaptor tube penetrations. An examination of 24 CRDM adaptor tube penetrations at Bugey 4 revealed axial ID cracks in 8 CRDM adaptor tube penetrations.
CRDM adaptor tube penetrations have been examined at 37 nuc, lear power plants in France, Sweden,. Switzerland, Japan, and Belgium and 59 of the 1,850 penetrations have revealed short, axial crack indications.
The primary safety concern associated with stress corrosion cracking in i
Alloy 600 in CRDM penetrations is the potential for circumferential cracks.
Extensive circumferential cracking could lead to the ejection
)
of a CRDM resulting in an unisolable rupture in the primary coolant system. As indicated above, the inspections to date have identified short axial cracks. However, two other inspection findings are of Particular interest.
First, the CRDM penetration that leaked during hydrostatic testing at Bugey-3 was removed and examined metallurgically during December 1992. A secondary crack that was 0.120 inches long and O.090 inches deep at about 30 degrees to_the axial direction was observed on this CROM. Second, in early in 1993, a J-groove weld at the Ringhals plant in Sweden was discovered to contain a circumferential crack. Preliminary indications are that this flaw is a fabrication defect. Additional work is in progress by the staff at the Swedish Nuclear Power Inspectorate to confirm this.
The Westinghouse CRDM adaptor tube penetrations are similar in design to the European PWR's and use Alloy 600 for the penetrations. The NRC staff met with the WOG on January 7, 1992 to discuss the experience at
_. _ _ _ _ _. _. _ ~ _ _ -
j 2
k l
the Bugey 3 plant and the relationship of the French design of the CRDM l
adaptor tube penetrations to the design of domestic Westinghouse plants.
The WDG informed the NRC staff th't a program had been initiated in Decembar_1991 to: (1) determine tne root cause of the CRDM penetration cracking; (2) analyze the stress distributions in the CRDM penetrations i
of a typical domestic plant; (3) compare the design and operational characteristics of domestic and French plants to determine the j
likelihood for cracking; and (4) identify the need for additional l
efforts. The NRC staff also met with the Combustion Engineering owners i
Group (CEOG) and the Babcock & Wilcox Owners Group (B&WOG) to discuss the PWSCC of CRDM adaptor tube penetrations. The Nuclear Management and l
Resources Council (NUMARC) coordinated the PWR Owners' Group efforts on 4
this subject.
On June 16, 1993, NUMARC submitted safety assessments to the NRC from WOG, CEOG, and B&WOG for review by the NRC staff.
These safety assessments present stress analyses, crack growth analyses, leakage analyses, and wastage assessment for flaws initiating on the ID of CRDM adaptor tube penetrations. NRC requested additional information on the safety assessments by letter dated September 2, 1993.
NUMARC submitted the response to NRC on September 22, 1993. The safety assessments submitted to the NRC did not address the secondary flaw observed at the Bugey-3 plant that was oriented approximately 30" from the longitudinal i
axis of the penetration nor the apparent fabrication flaw at the i
Ringhals plant.
Neither of these flaws posed a threat to the integrity of the CROM penetrations. However, NtMARC has committed to submit a safety assessment relevant to this type of cracking. After this safety assessment has been reviewed by NRC, a supplement to this SER will be issued.
2.0 STAFF EVALUATION 2.1 WOG WCAP-13565. ALLOY 600 REACTOR VESSEL HEAD ADAPTOR TUBE CRACKING SAFETY EVALUATION The WOG submitted the, " Alloy 600 Reactor Vessel Head Adaptor Tube Safety Evaluation," through NUMARC on June 16, 1993.
The safety evaluation addresses the following elements:
1.
A summary of the. stress analysis focusing on the type (orientation) of cracking that may be expected in the Alloy 600 material, and the stresses necessary for flaw propagation; 2.
A summary of the flaw propagation analysis along with the background of the flaw prediction method, 1
3.
An assessment of the WOG plants with respect to penetration flaw indication data from plant inspections at Ringhals, Beznau, and various Electricite de France plants, in which the key parameters for cracking are compared to WOG plants; i
. _. _.. _ _ _ _ _ _ _ _.. _ _ _ _ _. _ _. _ _ _. _ _.. _.. _. ~ _. _. _ _ _ _.
k 3
I i
4.
A leakage assessment suemarizing leak rate vs. flaw size, and i
postulatin may apply;g leaks for WOG plants for which leakage considerations l
- and, i
5.
A vessel head wastage assessment including the process that leads to wastage and an estimate of the allowable wastage.
l 2.1.1 REGULATORY BASIS AND DETERMINATION OF AEVIEWED SAFETY OUESTIONS l
4 i
The WOG prepared safety evaluation addresses the potential for cracking i
and the ramifications of such cracking of the reactor vessel head adaptor tubes at Westinghouse designed NSSS plants. The WOG compared i
the results of this safety evaluation to the criteria in the Title 10, Code of Federal Regulations; 'Section 50.59 (10 CFR 50.59). The W0G l
i concluded that an unreviewed safety question did not exist.
Its evaluation considered the following:
j t
1.
Continued plant operation will not increase the probability of an j
accident previously evaluated in the FSAR.
2.
The consequences of an accident previously evaluated in the FSAR are i
j hot increased due to continued plant operation.
l 1-3.
Continued plant operation will not create the possibility of an
)
accident which is different than any already evaluated in the FSAR.
1
-4.
Continued plant operation will net increase the probability of a malfunction of equipment important to safety.
5.
Continued plant operation will not increase the consequences of a malfunction of equipment important to safety previously evaluated in the FSAR.
6.
Continued plant operation will not create the possibility of a malfunction of equionent important to safety different than any already evaluated in the FSAR.
7.
The evaluation for the effects of continued plant operation with potentially cracked reactor vessel head adapters has taken into account the applicable technical specifications.
2.1.2 STAFF'S EVALUATION OF THE REGULATORY BASIS ARLDEIf3tMINAIl0fLDE UNREVIEWED SAFETY OUESTIONS The staff agrees that no unreviewed safety question exists, provided only axial flaws are found.
Those axial flaws would be expected to be short, and they would most probably leak noticeably prior to the flaw size reaching unstable dimensions. The existence of any unexpected leaks would not odversely affect plant operation, or accident / transient response.
No significant equipment degradation would be expected.
I Details of the staff's evaluation that led to the above conclusions is discussed in the following sections.
.. ~... -
l 4
l 2.1.3 Pit G RATION STRESS ANALYSIS The W0G conducted an elastic-plas+ 4c. finite element analysis of a 4-loop IEE plant vessel head penetrations.
The WOG concluded that the 4-a loop WGG plant is bounding since prior analyses showed that the 3
operating and residual stresses are higher on a 4-loop plant than on 2 i
or 3-loop plants on the outernost penetrations. Three penetration i
locations were modeled, the center location, the outermost location, and the location next to the outermost location. The stress history was simulated by using a load sequence of the thermal load from the first welding pass, the thermal load from the second weld pass, the l
fabrication shop cold hydrotest, the field cold hydrotest, and the steady state operational loading.
I The highest stresses are found in the zone around the weld and are the highest in the penetration, farthest from the center of the vessel (peripheral penetrations).
The highest stresses on that penetration are on the side of the penetration nearest to the center of the vessel (centerside) and on the side of the penetration farthest from the center of the vessel (hillside). Also, tne stresses are the highest below the j
weld and decrease significantly above the weld. The ratio of peak hoop i
stress to axial stress at the same location at the outermost j
penetrations was about 1.4 compared to a value of about 1.6 estimated l
based on the degree of ovaling measured on actual penetrations. The i
ratio of hoop stress to axial stress was about the same for'conter j
penetrations as for peripheral penetrations (1.6 for center penetrations compared to 1.4 for peripheral penetrations); however, the magnitude of
}
the stresses at the peripheral penetrations was higher.
The analysis indicates that axial flaws would be more likely than circumferential j
flaws, flaws are more likely below the weld than above the weld, and i
that axial flaws would appear at locations in the penetrations where j
they have been found in service.
4 2.1.4 STAFF EVALUATION OF THE PENETRATION STRESS ANALYSIS l
j The staff is in agreement with the results of the WOG stress analysis that predicts that the cracking will be predominately axial. These results are in qualitative agreement with field inspection findings.
However, the WOG did not address the effects of possible straightening of the CROM penetration tubes during fabrication.
Such straightening operations could significantly alter the residual stress fields within the penetration tubes.
Results of inspections to date have not identified any problems directly related to this process; however, the staff requests that NUMARC address this issua for all three owners groups' plants.
2.1.5 CRACK GROWTH ANALYSIS: FLAW TOLERANCE The WOG crack growth analysis was based on the assumptions that the flaw would be caused by primary water stress corrosion cracking, and that the crack growth is controlled by the hoop stress.
The maximum principal stress will be oriented at a slight angle to the hoop stress and flaws j
4
i 5
l:
l would be expected to be perpendicular to the maximum principal stress.
j-However, all of the flaws found in service with two exceptions have been j
axially located.
Hence, the WOG used the hoop stress as an j
approxjpation of the maximum principal stress.
The outer-most penetration for a 4-loop Westinghouse plant was selected for analysis j
since this location experiences the highest stresses.
The highest stress was located along the inner surface just below the center side of j
the weld.
The calculated hoop stress through the wall of the penetration was used for flaw growth calculations. The flaw growth data were obtained from steam generator field experience and laboratory data.
i Based on the stress fields that exist in the CR::M penetrations, any flaw j
growth that occurs is expected to be predominately axial in nature.
i Furthermore, the growth of any flaws inclined from the vertical would be limited in length due to the nature of the existing stresses.
These conclusions are consistent with the inspection results described above.
Accordingly, there is no significant potential for failure of a penetration by ejection of the CROM sleeve. With~ regard to axial l
cracking, WOG has concluded that the critical flaw length for an axial 4
flaw for A11ov 600 is sufficiently long that leakage would occur and be i
detected during surveillance walkdowns as required by GL 88-05.
1 Therefore, the consequences of cracking in the penetration sleeve are j
limited to the Effects of leakage as discussed below.
l The flaw growth analysis showed that under the most severe conditions of metallurgical micrcstructure, peak hoop stress, and operating l'
temperature, it would take about five years for a flaw to grow through j
wall.
Under the same conditions, it would take an additional 10 years i
for a through-wall flaw to grow I % inches above the weld on the lower j
hillside of the outemost head penetrations (Figure 3.2-2) and about the i
same time to grow two inches above the J-groove weld on the center side of the outermost penetrations (Figure 3.2-3).
The flaw growth analysis j
indicates that through wall flaws would essentially arrest before
{
growing a maximum of two inches above the weld.
These flaws would be 1
constrained within the head and could not significantly open thus j
limiting the amount of Teakage that could occur.
4 i
2.1.6 STAFF EVALUATION OF THE CRACK GROWTH ANALYSIS _
l j
The WOG stated that the crack growth analysis is in general agreement l
with the inspection findings. The crack growth rate data used in this l
analysis was limited, but the results predicted using these flaw growth 3
data bound the results of the inspections.
Crack growth rates are difficult to detemine precisely; however, the assumed growth rates compare well with inspection data available to date and the large margins that exist in the analyses will account for any possibly higher growth rates. There are large margins of safety in the analyses and the CRDM penetrations are constructed of inherently tough material with a critical flaw size of approximately 13 inches in the free span above the reactor vessel shell.
Therefore, the staff concludes that catastrophic failure of a penetration is extremely unlikely because a flaw would be
i l
]
6 detected during boric acid leakage surveillance walkdowns before it reached the critical flu size.
i i
2.
1.7 ASSESSMENT
OF WOG PLANTS 3
i The WOG compared the Ringhals and Beznau plants to the domestic i
Westinghouse plants and developed a model for the relative susceptibility to PWSCC. The WOG considered residual and operating stresses in the penetrations, the environment, material condition, i
operating temperature, and time-of-operation at temperature, and i
pressure.
Based on this evaluation, the WOG has evaluated domestic WOG i
PWR's with regard to their degree of susceptibility.
Based on what WOG l.
considers to be conservative assumptions, the Ringhals plants envelope l'
45 domestic plants.
None of these plants are expected to have any flaws other than some short, shallow, axial flaws. Nine additional WOG plants are not enveloped by the Ringhals plants.
Based on the stresses, operating temperatures, hours of operation, and the flaw growth curves provided in the WOG safety assessment, the WOG does not expect any CRDM penetration axial flaws to reach a length in excess of 1 inch before about the middle of 1995.
2.1.8 IIBFF EVALUATION OF THE WOG ASSESSMENT The susceptibility model developed by the WDG considers the appropriate parameters affecting IGSCC and should provide a reasonable ranking of plant susceptibilities.
In addition, this evaluation indicates that it is unlikely that U.S. plants should exhibit any cracking significantly worse than that found in European plants.
I 2.1.'9 LEAK RATE CALCULATIONS The leak rates were calculated based on the assumption that the leak rate will be controlled by the flow rate through the flaw in the head i
penetration or by the flow through the penetration annulus, whichever is j
smaller. WOG estimates the maximum leak rate would be 0.7 gpa for a 2 inch long flaw and an annular clearance of 0.003 inches.
Leakage above 1.0 gpa is detectable in domestic WOG plants according to WOG.
Growth of an axial flaw outside of the part contained within the reactor head will result in leakage greater than 1.0 gpm prior to reaching the j
critical flaw size.
The WOG stated that an axial flaw would remain stable for growth up to 13 inches above the reactor vessel head.
2.1.10 STAFFS EVALUATION OF THE WOG LEAK RATE CALCULATIONS The staff agrees with the WOG assumptions about leakage and concludes, that based on existing leakage monitoring requirements, there is reasonable assurance that leakage 'in excess of the 1.0 gpm technical specification limit would be detected prior to any unstable extension of the flaw.
l b
7 2.1.11 REACTOR VES$fL HEAD WASTAGE ASSESSMENTS This section assesses the potenttd wastage of the reactor vessel head due to deakage of primary coolant through the CRDM penetrations.
This i
assessment is based on wastage data from previous Westinghouse i
experiments and from the results of a penetration mockup test conducted by the Combustion Engineering Owners Group (CEOG).
This analysis assumed that coolant escaping from the penetration would i
flash to steam leaving boric acid crystals behind. WOG assumed that l
crystals would accumulate on the vessel head but would cause minimal corrosion while the reactor was operating.
The head temperature would be about 500"F during operation and significant wastage of the reactor i
head by the boric acid crystals would not be expected.
Dry boric acid crystals do not cause corrosion. Wastage would only occur during j
outages when the head temperature is below 212*F.
4 j
The CEOG provided all of the PW owners groups with the results of pressurizer penetration mockup test results.
The WOG oxamination of the j
CEOG mockup test results showed that the maximum penetration rate at the i
deepest pit was 2.15 inches / year while the average penetration rate was 0.0835inchys/ year. The maximum total metal loss rate or wastage volume was 1.07 in / year, and the greatest damage occurred where the leakage lefttheannulus. The WoG considered the maximum wastage would be 6.4 in of vessel head material.
The assumptions made were that any leakage i
over 1.0 gpa can be detected so only leak rates between 0.0 and 1.0 gpm were considered.
The WOG analyzed the situation using finite element i
analyses for a 2 loop, 3 loop, and 4 loop reactor vessel head where a l
1.0 gpm leak went undetected for 6 years and concluded that the ASME i
code minimum wall thickness requirement would be satir.fied and that the stresses remain within the ASME code allowable stresses.
l 2.1.12 THE STAFF'S EVALUATION OF THE REACTOR VESSEL HEAD WASTAGE i
ASSESSMENTS The assumption used in the'WDG corrosion assessment are based on experimental d..+.a and should provide a reasonable estimate of potential wastage of the 1: actor vessel head.
Based on these evaluations, there would be significant time between initiating a leak and experiencing wastage that would reduce the stractural integrity margins of the reactor vessel head to below accaptable levels. Considering the length of time involved, there 1: reasonable assurance that leakage, manifested by the accumulation of moderate amounts of boric acid crystals would be detected during a surveillance walkdown in accordance with GL 88-05.
3.0 CE0G SAFETY EVALUATION The CEOG safety evaluation is essentially the same as the WOG safety evaluation. The CEOG plants run at a slightly higher temperature than the European plants that have experienced cracking, have greater hillside angles, and have been in operation longer than many of the European plants.
The CE0G indicated that all of these factors would
~ - - -
e, r
i I
8 i
increase the probability of cracking for the CE0G plants.
However, the CEOG plants have significantly less weld metal in the J-groove welds and i
i the CE0G stated that this would significantly reduce the residual i
welding-induced stresses and would reduce the probability of PWSCC.
i CE0G concluded that any PWSCC that formed would be short, axial flaws.
i The CEOG states that they can detect a 0.12 gpa leak in the primary coolani; system.
CE0G also states that the boric acid accumulation as a result of a 0.12 gpa leak would not result in wall thinning below the code allowables in less than 8.8 years compared to 6 years for WOG plants and that surveillance walkdowns would detect boric acid crystals j
long before the 8.8 years.
3.1 STAFF EVALUATION OF THE CEOG SAFETY EVALUATION l
The staff has concluded that the potential for PWSCC of CRDM/CEDM for CEOG plants does not create an immediate safety issue as long as the surveillance walkdowns required by GL 88-05 continue and corrective action is instituted when leaks are discovered. 'The CEOG analyses indicating that the stresses would favor development of axial rather than circumferential cracks and that significant time would be required to reduce the wall thickness of the vessel head to below the ASME code allowables demonstrates that an immediate safety concern does not exist.
4.0 B&WOG SAFETY EVALUATION The B&WOG safety evaluation was essentially the same as the WOG and CEOG safety evaluations. The B&WOG analysis indicates that B&WOG plants have essentially the same susceptibility to PWSCC as the European plants based on operating temperature, residual stresses, and operational life.
The B&WOG predicts short, axial flaws on the peripheral locations based on the results of' finite element analyses.
The B&EOG estimates that it would take 10 years from the time a flaw initiates on the inside diameter of a CROM penetration until a leak appears.
Once a leak starts, B&WOG concluded that it would take 6 years before enough corrosion would occur to reduce the wall thickness of the reactor vessel head to below ASME code minimums, and that this amount of leakage would be detected during surveillance walkdowns.
4.1 STAFF EVALUATION OF THE B&WOG SAFETY EVALUATION The staff has concluded that the potential for PWSCC of CROM for B&WOG plants does not create an famediate safety issue as long as the surveillance walkdowns required continue and as long as any leakage is corrected.
The B&WOG analyses, indicating that the stresses would favor development of axial rather than circumferential cracks and that significant time would be required to reduce the wall thickness of the vessel head to below the ASME code allowables, demonstrates that an immediate safety concern does not exist.
9 5.0 PROPOSED FLAW ACCEPTANCE CRITERIA On July 30, 1993, NUMARC submitted the proposed flaw acceptance criteria for f1ms identified durirg inservice inspection of reactor vessel upper head penetrations to the NRC for review.
These criteria were developed by utility technical staffs and tha domestic PWR vendors.
NUMARC proposes that axial flaws are perisitted through-wall below the J-groove weld and 75 percent through-wall above the we' d.
There is no limit on the length of the flaws. NUMARC proposes that circumferential flaws through-wall and 75 percent around the penetration be allowed below the J-groove weld and that circumferential flaws above the weld could be 75 percent through-wall and 50 percent around the penetration.
Proximity rules found in ASME Section XI, Figure IWA 3400-1 are proposed for determining the effective length of multiple flaws in one location.
t NUMARC proposos that the flaws be characterized by length and preferably depth. NUMARC proposes that if only the length is characterized, that-the depth be assumed to be one half of the length based on inspection findings to date.
5.1 STAFF EVALUATION OF THE PROPOSED FLAW. ACCEPTANCE CRITERIA The staff finds the proposed flaw acceptance criteria acceptable for axial flaws because the criteria conform to the American Society of Mechanical Engineers (ASME)Section XI criteria.
The assumption that flaw depth is one half the flaw length for flaws whose depth cannot be determined will limit the flaw length to 1.5 times the thickness of the penetration sleeve. However, it is expected that reasonable attempts will be made to deter 1eine flaw depths.
Flaws found through inservice inspection (ISI that are primarily axial (les.s than 45' from the axial direction) will)be treatedias axial flaws as indicated in Figure I (d), and (f) of NUMARC'S July 30, 1993 letter.
Flaws more than 45 from the axial direction are considered to be circumferential flaws.
Based upon information submitted to date and the more serious safety consequences of circumferential flaws, the staff has concluded that criteria for circumferential flaws should not be pre-approved.
Detection of such flaws would be contrary to inspection results to date and to the conclusion of the Owners Groups evaluations. The curcumstances associated with such a flaw would have to be well understood.
Therefore, any circumferential flaws found through ISI, which a licensee proposes to leave in service without repair, will be reviewed on a case-by-case basis by the staff.
6.0 LEAKAGE MONITORING NUMARC, through the owners groups' reports, determined that any leakage in excess of I gpa would be detected prior to any unstable extension of axial flaws. Also, leakage at less than 1 gpa would be detectable over time based on boric acid buildup as noted during periodic surveillance walkdowns. Although NUMARC has proposed, and the staff agrees, that low level leakage will not cause a significant safety issue to result, the staff determined that NUMARC should consider methods for detecting smaller leaks to provide defense-in-depth to account for any potential S
10 uncertainty in its analyses. The reported leak rate at Bugey 3 was about 0.003 gpa and was detected using acoustic monitoring techniques during the perOsmance of a hydrostatic test. The staff does not think that tis necessary to detect a 0.003 gpm leak, but does think that permitting leakage just below 1.0 gpa as currently proposed may be undesirable.
Leakage of this magnitude would produce significant deposits (thousands of pounds / year) of boric acid on the reactor vessel head. Further, most facilities
- technical specifications state that no pressure boundary leakage is permitted.
The staff notes that small leaks resulting from flaws which progressed through-wall just prior to a i
refueling outage would be difficult to detect while the thermal insulation is installed. Although running for an additional cycle with 4
that undetected leak would not result in a significant safety issue, the NUMARC should consider proposing a method for detecting leaks that are significantly less than 1.0 gps, such as the installation of on-line monitoring equipment.
70 CONCLUSIONS Based on reviw of the NUMARC submittal and the overseas inspection results, the staff concludes that the CROM/CEDM cracking at the reactor vessel heads is not a significant safety issue at this time as long as the surveillance walkdowns in accordance with GL 88-05 continue.
'he staff agrees with the NUMARC's determination that there are no 4
unreviewed safety questions associated with stress corrosion cracking of CROM penetrations.
However, new information and events may require a i
reassessment of the safety significanew.
Furtherimore, there is a need to verify the conclusions of the NUMARC's safety evaluations.
1 Therefore, nondestructive examinations should be performed to ensure there is no unexpected cracking in domestic PWRs. These examinations do not have to be conducted immedtately since only short, shallow, axial flaws are likely to be present in the CROM penetrations. The industry has comeitted to conduct inspections at three units in 1994. They are:
(a) Point Beach I! nit 1 in the Spring of 1994 (b) D.C. Cook Unit 2 in the third quarter of 1994, (c) Oconee Unit 2 in September 1994.
As the surveillance walkdowns proposed by NUMARC are not intended for detecting small leaks, it is conceivable that some affected PWRs could potentially operate with small undetected leakage at CRDM/CEDM penetrations.
In this regard, the staff believes it is prudent for NUMARC to consider the implementation of an enhanced leakage detection method for detecting small leaks during plant operation.
The staff found NUMARC's flaw acceptance criteria acceptable for axial flaws but NRC review and approval of the disposition of any circumferential flaws will be required.
Technical Contacts:
Robert A. Hermann (301) 504-2768 William H. Koo (301) 504-2706 James A. Davis (301) 504-2713
=.
.--na
,.-,L,...
,4.=
4--
1..w.,,
im.s a
f i
I l
?
i l
i t
i I
i 1
\\
l I
l
[
ATTACHMENT D 1
i l
i I
m l
1 l
i I
i
DOMINION ENGINEEPJNG,INC.
DEI Letter L-5061-01-03 Time from 1/1/97 until 75% through-wall Cracking Above the Bottom of the J-groove Weld for Zion Units 1 & 2 Probability
.g Zion 1 Zion 2 Level Worst Nozzle (1)
Any Nozzle (2)
Worst Nozzle (1)
Any Nozzle (2) m 21.6 3.4 79.1 22.6 5%
2
{
l89,000 30,000 693,000 198,000 m
36.5 8.0 123.2 37.1 s
16 %
2
{
320,000 70,000 1,080,000 325,000 m
65.5 17.5 211.2 65.4 50 %
2 i
[
574,000 153,000 1,851,000 573,000 m
(1) Time for highest susceptibility CRDM nozzle in each head (2) Integrated time for any of the 78 CRDM nozzles in each head l
4
DOMINION ENGINEERING,INC.
DEI Letter I 5061-01-03 l
l Time from 1/1/97 until 75% through-wai'. Cracking Above the Bottom of the J-groove Weld for Braidwood Units 1 & 2
. Probability
,g Braidwood 1 Braidwood 2 Level Worst Nozzle (1)
Any Nozzle (2)
Worst Nozzle (1)
Any Nozzle (2) 71.2 28.7 57.0 22.7 i
e 5%
2{
624,000 252,000 500,000 199,000 104.7 39.7 82.9 31.8 b
16 %
2 l
{
918,000 348,000 727,000 279,000 m
l 171.2 61.4 136.2 49.5 50 %
1,501,000 538,000
'1,194,000 434,000 m
(1) Time for highest susceptibility CRDM nozzle in each head (2) Integrated time for any of the 78 CRDM nozzles in each head l
l l
l
DOMINION ENGINEERING,INC.
DEI Letter L-5061-01-03 Attachment I i
Time from 1/1/97 until 75% through-wall Cracking Above the Bottom of the J-groove Weld for Byron Units 1 & 2 Probability lp Byron 1 Byron 2 Level Worst Nozzle (1)
Any Nozzle (2)
Worst Nozzle (1)
Any Nozzle (2) 114.5 42.8 166.8 61.5 b
1 5%
l 3{
1,004,000 375,000 1,462,000 539,000 l
w 170.2 60.9 247.0 87.7 16 %
3 l
{
1,492,000 534,000 2,165,000 769,000 i
281.3 97.1 409.3 139A 50 %
2 l
{
2,466,000 851,000 3,588,000 1,222,000 m
(1) Time for highest susceptibility CRDM nozzle in each head (2) Integrated time for any of the 78 CRDM nozzles in each head l
b I
d e
i I
l l
f a
1 1
l l
l l
l l
l 1
1 l
l ATTACHAENT E t
1 I
l l
l t
l l
1 k
j 3
l I
I i
l l
I l
l
1 DOMINION ENGINEERING, INC.
July 22,1997 L-5057-00-1 Mr. Thomas Spry Commonwealth Edison 1400 Opus Place, Suite 400 i
Downers Grove,IL 60515 l
Subject:
CRDM Nozzle Stratede Planning for Conunonwealth Edison Plants
Dear Mr. Spry:
Enclosed are summary appendices providing the inspection and test data used to develop crack initiation and growth models, and the plant specific inputs to the models for the six Commonwealth Edison plants. The crack initiation and growth prediction methodology and the validation of the stress analysis method is reponed in EPRI TR-103696, "PWSCC of Alloy 600 Materials in PWR Primary System Penetrations."
l r
If you have any questions, do not hesitate to contact me at (703) 790-5544.
Sincerely, E. Stephen Hunt Enclosure cc:
Dr. Raj Pathania, EPRI l
i i
i 6862 ELM STREET McLEAN. VIRGINIA 22101 PHONE: 703n90-5544 FAX: 703n90 0027 i
~ _. -..
DOMINION ENGINEERING,INC.
1 Appendix A-Selection of Crack Initiation Reference Data l
The October 1994.Oconee Unit 2 CRDM nozzle inspection data are considered the best reference point for the Comed vessel analyses for several reasons.
i The Oconee nozzles were all inspected and the inspection showed that one nozzle had shallow indications which can be conservatively assumed to be crack initiation.
This provides a basis for establishing a reference time for cracking.
i L
The CRDM nozzle materials for Oconee Unit 2 and all six Comed units were L
supplied by B&W Tubular Products. Accordingly, it is not necessary to include materials correction factors.
The Oconee 2 and all six Comed vessel heads were fabricated by B&W. The two Zion vessels were fabricated about the same point in time as the Oconee vessels j.
using materials with similar heat numbers.
1 By selecting a reference plant fabricated by the same company using materials from the same j
supplier, uncertainties associated with differences in materials and fabrication methods are
)
reduced.,
i While the Oconee 2 inspection data serve as the main reference for the Comed predictions, the distribution for statistical purposes considers other industry inspection results.
Table A-1 shows the reference conditions assumed for crack initiation analysis of the six Comed PWR plant CRDM nozzles.
l i-1 I
i 4
5 1
1 e
m-+
--m.,
w.~
DOMINION ENGINEERING,INC.
l c
l l
Table A-1 Kev Input Parameters for Comed CRDM Nozzle Analyses i
Distributed input Data CIRSE Inputs Units Lower Nominal Upper Distribution Bound Bound i
Nozzle Temperatures
- Oconee 2 (ref)
F N/A 602.0 N/A N/A
- Zion 1 F
585.8 590.8 595.8 triangular
- Zion 2 F
585.8 590.8 595.8 triangular l
- Byron 1 F
546.0 551.0 556.0 triangular
- Byron 2 F
545.4 550.4 555.4 triangular
- Braidwood 1 F
551.0 556.0 561.0 triangular
- Braidwood 2 F
547.0 552.0 557.0 triangular Nozzle Stress
- Oconee 2 (ref) ksi N/A 56.7 N/A N/A
- Com Ed Plants ksi
-10%
FEA 10 %
triangular Other Factors
- Material 1.0 1.0 1.0 triangular
- Fabrication 1.0 1.0 1.0 triangular
- Chemist 1y 1.0 1.0 1.0 triangular i
Crack Initiation
-Time to 10% (ref)
EFPY 20.0 32.0 50.0 triangular l
- Weibull Slope 2.0 3.0 4.0 triangular
- Activation Energy kcal/ mole 45.0 50.0 55.0 triangular 3.0 4.0 5.0 triangular
- Stress Exponent Crack Growth
- Ref. Temperature F
N/A 617.0 N/A N/A
- Constant (A) m, s, MPa
.366E-12
.164E-11
.310E-11 log-triangular 1.16 1.16 1.16 triangular j
- Exponent (n)
-Threshold K (Kth)
MPaVm 4.0 4.0 4.0 tri mgular
- Activation Energy kcal/ mole 30.0 33.0 35.0 triangelar
)
-Initiation Depth mm 0.1 0.1 0.1 triangular 1
1 1
l l
t
~ _ _ -. - _ _ _ _ _ _ _ _ _.
DOMINION ENGINEERING,INC.
Appendix B i
i Selection of Crack Growth Reference Data l
i Crack growth predictions for the six Comed PWR vessel heads are based on the linear-clastic l
fracture mechanics model proposed by Scott.
i A (K - K )"
A'
=
th t
I
- where, a'
crack growth rate (m/s)
=
power law constant (m, s, MPa)
A
=
stress intensity (MPa6)
K
=
threshold stress intensity (MPa6)
Kth =
powerlaw exponent l.
n
=
l l
This model has become the industry standard for CRDM nozzle PWSCC. Recent industry field inspections 2 and laboratory testing.4 have substantiated this basic model. While 3
laboratory testing and industry experience substantiates the Scott model, these data show that there is a significant amount of scatter. This scatter has been handled in the Monte Carlo analysis by assuming a range of distributed parameters as indicated in Table A-1. These data bracket fifty-one field and laboratory test data points.
l 1
l 1
Scott, P. M. " Prediction of Alloy 600 Component Failures in PWR Systems," presented at Corrosion %,
Denver, CO. Published by NACE International, Houston.
2 Hong, S. L. " Crack Growth Rate Measurements of Vessel Head Penetration Alloy 600 in PWR Environment," presented at the 4th EPRI Workshop on PWSCC of Alloy 600 in PWRs, Daytona Beach, i
i FL, February 1997.
3 Bamford, W. " Crack Growth Behavior of Alloy 600 Head Penetration Materials " presented at the 4thEPRI Workshop on PWSCC of Alloy 600 in PWRs, Daytona Beach, FL, February 1997.
4 Cassagne, T. "PWSCC Crack Propagation Rate in Alloy 600 in Primary Water," presented at the 4th EPRI Workshop on PWSCC of Alloy 600 in PWRs, Daytona Beach, FL, February 1997.
I i
. ~.
l I
i l
[
i l
ATTACHMENT F r.
I l
1
\\
l l
l l
)
1 l
1 t
I t
Se W Criteda For Evaluation of Reactor Coolant Chemmistry Relative to RCS Resin l
Ingress During Power Operation Category 1:
Pre Sulfate Determination Data Large Ingress (>l cu. ft., > 30 L, Cation Resin)
Specific Conductance @ 25 C >28 micro S/cm Elevation /ft resin 3
pH @ 25 C >1.0 pH Unit Depression i
Li Increase to greater than vad operating band SoWad Solids >1.0 ppm Corroborating Evidence:
A. Increased delta P in RC and SWI filters l
B. Sampleline plugging C. Resinin sample sink Category 2:
Pre Sulfata Dh. :udon Data Small Ingress (<0.lft', <3L, Cation Resin)
Specific CoWaw @ 25 C >3 times the expected deviation from the mean value.
l Conoborating Evidence:
A. Trend in pH reduction B. Trend in Li elevation Category 3:
Small Ingress (<0.lft', <3L, Cation Resin) with sulfate data.
Sulfate >50 ppb Sulfate up to 1.7 ppm peak concentration for a lesk of 0.lft3 l
Corroborating Evidence:
i A. Trendin pH reduction l
B. Trend in Lireduction
JLA. 83 '9716:05 FRAC 412 374 6337 TO 01AM f71'71 P.0907 4
Rationale For Selection of Semening Criteria Parameters Relative To Resia Ingress To The RCS
- 1. Of all PWRs known to have experienced resin ingress to the RCS, the only plant which had accumulated sufHeient ch-mi*y data associated with a resin in leakage incident to be usefbl in establishing a screening criteria was Zorita in 1980 and 1981. The 1981 resin leak was so large that some ofthe a= *47 p -.-_c-2 were driven beyond the limits ofdetection.
Therefore, the 1980 incidant provided the better data for use in developing a parametric basis for establishing an evaluation criteria with PWR historical data.
- 2. The 1980 Zorita leak was estimated to be 40 L of cation resin. Within the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, the y
specific nanthwance peaked at 165 micro S/cm befom it started to decline. 'Ihat peak preceded hydrazine additions for pH coutml which would have complie=*ad interpretation of conductivity data. The pH decreased to about 4.0 and lithium inemased to nearly 5.0 ppm.
'Ihe waaadad solids (crud plus possibly some resin fines) peaked at greater than 8 ppm.
- 3. The volume of the Zorita RCS was given as 76.4 m' or 7.64 x 104 L. By comparison, the averase KC5 "olume of a large plant PWR is 3.1 x105 L or 4.1 times larger than Zorita.
The ?Ae, t!r. conductivity imparted to reactor coolant from the ionic degradation products of c# w mn released from a given quantity of resin would be much less than that expencoced at Zorita 1-m=== of the greater coolant volume in large PWRs. h peak specific candiwanne r.;s,ed at Zorita,165 micro S/cm for 40 L resin, leak nuy be v
normalimi to 1.0 ff ofresin, or approximataly a 122 micro S/cm elevation /1.0 ff. Relative to the volane of a large PWR, the specific candne+aar= would be approximately a 28 micro S/cm elevation / cubic foot of cation resin.
- 4. The Zorita resin leak of approximately 40 L, or 1.3 ff, caused a pH dor aion of betwas I and 2 pH units. Therefore, a leak greater than 1.0 ff would be arnaetad to dor <
the pH greater than 1 pH unit. h increase in coolant lithium concentration would be a function of lithium released from degraded cation resin. The lithium inventory available for release would depend on the extent oflithiation of the cation bed resin leaked or the quantity of the fully lithiated cation portion ofmixed bed resin which was leaked. A qualitativej%=maat would be that cation rusin (>l 19) degradation from either a cation bed or a mixed bed could add sufEcient lithium to the coolant to drive the lithium concentration above the nonnal control limit of 2 ppm. This would be more likely with fully lithiated cation resin from a mixed bed.
- 5. The Zorita coolant suspended solids increased to values well into the ppm range from nonnal concentrations in the low ppb range. Therefore, qualitatively, a cation resin leak of greater than I ff, causing a period of acid chemistry operation, would be expected to cause a par +i& -+- crud burst above I ppm.
8
- 6. In the absence of sulfate data, a smeil ingress of cation resin could be identified by comparing
JLL. 83 '97 16:05 FR 1.EC 412 M4 @MFP T@ @i@K12!$fPaTB
@.5PMEFF I
e plot of meastad specific conductance to a plot of the theoretical conductivity. If a deviation greater than 3 times the v~I value is observed, indicating the possibility of resin ingress, co ro!~ratmg evidence such as trends in pH reduction or an elevation in lithium concentration should be confumed.
- 7. In order to detect a small ingress of cation resin (<0.1 ft' or <3L) when sulfate data are available, the initial criteria for evaluation which should be used was established in the EPRI PWR Fameiy Water N-- -;g Guidelines, Rev. 3, at a sulfate concentration of 50 ppb. If a cation resin leak of up to 0.1 ft' has occurred, a sulfate concentration up to a peak of 1.7 ppm could be observed. The 1.7 ppm maximum concentration was derived from a sulfur content in cation resin of 0.046g S/g resin, convened to the mass of sidrate in 3L ofresin and related to an RCS volume of 3.1 x 105 L. An observation of the maximum concentration of sulfate would not be expected, however, because the production of the degradation pmduct would not be instantaneous and the purification system would tend to reduce the concentration as it was being releasedinto the coolant.
- 8. In both categories 2 and 3, there could be evidence of a trend in pH reduction and lithium elevation, Wii on the magnitude of the leak. If detennined, each would provide corroborating evide ofa low level leak.
..... - ~ ~. -
A 4
1 i
l I
l i
1 I
ATTACHMENT G 1
i i
I i
4 1
l i
l I
i
Chemistry and Conductivity Excursion Summary Braidwood Units 1 and 2, Byron Units I and 2, and Zion Units 1 and 2.
Bnidwood Units 1 and 2 The r. iew period for Braidwood ' Units 1 and 2 was from 1987 to the present. There were 21 instances where the s. 2ning criteria were exceedal. The date, description, technical assessment, and disposition ofeach excursion are as follows:
Braidwood Unit 1 2/4/91 - 4/27/91 - Unit One, Modes 5-6. During AIRO2, an oil intrusion into RCS from the SI Accumulators occurred. Old N2 compressor diaphragms leaked and oil was pumped into each accumulator during N2 pressure adjustments. Accumulator dumps into the RCS cavity were performed during the outage. Elevated sulfate concentrztions as high as 165 ppb were reported during this time frame and was attributed to the oil intrusion. No other elevated parameters associated wi.th resin intrusion were noted. No resin ingress occurred.
9/6/92 - 9/8/92 - Unit 1, Mode 5. During shutdown for refuel outage AIRO3, the IB CVCS mixed bed demineralizer released sulfates and chloride due to exhaustion / chemical exchange. RCS chlorides exceeded 150 ppb for approximately 47 hours5.439815e-4 days <br />0.0131 hours <br />7.771164e-5 weeks <br />1.78835e-5 months <br />. RCS sulfates were reported as high as 176 ppb and chlorides were reported as high as 392 ppb. Comed and Westinghouse evaluations were performed and determined that no degradation of materials occurred. RCS temperatures ranged from 198*F to 140 F, which is below the temperature required for cation resin decomposition. No other eles2ted parameters associated with resin intrusion were noted. No resin ingress occurred.
3/23/94 - 3/30/94 - Unit 1, Mode 6. During AIRO4 fuel moves, elevated sulfate concentrations were noted. These sulfate concentrations ranged from 70 ppb to 98 ppb and were attributed to resolubilization of sulfate within the RCS following RCS cooldown. No other elevated parameters associated with resin intrusion were noted. RCS temperature was below 120 F. No resin ingress occurred.
3/31/94 - 4/3/94 - Unit 1, Mode 6. During AIRO4, a reverse osmosis unit was operated on the reactor cavity for cleanup of silica. Localized elevated sulfate concentrations as high as 1.87 ppm and chloride concentrations as high as 6.38 ppm were reported in the reactor cavity during this time frame. However, RHR samples indicated a concentration of 20 to 35 ppb sulfate, and the steady state chloride limit of 150 l
ppb was not exceeded in the RCS. The elesated sulfate and chloride coincided with the operation of the l
reverse osmosis (RO) unit. Although the source of the contamination was not confirmed, a layup l
chemical used to protect the RO membranes was highly suspect. Continued operation of the RO unit and l
normal use of demineralization decreased the sulfate and chloride concentrations to acceptable values on j
4/4/94. Comed and Westinghouse evaluations were performed and determined that no degradation of materials occurred. No resin ingress occurred.
4/18/94 - 4/19/94 - Unit 1, Mode 6. During AIRO4, RCS loop fills, an increase in sulfates was reported to as high as 267 ppb. The increase in sulfates is attributed to resolubilization of sulfates in the RCS following RCS cooldown. No other elevated parameters associated with resin intrusion were noted. RCS temperature was below 120 F. Therefore, no resin ingress occurred. l
10/9/95 - 10/12/95 - Unit 1, Mode 6. During AIRO5 shutdown, an increase in sulfates was reported in i
the 97 to 132 ppb range. The increase in sulfates is attributed to resolubilization of sulfates in the RCS following RCS cooldown. No other elevated parameters associated with resin intrusion were noted. RCS temperature was below 120'F. Therefore no resin ingress occurred 10/19/95 - 10/20/95 - Unit 1, Mode 6. During AIROS, hydrogen peroxide was added to the RCS cavity to improve clarity of the water. Coincident with the hydro;;en peroxide addition and reactor cavity inspections (by submarine), elevated sulfates ranged from 112 to 125 ppb. The elevated sulfates are attributed to resolubilization of sulfates in the RCS following RCS cooldown and the reactor cavity inspections. No other elevated parameters associated with resin intrusion were noted. RCS temperature was below 120 F. No resin ingress occurred.
10/26/95 - Unit 1, Mode 6. During AIROS core reload, elevated sulfates as high as 132 ppb were i
observed. The elevated sulfates are attributed to sulfate resolubilization within the RCS following RCS cooldown. No other elevated parameters associated with resin intrusion were noted. RCS temperature was below 120 F. No usin ingress occurred.
10/14/96 - Unit 1, Mode 5. During AlPO2, an addition of hydrogen peroxide was performed for the purpose of RCS oxidation. Elevated sulfates ranging from 75 to 157 ppb were noted. The elevated sulfates are attributed to ruotubilization of sulfates within the RCS following RCS cooldown. No other elevated parameters associated with resin intrusion were noted. RCS temperature was below 140'F. No resin ingress occurred.
Braidwood Unit 2 2/24/88 - Unit 2 Mode 3. At approximately 2% reactor power RCS specific conductivity increased from 15.4 mho on 2/23 to $1.7 mho on 2/24 and back down to 14.5 mho on 2/25. The RCS pH followed a slow reduction in lithium concentration over the three days. There was no significant depression in pH on 2/23 nor was there any appreciable change in the other parameters from the previous days. The calculated conductivity and pH based on Boron and lithium values across the three day time frame indicate a steady well controlled chemistry. This entry is considered to be in error and actually appears to i
have the first two numbers of the conductivity valuejuxtaposed. 51.7 mho should likely have been l
recorded as 15.7 mho. The calculated conductivity for 2/23 was 15.2 mho. No resin ingress occurred.
l 4/18/93-4/21/93 - Unit 2, Mode 5. During this time period in A2R03, RCS sulfe ranged from 95 ppb to l
455 ppb. Unit 2 RCS loops were filled on 4/13, the RCS went solid on 4/16, and RCP bumps were performed on 4/18. RCS sulfate staned trending up on 4/17 and, eventually, the online CV mixed bed was sluiced out and refilled with anion resin, which cleaned up the RCS sulfate from 4/21 to 4/22. Both the RCS pH and conductivity remained stable during this time frame and the temperature of the RCS was approximately 120*F, which is too low for the thermal decomposition of cation resin. The cause of the high sulfate values is believed to be sulfate resolubilization or chemical contamination caused by startup RCS activities. No resin ingress occurred.
l 4/18/94-4/22/94 - Unit 2 Mode 5-6. During this time in A2F27 RCS sulfates ranged from 63 ppb to 126 ppb. A hydrogen peroxide addition to initiate a crud burst was performed on 4/16, and Unit 2 entered mode 6 on 4/22. RCS fluoride and chloride values remained less than 4 ppb during this time period, and the temperature of the RCS was less than 140 F, which is too low for the thermal decomposition of cation resin. The cause of the high sulfate values is believed to be sulfate resolubilization or chemical contamination caused by startup RCS activities. No resin ingress occurred.
4/28/94 - 5/13/94 - Unit 2, Mode 6. During this time period in forced outage A2F27, RCS sulfates ranged from 51 to 396 ppb. The increase was noted following the accumulators dump to the RCS cavity on 4/26
^
r l
l and the removal of the upper internals on 4/27. During this time frame, RCS conductivity was less than 2 mhos and pH remained stable. In addition, the temperature of tb RCS was less than 140*F, which is too low for the thermal decomposition of cation resin. The cause of the high sulfate values is believed to be chemical contamination during the outage. No other elevated parameters associated with resin intrusion were noted. No resin ingress occurred.
5/16/94 - 5/18/94 - Unit 2, Mode 5, During this time period in A2F27, sulfates ranged from 50 to 318 l
ppb. The RCS went solid and RCPs were bumped or 5/17. The RCS started heating up from 108 F to 180 F on 5/18. The pH and conductivity values dunng this time frame were stable and the temperature of the RCS was less than 140 F, which is too low for the thermal decomposition of cation resin. The cause i
of the high sulfate values is believed to be sulfate resolubilization or chemical contamination caused by stanup RCS activities. No resin ingress occurred.
10/23/94 - Unit 2, Mode 6. During this time period in A2RO4 sulfate was 51 ppb. Core fuel reloading was halted on 10/23 due to RH heat exchanger operability questions, and then resumed later that day.
RCS fluoride and chloride remained less than 6 ppb and the temperature of the RCS was less than 140 F, which is too low for the thermal decomposition of cation resin. The elevated sulfates are attributed to resolublization of sulfates within the RCS following cooldown of the RCS. No resin ingress occurred.
I1/1/94 - Unit 2, Mode 5. During this time period in A2RO4, sulfate was 58 ppb. The RCS loops were being fdled on 11/1. RCS F and C1 were less than 6 ppb, RCS Na was 4 ppb, and the temperature of the l
RCS was less than 140 F, which is too low for the thermal decomposition of cation resin. The sulfates are l
attributed to the resolublization of sulfates in the RCS following cooldown of the RCS. No resin ingress l
occurred.
I1/10/94 - Unit 2, Mode 5. During this time period in A2RO4, sulfate was 59 ppb. Oxy,en was being degassed from the RCS on 11/10. RCS fluoride and chloride were less than 5 ppb RCS pH and conductivity values remained stable. No resin ingress occurred.
3/19/96 - Unit 2, Mode 5, During this time period in A2ROS, sulfate was 64 ppb. Following the hydrogen peroxide addition to promote crud burst, elevated sulfates were noted. RCS fluoride and chloride were less than 8 ppb and the temperature of the RCS was less than 140*F, which is too low for the thermal decomposition of cation resin. These elevated sulfates are attributed to resolublization of sulfates in the RCS following cooldown of the RCS and the crud burst. No resin inc.ress occurred.
3/24/96 - 3/26/96 - Unit 2 Mode 5. During this time period in A2ROS, sulfates ranged from 50 to 318 ppb. The RCS cavity was being filled at this time. RCS fluoride and chloride were less than 10 ppb and the temperature of the RCS ns less than 140tF, which is too low for the thermal decomposition of cation j
resin. These elevated sulfates are attributed to resoltMization of sulfates in the RCS following cooldown i
i of the RCS. No resin ingress occurred.
l l
Byron Units 1 and 2 The review period for Bymn Units I was from 1984 to the present, and for Unit 2 was from 1986 to the present.
There were three instances where the screening critena were execx:ded. The date, description, technical assessment, and disposition of each excursion are as follmvs.
i I 1
Byron Unit !
1/11/85 -Unit 1 - Chloride 195 ppb Chloride was the only parameter identified as elevated, the remaining parameters were stable as compared to previous sampling results. The chloride spike was due to organic sohent breakdown at 380 degtees F upon unit heat up.
8/9/85 - Unit 1 -Dissohed Oxygen 801 ppb Based on review of all the data for the day, it appears that this was a data entry error. The boron results were inadvertendy recorded as dissolved oxygen.
4/15/96 -Unit 1 -Sulfates 59 ppb The unit was in a refuel outage, samples were being obtained from the residual heat removal system and refueling aethities had started. The sulfate increase was attributed to the opening the transfer canal in support of refueling activities. Samples analyzed prior to and aner 4/15/96 were well below the 50 ppb limit. There were no appreciable changes in any of the other parameters monitored.
Byron Unit 2 No instances where the screemng criteria were exnwW were identified for Unit 2.
Zion Units 1 and 2 The review period for Zion Units 1 and 2 was from 1973 to the present. There were 37 instances where the screening criteria were exceeded for Unit I and 34 instances where the screening criteria were excx:eded for Unit
- 2. The date, description, technical assessment, and disposition of each excursion are as follows:
Zion Unit 1 1
February 9,1989 - During ZlFO2, the reactor coolant dissolved oxygen concentration was measured 0.300 ppm while in hot shutdown. The plant continued into cold shutdown the following day and remained there until March 3. This event is not considered to be indicative of a resin intrusion.
2 May 11 through May 12,1986 - For two days during a cycle 9 cold shutdown, the RCS chloride concentration was above 0.15 ppm. The highest measured concentration was.800 ppm. The chloride increase was coincident with mosing into Mode 5 for a main steam valve repair. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. Reactor coolant chloride concentration was returned to normal on May 13. This event is not considered to be indicative of a resin intrusion.
3 May 8 through May 17,1985 - While in cold Fhutdown during ZlRO8, reactor coolant chloride concentration became elevated. From May 8 to May 17, reactor coolant chloride sanged between 0.26 ppm and 0.476 ppm. The coolant temperature was below that uhich is required for cation resin thermal decomposition and there was no coolant pH depression. The chloride level was reduced by placing a new CVCS mixed bed demineralizer in ser ice prior to startup. This event is not considered to be indicative of a resin intrusion. l
t 1
4 August 6 through August 18,1984 - Reactor coolant chleride concentration was elevated while in cold shutdown during Cycle 8. The chloride concentration ranged between 0.160 ppm and 0.175 ppm on August 8,11, and 13. The chloride increase was coincident with changing residual heat removal trains. The chloride inventory originated in the refueling water storage tank. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. The chloride concentration was reduced to less than 0.15 ppm on August 14 after placing the CVCS demineralizer in service. This event is not considered to be indicative of a resin intmsion.
a 5
November 18 through November 22,1983 - For five days while in cold shutdown during ZlRO7,
the RCS chloride concentration ranged from 0.18 to 0.20 ppm. Although no refueling water storage tank chloride data exists for September through November 1983, the tank chloride concentration was 0.248 ppm in August. The coolant temperatore was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. Coolant chloride was reduced via dilution on filling of the reactor coolant system. This event is not considered to be indicative of a resin intrusion.
6 July 5 and July 6,1982 - During the cycle 7 startup the reactor coolant dissolved oxygen
<:oncentration was recorded at or greater than 0.100 ppm. These measurements are considerul to be in error as there was between 20 and 30 cc/kg dissolved hydrogen in the coolant at the time of the dissolved oxygen measurement. Radiolytic recombination with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. This event is not considered to be indicative of a resin intrusion.
7 February I through Februe.ry 6,1981 - Reactor coolant chloride concentration became elevated while in cold shutdown during ZlROS. Between February I and February 6, the reactor coolant chloride ranged between 0.15 ppm and 0.68 ppm. Although not noted, the source is suspected as being chloride contamination in the refueling water storage tank during the reactor cavity flood up.
The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. The chloride concentration was reduced to 0.014 ppm by February 8. This event is not considered to be indicative of a resin intru ton.
8 August 8 and i1,1981 - During cycle 6 while at full power, the reactor coolant dissolved oxygen concentration was reported at 0.100 ppm. The values on August 10 and 12 were reported as "not detected". Suspect 0.100 ppm values are in error. This event is not ccasidered to be indicative of a resin intrusion.
9 January 13 through February 18,1980 - While in cold shutdown during ZlRO4, the reactor coolant chloride concentration became elevated. Coolant chloride concentration ranged as high as 0.53 ppm. The most likely contamination source was the refueling water storage tank. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no increase in specific conductivity. The coolant chloride concentration was returned to normal prior to power operation. This event is not considered to be indicative of a resin intrusion 10 January 23 through 28,1979 - While at power dming cycle 4, the reactor coolant dissolved oxygen concentration was measured at 0.10 ppm. The dissolved hydrogen concentration ranged between 16 and 29 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. The high dissolved oxygen concentrations reported are attributed to analytical error as was noted in the log sheets.
t 11 March 2,1979 - During cycle 4, the reactor coolant chloride concentration was recorded at 0.25 ppm following a unit trip. The source is suspected as being the boric acid delivered to the system following the trip from either the refueling water storage tank or boric acid tank. Although system j
temperatures existed that could cause thermal decomposition of cation resin there was no depression of coolant pH. This event is not considered to be indicative of a resin intrusion.
12 March 15 and 17,1979 - During cycle 4 on March 15 and 17, the reactor coolant dissolved oxygen concentration was reported at 100 ppb. Each day was bounded by days at below detectable concentrations and RCS hydrogen was reported at 15 to 38 cc/kg around this time period.
]
The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. Suspect measurement error.
13 March 31 through April 3,1979 - During a cycle 4 cold shutdown, the reactor coolant chloride concentration became elevated. The reactor coolant chloride was measured at 0.37 ppm,0.16 ppm, and 0.175 ppm on March 31, April 1, and April 3 respectively. Thu source of the chloride is
. suspected as being the boric acid delivered to the system from either the refucIing water storage tank or boric acid tank following the reactor trip. There was a significant specific conductivity spike (40 uS) associated with this event however there was no coolant pH depression and the coolant temperatu-e was below that which is required for cation resin thermal decomposition. This event is not considered to be indicative of a resin intrusion.
14 June 9 through June 14,1979 - During a cycle 4 cold shutdown the reactor coolant chloride became elevated. From June 9 through June 12 chloride concentration ranged between 0.150 ppm and 0.195 ppm. The source of the chloride is suspected as being the boric acid delivered to the system from either the refueling water storage tank or boric acid tank following the reactor trip. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no pH depression. This event is not considered to be indicative of a resin intrusion.
l' 15 October 13 to November 25,1979 - While in cold shutdown during ZlRO4, the reactor coolant chloride concentration became elevated. Chloride concentration ranged up to 0.40 ppm. The source of the chloride is suspected as being the boric acid delivered to the system from either the refueling water storage tank or boric acid tank. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. This event is not considered to be indicative of a resin intmsion.
l 16 January 18 to January 19,1978 - While at full power during cycle 3, the reactor coolant dissolved oxygen was reported at 0.100 ppm. Reactor coolant dissolved hydrogen concentration at the time was measured at 48 cc/kz The radiolytic recombination of oxygen with dissolved nydrogen would likely have prevented free dissolved oxygen from existing under Mode I conditions. The 0.100 ppm measurement was likely a sampling or analyses error.
l 17 April 5, 9,16,1978 - During cycle 3 power operation, the reactor coolant dissolved oxygen was
)
reported at 0.100 to 0.150 ppm. The reactor coolant dissolved hydrogen concentration ranged from 18 to 33 cc/kg at the time. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing under Mode 1 conditions. The measurements were likely m error.
18 May 7,1978 - During cycle 3 power operation, the reactor coolant dissolved oxygen was reported at 0.100 ppm. The reactor coolant dissolved hydrogen concentration ranged from 25 to 27 cc/kg at the time. The radiolytic recombination of oxygen with dissolved hydrogen would likely have j
prevented free dissolved oxygen from existing under Mode I conditions. The 0.100 p,m measurement was likely a sampling or analysis error.
19 July 10, 1978 - During cycle 3 power operation, the reactor coolant dissoh ed oxygen reported at
.100 ppm. The reactor coolant dissolved hydrogen concentration was measured at 26.7 cc/kg. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing under Mode I conditions. The 0.100 ppm measurement was likely a sampling or analysis error.
20 October 14 to October 28,1978 - During ZlRO3, while filling the reactor coolant system from the refueling water storage tank, coolant chloride concentration was measured as high as 0.27 ppm.
High concentrations of chloride in the refueling water storage tank were the result of elevated boric acid evaporator concentrate chloride concentrations. Reactor coolant lithium concentration and conductivity were also elevated due to evaporator concentrate contamination.. There was no coolant pH depression. This event is not considered to be indicative of a resin intrusion.
21 November 13,1978 - Reactor coolant dissolved oxygen was reponed at.100 ppm. The reactor coolant dissolved hydrogen concentration was reported at 24 cc/kg. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. The 0.100 ppm measurement was likely a sampling or analysis error, i
22 December 5 and December 6,1978 - Following a cycle 4 reactor trip and safety injection on December 4, reactor coolant chloride concentration ranged between 0.10 ppm and 0.22 ppm. The boric acid injection tank was identified as the source of the contaminant. There was no coolant pH depression. This event is not considered to be indicative of a resin intrusion.
23 September 16 through November 29,1977 - During 21RO2 the reactor coolant chloride concentration ranged from 0.15 to 0.65 ppm. The source appeared to be chloride contamination of the boric acid in the storage tank system and refueling water storage tank. Specific conductivity was also elevated between September 16 and September 25 ranging form 26 uS to 61 uS. Elevated conductivity is assumed to be the result of high sodium concentration from sodium contamination of the refueling water storage tank. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. This event is not considered to be indicative of a resin intrusion.
24 March 15 through May 18,1976 - Throughout the cold shutdown period of ZlROI reactor coolant chloride concentration was elevated. The chloride increase was coincident with RCS boration. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. This event is not considered to be indicative of a resin intmsion.
25 June 9 through June 15,1976 - During ZlROI while in hot shutdown, the reactor coolant dissolved oxygen concentration was measured at 0.10 ppm. Reactor coolant dissolved hydrogen was reported at 35 to 39 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. The 0.100 ppm measurements were likely due to sampling or analyses error.
26 June 16,1976 - During ZlROI on June 16 the reactor coolant cnloride concentration was reported at 155 ppb. Although the coolant contained measurable chloride concentration, the June 16 concentration was twice that on June 15 and June 17 suggesting that the measurement may have been biased high. No other chemistry changes were noted. There was no coolant pH depression.
This event is not considered to be indicative of a resin intrusion.
27 August 5,1976 - While at power in cycle 2 the reactor coolant chloride concentration increased to 0.175 ppm. No other chemistry changes were noted. The coolant chloride inventory was removed within 2 days. There was no coolant pH depression at the time. This went is not considered to be indicative of a resin intrusion.
j 7
l
4 28 September 22, 23,1976 - During cycle 2 power operation the reactor coolant dissolved oxygen was reported at 0.100 ppm. The reactor coolant dissolved hydrogen concentration was reported at 34 cc/kg. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. The 0.100 ppm measurement was likely a sampling or analyses error.
29 October 1 through October 8,1976 - During a cycle 2 cold shutdown the reactor coolant chloride concentration became elevated. On October I and 8 the concentration was measured at.150 ppm.
The chloride increase was coincident with the boration of the reactor coolant system. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. This event is not considered to be indicative of a resin intrusion.
30 March 5 through March 24,1975 - While in a cycle I cold shutdown the reactor coolant chloride concentration became elevated ranging as high as 0.22 ppm. The refueling water storage tank sia by the residual heat removal system is the suspected source. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no coolant pH depression. This event is not considered to be indicative of a resin intrusion.
31 April 29,1975 - During cycle 1 power operation the reactor coolant dissolved oxygen was reported at.100 ppm. The unit was at 84% power with a reported 25 cc/kg dissolved hydrogen. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. The 0.100 ppm measurement was likely a sampling or analysis error.
32 June 7,1975 - During a cycle I cold shutdown the reactor coolant chloride concentration was reported at.36 ppm. Aner review of coolant data prior to and after this measurement no apparent cause could be identified. Suspect analysis error.
33 January 9 through Arvil 2,1974 - While in a cycle I cold shutdown, the reactor coolant chloride concentration and i,pecific conductivity were elevated. Coolant chloride concentration ranged between 0.15 and 0.275 ppm during the period. A significant rise in specific conductivity started 35 days into the cold shutdown reaching a maximum of 119 uS. A review of chemistry data during this period indicated that a significant amount of sodium may have entered the system from the refueling water storage tank. Computer codes estimating solution pH and specific conductivity j
predict that as much as 70 ppm sodium may have existed at this time. Specific conductisity was i
returned to normal prior to restart of the plant. There was no coolant pH depression during this period. This event is not considered to be indicative of a resin intrusion.
34 June 30 through July 9,1974 - As a result of a cycle 1 down power on June 30, an increase in reactor coolant lithium, chloride and specific conductivity occurred. The transient began when the coolant boron concentration was increased 30 ppm to 975 ppm. Coincident with the boron increase was an increase in lithium and chloride and specific condrnivity. On July I with the reactor power at 10%, the coolant boron concentration was increased to 944 ppm. Coincident with this boron increase was another increase in lithium, chloride and specific conductivity. Although no refueling water storage tank data is available, a pattern of increasing coolant chloride concentration and increased specific conductivity following coolant borations is apparent. It is reasonable to assume that the chloride, lithium and other possible contaminants (sodium) were delivered to the reactor coolant system during the June 30, July I boration. Computer codes estimating coolant pH and specific conductivity based on lithium and boron concentrations predict values close to those measured by laboratory personnel. Although the existing reactor coolant temperature was sufficient to cause the thermal decomposition of ion exchange resin, there was no coolant pH depression that would indicate that resin had been delivered to the system. This event is not considered to be indicative of a resin intrusion.
8-
1 1
[
35 Eeptember 2, through 5,1974 - During a cycle I hot shutdown outage the reactor coolant chloride concentration was elevated. The chloride concentration reached 0.160 ppm on September 1 The source appeared to be chloride leakage from the CVCS mixed bed. There was no coolant pH depression during this period. This event is not considered to be indicative of a resin intrusion.
36 November 7, 8,1974 - For two days during a cycle I hot shutdown outage, the reactor coolant chloride concentration was elevated. The increase was coincident with a 168 ppm boron increase.
Although no refueling water storage tank data is available, chloride contamination in the refueling water storage tank boron is the most likely source. There was no coolant pH depression during this period. This event is not considered to be indicative of a resin intrusion.
37 May 16 to May 20,1973 - Plant was in cold shutdown. Reactor coolant lithium concentration was elevated from baseline substantially during this period. Specific conductivity generally followed i
this trend as well. Suspect this may have been initial lithium hydroxide chemical additions for initial plant startup. Computer codes estimating pH and specific conductivity predicting values based on lithium and boron concentrations were close to those values measured by laboratory personnel.
Zion Unit 2 1
October 13 and 14,1988 - A reactor trip occurred on October 12,1988 initiating the Z2R10 refueling outage. While in hot shutdown the reactor coolant dissobed oxygen was 0.100 ppm. The reactor coolant dissolved hydrogen concentration was 13.9 cc/kg on October 14. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. The 0.100 ppm measurement was likely a sampling or analysis error.
2 April 18,1987 - During Z2RO9 while in cold shutdown, the reactor coolant chloride concentration was measured at 0.175 ppm. Just prior to the increase, the refueling water storage tank, whici-contained 0.185 ppm chloride, was delivered to the reactor coolant system for cavity flood rp. On April 19 coolant chloride concentration was reduced to 0.105 ppm. This event is not considered to be indicative of a resin intrusion.
3 May 26,1987 - High chloride concentration continued to exist in cold shutdown following core reload during Z2RO9. The coolant chloride concentration was 0.17 ppm on May 26. Coolant chloride concentration was reduced to 0.035 ppm May 31. This event is not considered to be indicative of a resin intrusion.
4 June 4,1987 - High chloride concentration continued to exist in cold shutdown following core reload during Z2RO9. On June 4 coolant chloride concentration was 0.190 ppm. Coolant chloride concentration was reduced to 0.045 ppm and 0.088 ppm on June 3 and June 6 respectively. This event is not considered to be indicative of a resin intrusion.
5 October 16 through October 18,1985 - While in cold shutdown during Z2RO8, the reactor coolant chloride was 0.185,0.150 and 0.155 ppm on October 16,17, and 18 respoctively. Flood-up of the cavity from the refueling water storage tank occurred during the October 10 through 15 time period. The refueling water storage tank chloride concentration was at 0.137 ppm on October 8 two days prior to the cavity flood-up. The refueling water storage tank is considered to be the source of the contamination. This event is not considered to be indicative of a resin intrusion.
6 November 16,18, and 30, and December 6,1985 - While in cold Shutdown during 22RO8, reactor coolant chloride concentration was recorded at 0.180,0.150,0.170 and 0.150 ppm on November 16,18, and 30 and December 6 respectively. On December 16, a new mixed bed demineralizer was placed in service and on December 17 reactor coolant chloride concentration was at < 0.018 ppm.
l This event is not considered to be indicative of a resin intrusion.
9
l l
7 April 19 through May 26,1984 - While in cold shutdown during the Z2RO7, the reactor coolant chloride concentration ranged from a low of 0.31 ppm to a high of 0.96 ppm. The increase in coolant chloride was coincident with flooding the reactor cavity from the refueling water storage tank. The refueling water storage tank chloride concentration at the time was between 0.074 ppm l
0.95 ppm. On May 26, a new CVCS mixed bed demineralizer was placed in senice. On May 27, I
one day after the new CVCS mixed bed was placed in service, the reactor coolant chloride concentration was recorded at 0.106 ppm. This event is not considered to be indicative of a resin l
intmsion.
I 8
March 30, April 4, April 7, April 8, and April 29,1983 - While in cold shutdown during Z2RO6, the reactor coolant chloride concentration for the dates indicated was 0.150,0.183,0.156, 0.180, and 0.150 ppm respectively. The refueling water storage tank chloride concentration on February 8 was 0.240 ppm. On March 27, reactor coolant conductivity increased to 27.5 uS coincident with the reactor cavity flood-up from the refueling water storage tank. Fuel moves began on March 30 and continued into early April. This event is not considered to be indicative of a resin intrusion.
9 January 4,1982 - While in hot shutdown during cycle 6 the reactor coolant dissolved oxygen concentration was reported at 0.100 ppm. The reactor coolant dissolved hydrogen concentration was above 35 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented '.ee dissolved oxygen from existing. The 0.100 ppm measurement was likely a sampling or analysis error.
10 April 8,1982 - While at power operation during cycle 6 the reactor coolant dissolved oxygen concentration was reported at 0.150 ppm. The reactor coolant dissolved hydrogen concentration was reported at between 33 and 35 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing.
The 0.150 ppm measurement was likely a sampling or analysis error.
1 11 April 25,1982 - While at power operation during cycle 6 the reactor coolant dissolved oxygen concentration was reported at 0.10 ppm. The reactor coolant dissolved hydrogen concentration ranged between 33 and 46 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxyten from existing. The 0.10 l
ppm measurement was likely a sampling or analysis error.
12 May 28,1982 - While at power operation during cycle 6, the reactor coolant the reactor coolant chloride concentration was reported at 0.170 ppm and pressurizer chloride concentration at 0.150 ppm. The CVCS demineralizer efIluent chloride concentration and all other reactor coolant chemistry parameters remained stable. The reactor coolant was reported to have no detectable chloride for days prior to and aller May 28. The elevated chloride measurements are suspected as being in error.
13 October 3,1982 - During a cycle 6 cold shutdown the reactor coolant chloride concentration was reported at 0.170 ppm and pressurizer chloride concentration at 0.210 ppm. A 32 ppm reactor coolant boron concentration change was coincident with the chloride increase. The refueling water storage tank chloride concentration was reported at 3 ppm during this period. Although not noted, both increases can be explained by a residual heat removal train swap. No other substantial chemistry changes are noted. Coolant chloride concentration was returned to normal prior to re-start. This event is not considered to be indicative of a resin intrusion.
t 14 November 22, 23,1982 - While at power operation during cycle 6 the reactor coolant dissolved l
oxygen concentration was reported at > 0.10 ppm. The reactor coolant dissolved hydrogen concentration ranged between 39 and 50 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing.
l The > 0.10 ppm measurement was likely a sampling or analyses error.
15 September 17 through November 6,1981 - While in cold shutdown during Z2ROS stistantial transients occurred involving reactor coolant chloride, lithium and conductisity. Rea >r coolant chloride concentration increased from "none detectable" to 0.46 ppm. Some ch!cride, :sults later l
in this transient were approximately 1 ppm. Conductivity incr:ased from a baselira # H S to approximately 50 uS. Later in the transient conductivity incre ased well above 1% uS. Lithium increased from a baseline of about 0.8 ppm to 2 to 3 ppm. Altrough no refnimg wcter storage tank sodium or lithium concentrations are available, the most likdy explanation of this cicat is i
gross sodium and chloride contamination of the refueling water stoiage tank. Based on measured lithium and boron concentrations, computer calculated pH codes accurately predict the measured specific conductivity and pH if 60 ppm sodium were prer2nt. The coolant temperature was below that which is rec l aired for cation resin thermal decomposition. This event is not considered to be indicative of a resin intrusion.
l 16 May 13 through June 30,1980 - While in cold shutdown daring Z2ROS the reactor coolant l
chloride concentration became elevated. Reactor coolant chloride concentration reached a reported peak of approximately 3 ppm for 2 days. Reactor coolant conductivhy was elevated as high as 55 l
uS for about 2 weeks. Although no refueling water storage tank sodium data is availabk, the most l
likely cause of the event was significant sodium contamination of the refueling water storage tank which entered the reactor coolant system during the residual heat exchanger pump operation. The coolant temperature was below that which is required for cation resin thermal decomposition, i
therefore this event is not considered to be indicative of a resin intrusion.
17 January 24,1979 - While at power during cycle 3 the reactor coolant dissolved oxygen reported at l
0.100 ppm. Reactor coolant hydrogen was reported as 12 cc/kg. The radiolpic recombination of l
oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing.
(
Remarks on the data sheet indicate value is due to poor sampling.
18 February 12,15,1979 - During a cycle 3 hot shutdown, the reactor coolant chloride and specific conductivity were elevated. On February 12 the coolant chloride was reported at 0.250 ppm and specific conductivity increased from a baseline of 20 uS to 42 uS. The increases were coincident l
with borating the system to 674 ppm. Although no refueling water storage tank data is available, the most likely source of the increases are due to sodium and chloride contamination of the boric acid in the tank. There was no pH depression during this event. On Febmary 15 the reactor coolant dissolved oxygen was reported as 0.100 ppm. Remarks on the data sheet indicate the oxygen value may be in error. This event is not considered to be indicative of a resin intrusion.
19 March 11 through April 13,1979 - While in cold shutdown d::ing Z2RO3 the reactor coolant chloride, fluoride and specific conductivity became elevated. Reactor coolant chloride concentration ranged as high as 0.87 ppm during this period. The increase in coolant chloride, fluoride and specific conductivity was coincident with borating the system on March 11. Reactor coolant conductivity increased from a baseline of about 11 uS to about 45 uS. Lithium l
concentration increased from 0.382 ppm to 1.5 ppm and continued to increase to a peak of approximately 3.3 ppm over the next several days. Although no refueling water storage tank data is available, the most likely source of the increases was sodium and chloride contamination of that l
tank or residual heat removal suction piping coming from the tank. The coolant temperature was l
below that which is required for cation resin thermal decomposition and there was no pH depression. This event is not considered to be indicative of a resin intrusion.
20 December 29,1979 - During cycle 4 while in a cold shutdown reactor coolant chloride concentration was reported at 0.160 ppm. No other baseline changes in reactor coolant chemistry were noted. Chloride concentration bracketing this day were typical of the day to day baseline.
Suspect a sampling or analysis error.
l
-11 l
l
l
\\
j 21 February 11 thrcugh March 13,1978 - While in cold shutdown during Z2RO2 the reactor coolant i
chloride concentration and specific conductivity became elevated. Remarks indicate the source l
appeared to be chloride contamination of the BAST via the RWST. Upon further resiew it appears l
that the most likely source of the increases were due to sodium and chloride contamination of that tank or residual heat removal suction piping coming from the tank. The increased specific l
conductivity is suspected to be related to a high sodium contamination. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no pH l
depression. This event is not considered to be indicative of a resin intrusion.
22 August 6,1978 - During cycle 3 while at power the reactor coolant dissolved oxygen concentration reported as 0.100 ppm. The reactor coolant dissolved hydrogen concentration was 28 cc/kg. The I
radiolytic recombination of oxygen with dissolved hydrogen would likely has e prevented free dissolved oxygen from existing. Surpect sampling or analysis error.
23 October 9,1978 - During cycle 3 while at power the reactor coolant fluoride concentration was 0.190 ppm. No other baseline changes in reactor coolant chemistry were noted. Fluoride l
concentrations prior to and following October 9 were normal. Suspect analytical error, l
24 October 24,1978 - During cycle 3 while at power the reactor coolant reactor coolant dissolved oxygen was reported at 0.100 ppm. The reactor coolant hydrogen concentration was 27 cc/kg. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. Suspect sampling or analysis error.
25 January 15,1977-through March 21,1977 - While in cold shutdown during Z2ROI the reactor coolant chloride concentration became elevated. Coolant chloride concentration ranged as high as 0.58 ppm during the period. CVCS demineralizer chloride leakage is suspected as the source.
Periodic operation of CVCS revealed a pattern of chloride increases while letdown was in 1
operation. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no pH depression. This event is not considered to be indicative of a t
resin intrusion.
l 26 January 18,1976 through February 14,1976 - During cycle I while in a cold shutdown, the reactor coolant chloride concentration became elevated. Chloride ranged from 0.31 ppm on January 18, peaked at 0.800 ppm on February 4 and then trended down below 0.150 ppm after February 14.
l (Note: 7.600 ppm chloride peak on February 5 is suspect, and is most likely.760 ppm, not 7.60 ppm based on RCS chloride trends) Specific conductivity ranged as high as 65 uS. The chloride ingress and conductivity rise was coincident with the delivery of boron to the system. Although no refueling water storage tank data is available, the most likely source of the increases were due to sodium and chloride contamination of that tank or residual het removal piping. The increased specific conductivity is suspected as being the result of high sodium contamination. The coolant l
temperature was below that which is required for cation resin thermal decomposition and there was no pH depression. This event is not considered to be indicative of a resin intrusion.
27 March 27,1976 - During cycle I while at power the reactor coolant dissolved oxygen was > 0.100 l
ppm. The coolant dissolved hydrogen concentration was recorded at 35 cc/kg on March 24 and I
33.9 cc/kg on March 31. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. Suspect sampling or analysis error.
28 April 22,1976 - During a cycle I cold shutdown the reactor coolant chloride concentration was recorded at 0.270 ppm. All other chemistry parameters were stable. The coolant chloride concentration recorded for April 21 and 23 was 0.055 ppm and 0.025 ppm respectively. Suspect sampling or analysis error. This event is not considered to be indicative of a resin intrusion.,
i l
l
I i
l
(
29 June 9,1976 - During a cycle I cold shutdown the reactor coolant chloride concentration was l
recorded at 0.270 ppm. The coolant chloride concentration recorded around June 9 was 0.043 ppm and 0.040 ppm. Suspect sampling or analysis error. This event is not considered to be indicative of a resin intrusion.
l l
30 August 30,1976 - During cycle I while at power the reactor coolant dissolved oxygen was 0.100 ppm. The coolant dissolved hydrogen concentration ranged between 33 cc/kg and 39 cc/kg during this period. The radiolytic recombination of oxygen with dissolved hydrogen would likely have prevented free dissolved oxygen from existing. Suspect sampling or analysis error.
f l
31 September 28,1976 - During cycle I while in cold shutdown, the reactor coolant chloride l
concentration was recorded at 0.152 ppm. Although no refueling water storage tank data is available, the most likely source of the increase was chloride contamination of that tank or residual heat removal suction piping. The coolant temperature was below that which is required for cation resin thermal decomposition and there was no pH depression. This event is not considered to be indicative of a resin intrusion.
(
32 April 8,1975 - During cycle I while in cold shutdcwn, the reactor coolant chloride concentration l
was recorded at 0.174 ppm and specific conductivity at 25.5 uS. These increases were coincident l
with boration of the system. Although no refueling water storage tank data is available, the most l
likely source of the increases were due to sodium and chloride contamination of that tank or I
residual heat removal suction piping. The increased specific conductivity is suspected as being the re.; ult of high sodium contamination. The coolant temperature was below that which is required for l
cation resin thermal decomposition and there was no pH depression. This event is not considered to l
be indicative of a resin intrusion.
December 7,1975 - During cycle I while at power the reactor coolant dissolved oxygen recorded at 33 O.100 ppm. The coolant dissolved hydrogen concentration ranged between 22 cc/kg and 30 cc/kg during this penod. The radiolytic recombination of oxygen with dissolved hydrogen would likely l
have prevented free dissolved oxygen from existing. Suspect sampling or analysis error.
1 34 December 13 through 18,1973 - With the unit in cold shutdown the reactor coolant lithium concentration and specific conductivity became elevated. The lithium and specific conductivity increases were not coincident. The specific conductivity increased from 9.4 uS on December 12 to 81 uS on December 13. The lithium concentration increased from 0.53 ppm on December 13 to 5.8 l
ppm on December 15. The reactor coolant boron and chloride concentration were somewhat stable during this period. The events that produced these transients are not clearly understood. However, the coolant temperature was below that which is required for cation resin thermal decomposition and there was no pH depression. This event is not considered to be indicative of a resin intrusion.
l l i
+u d
l 1
i l
I I
i l
t ATTACHMENT H l
l i
1 I
1 j
i I
L -
WESTINGHOUSE NON-PROPRIETARY CLASS 3 i
WCAP-14902 Background Material for Response to NRC Generic Letter 97-01:
Reactor Vessel Closure Head Penetration Integrity for the Westinghouse Owners Group W. H. Bamford B. A. Bishop J. F. Duran D. E. Boyle June 1997 Reviewed by:
bh
/
AO G.V.Rao' ineering Materials Technology
!~3 1
1 1
Approved C
D. A. Howe'l, Manager \\
Mechanical Systems integration l
Westinghouse Electric Corporation Nuclear Services Division P.O. Box 355 Pittsburgh, PA 15230 C1997 Westinghouse Electric Corporation l
All Rights Reserved Rev.O July 1997 o 0709. doc:1b:07/14/97 Qoo: '?o s c r s
.. ~.
f I
4 TABLE OF CONTENTS k
i l.
Executiv e S u mm a ry...................................................................................................
1.0 I nt rod u ct io n...................................................................;,........................................... 1 - 1 l
2.0 Development of a Crack Growth Rate Model for Alicy 600 Head Penetrations.......... 2-1 3.0 Westinghouse Crack initiation Model Development and Crack initiation Testing....... 3-1 4.0 Ref e re n ce s................................................................................................
4-1 I
l l
t o
c 1
4 1
l l
i I
l f-l L
i r
Rev.O i
July 1997 c:0709.coc:1b:07/14/97
EXECUTIVE
SUMMARY
This report is intended for use in response to NRC Generic Letter 97-01. Cracking in Alloy 600 reactor vessel head penetrations is a relatively new issue to the nuclear industry. The issue was first brought to the world's attention in 1991 when, after 10 years of operation, a leak was detected during a hydrotest of the reactor coolant system at the Bugey Unit 3 power plant in France. Since then a significant number of studies and research programs have been funded by the industry to determine the causes of the problem and develop strategies for repair and management.
l Through these programs and subsequent studies it was concluded that reactor pressure vessel head CRDM penetration cracking at Bugey Unit 3 is induced by is a thermally activated stress corrosion mechanism operative in primary water environments, more commonly known as l
primary water stress corrosion cracking (PWSCC). Based on conservative evaluation resuits, the NRC and industry concluded that PWSCC cracks were most likely to initiate from the inside surface of the penetrations, in the axial orientation, and would take at least six years to propagate through the wal! under the typical plant operating conditions. Fracture mechanics evaluations have determined that the crack is non-critical until its axial length reaches 8.5 inches to 20 inches, depending on plant design. Therefore this issue is an economic one, and does not constitute a serious challenge to plant safety.
Extemal circumferential cracking is less probable. It may occur only in the presence of an above the weld through-wall crack, with active leakage. Assuming coolant is present on the outer diameter of the penetration, one conservative analysis estimated that it would take more than 90 years before penetration failure would occur. In the presence of reactor coolant, corrosion wastage of the alloy steel RV head is possible. Conservative evaluations estimate that it would take longer than six years after a through-wall crack occurs before the code structural integrity margin for the RV head would be impacted by corrosion. It was concluded l
that periodic visual inspection of the RV head in accordance with Generic Letter 88-05 is adequate to maintain plant safety, and sufficient to detect leakage prior to significant penetration cracking and vessel head corrosion.
Worldwide, approximately 5,200 Alloy 600 RV head penetrations have been inspected since the l
first cracking was observed in 1991. Approximately 2 percent of these penetrations are l
reported to be cracked. Most of the cracks were observed in French RV head penetrations. If the French inspection records are removed from the inspected population, the percentage of i
head penetrations with indications is only about 0.5 percent. Only one plant worldwide has experienced PWSCC head penetration through-wall leakage, and this was from a single penetration.
i Specialized NDE methods have been developed and verified using mock-ups to ensure l
accurate inspections. Flaws were introduced into the mock-up penetrations by artificial means.
The ability of these NDE methods to detect and size the potential PWSCC indications in the vessel head penetrations was demonstrated. Flaw acceptance criteria were established by the industry, and approved by the NRC staff.
The Westinghouse Owners Group has developed methods to evaluate the PWSCC l
susceptibility and the probability of a penetration initiating a crack, or a leak, as a function of l
l f
Rev.O ii July 1997 c:\\3709.cloc:1b:07/14/97
l 4
plant operation time. This information has been used to evaluate the need for inspection of the j
reactor vessel head penetrations or other appropriate actions.
Through participation in WOG and U.S. industry programs the Westinghouse plant owners have taken a proactive approach to address the cracking issue in RV head penetrations. This approach is based on the conclusion that the issue is not an immediate safety concem, because (1) the PWSCC process is slow; (2) the allowable or critical flaw size is large; (3) leak-before-break (LBB) will occur to allow safe shutdown of a plant and (4) at least six additional years of operation with a penetration leak is required before ASME Code structural margins are l
challenged.
l In addition to the material contained in this report, detailed integrity assessments have been i
completed for all Westinghouse plants, and these results are being incorporated into an integrated response to the Genaric Letter 97-01, which is being prepared in cooperation with i
the Nuclear Energy institute. This response will be transmitted to the NRC by the end of 1997.
j l
l l
l l
l l
l l
l l
l Rev. O iii July 1997 c:U709. doc:1b:07/14/97
- g 1
3 j
1.0 ' INTRODUCTION 1
)
1.1
SUMMARY
OF THE SAFETY EVALUATIONS The purpose of this section is to review the significance of cracking in pressurized water reactor i
(PWR) vessel head penetrations and to describe the management of the issue in response to j
i the recently released NRC Generic Letter 97-01. This report covers the following areas:
1 l
worldwide PWSCC history in head penetrations; safety evaluation conclusions reached by WOG and industry and approved by the NRC relative to PWSCC; and a number of supporting tasks performed by Westinghouse for the WOG concoming this issue. The latest findings on j
this subject are summarized, along with response to specific questions in Generic Letter 97-01.
t l
In February of 1993, Westinghouse and the Westinghouse Owners Group performed an assessment of the continued safe operation of Westinghouse designed NSSS plants in light of l
the cracking that had been reported in French supplied and operated plant reactor vessel head penetrations.
l Westinghouse reviewed the available metallographic and tractographic data from the French j
plant and concurred with the EdF conclusion that the mechanism of degradation of the Bugey 3 reactor vessel penetration was due to primary water stress corrosion cracking.
- l The Westinghouse safety evaluation [1] provided the following elements
- 1. A summary of the vesselheadpenetration stress analyses that focuses on the nature and orientation of cracking that may occurin the Alloy 600 penetration material. The Westinghouse evaluation concluded that the penetration residual stress induced by welding into the reactor vessel head was the initiating source promoting crack initiation and growth in a susceptible microstructure.
- 2. A summary of the crack propagation analysis along with the basis of the prediction methodology. As indicated in Section 2 of this report, continued crack growth testing has confirmad the initial expectations. The analysis also predicted that cracking would be axial and any cracks formed would be limited in extent by the penetration stress field distribution.
The crack lengths predicted were found to be much smaller than the length of cracking required for any instability. The existence of circumerential cracking is unlikely due to the nature of stress distribution in the penetrations (i.e., hoop stress dominates the stress field).
- 3. A description of an assessment of the Westinghouse Owners Group vessels with respect to crack indications reported at Ringhals, Bernau, and various EdFplants. Important parameters applicable for crack initiation (i.e., time, temperature, stress, and material) were compared to those of Ringhals, Beznau and EdF plants. A comparison of susceptibility predictions suggested that the WOG vessels were generally less susceptible than Ringhals.
However, several vessels were found to be more susceptible. Since this initial evaluation, three of these vessels were inspected for penetration cracking. One vessel head was found with cracking in a single penetration and no cracking was found in the penetrations of the other two plants. The level and depth of cracking was found to be covered by the Westinghouse Safety Evaluation.
l Rev.0 1-1 July 1997 0:0709. doc:1b:07/14/97
\\
i i
1 i
+
Y
- 4. A penetration leakage assessment summarizing leak rate vs. crack size. Expectations from i
this evaluation were that (a) leakage would be detected well before cracks extended to their j
critical flaw cize (through-wall, and 8.5 20 inches long) and (b) Boron deposits would be l
r,ignificant enough from small flaws to be readily visible during a Generic Letter 88-05 walkdown.
].
- 5. A vesselhead wastage and structuralevaluation. The evaluation showed that the loss of approximately 1.0 in3 of vessel head material per year could be expected if cracks initiated i
and propagated through wall, however, vessel structural margins would be maintained for at least six additional years following the through wall leak.
1.2 HISTOR! CAL BACKGROUND 1
in 1991, during a hydrotest of the reactor coolant system at the Bugey Unit 3 power plant n 3
France, a leak from the reactor vessel head was detected by acoustic monitoring [2].
i l
Subsequent investigation, by visual examination and destructive testing, revealed that the leak j
came from a through wall flaw in one of the head penetrations. Further inspections on this and many other plants in France led to the discovery of flaws in the head penetrations of seversi i
alants. Examinations confirmed that the problem was directly related to Primary Water Faress l
Corrosion Cracking (PWSCC).
EdF conducted additional CRDM (Control Rod Drive Mechanism) penetration inspections at its nuclear plants, using eddy current techniques for indication detection and ultrasonic methods for defect size de'ermination. Inspection results and metallurgical examinations confirmed PWSCC in CRDM penetrations at several other EdF plants. This was a concem to the French l
regulabry autoWies as well as to the other PWR owners and regulatory authorities around the i
world.
These incidents are similar in nature to what occurred to other Alloy 600 tubular parts used in I
de Reactor Coolant System (RCS). Over the past few years, cracks in Alloy 600 pressurizer heater sleeve penetrations and instrumentation nozzles [3,4) have been reported at non-Westinghouse supplied domestic and French PWR plants. In February 1990 the USNRC issued information Notice 90-10 on this issue (5). The Notice informed PWR utilities of a j
number of incidences of PWSCC of Alloy 600 in applications other than steam generator tubing j
and suggested that utilities review their Alloy 600 applications and implement an augmented j
inspection program as necessary. In 1990, EPRI issued a report (4) which suggested that utilities should identify locations where Alloy 690 is used on the primary side, review the j
material and fabricaion records to assess material susceptibility to PWSCC in terms of microstructure, stress, and environment, and 'mplement an inspection program to detect leakage or cracking with the view of, repitcirig susceptible components, as appropriate.
3 t
The Westinghouse Owners Group (WOG) and Westinghouse initiated and helped to lead a joint industry owners group under NUMARC, now the Nuclear Energy Institute (NEI), beginning in 1992. The group consists of all owners of Pressurizer Water Reactors in the USA along with l
EPRI. This group shared technicalirdormation and deve!ored consistent safety evaluations and evaluation precedures for flaws that may be found dur.ng inspections. The group also worked with EPRI to develop inspection performance demonstrations for the head penetration inspections. The group demonstrated to the US Nuclear Regulatory Commission that cracking Rev.0 1-2 July 1997 o$3709.coc:1b:07/14/97
[
1
)'
on the head penetrations was not an immediate safety issue. The NRC concurred witn the Westinghouts conclusion, stating that vessel head penetration cracking is not an immediate i
safety issue [5).
1.3 INSPECTIONS PERFORMED TO DATE 4
i j
in 1994, two WOG/ Westinghouse PWR plants in the US (Point Beach Unit 1 and D. C. Cook j
Unit 2) voluntarily performed inspections of the CRDM penetrations. The results showed that there were no indications found in Point Boach Unit 1. Three indications were found in a single 4
penetration at D.C. Cook Unit 2. These were significant cracks but considerably smaller than i
the NRC approved acceptance limit.
5 I
in Spring of 1996, D. C. Cook Unit 2 ra-inspected some of their penetrations that had been l
previously inspected and confirmed 6e same indications reported earlier. No new indications 4
were found and the existing indicatien was successfully repaired. Meanwhile, North Anna Unit 1 inspected 20 out of the total complement of 65 penetrations. No indications were found.
I 4
A large number of inspections have been performed on Westinghouse supplied reactor vessel head penetrations throughout the world, and this section will document those inspections, and l
the findings to date.
ASME Code Section XI inspections (VT-3) have been performed for a number years on the head penetration to reactor vessel partial penetration weld, and the weld between the head l
l penetration tube and the control rod drive mechanism (CRDM). While these inspections do not i
cover the Alloy 600 inside diameter surface region of the head penetration directly, they do l
provide surveillance information on the head penetration region, and must be performed on every penetration once every ten years. To date no indications have been reported.
A second series of inspections which have been carried out regularly since 1988 involves visual surveillance of the head for boron deposits which would be evidence of leaks, following NRC Generic Letter 88-05. Some boron deposits have been fcund by this surveillance, but the sources of the leakage were Dgl from cracked head penetrations. Generally these leaks have been associated with mechanical seals or canopy seals on the vessel head.
Westinghouse supplied NSSS plants in Spain,. Sweden, Switzerland, Belgium, Brazil, and Korea have conducted NDE inspections on Reac*or Vessel Head Penetrations. By.the beginning of 1996, some 5200 penetmtions had been inspected worldwide. The results are summarized in Table 1-1. On average, indications were found in approximately 2% of the penetrations that were inspected. Based on Table 1-1, it appears that the rate of indications at U.S. plants.is significantly less than that of tiia French plants. The operating time for the plants of US manufacture where the inspections hava been performed has in most cases been much longer than for the French plants. Of d these inspections, only one penetration was found to have through-wall cracking: the Bugey plant where cracking was first identified.
It will be of interest to examine the history of inspections of the plants of Westinghouse design -
worldwide, as well as the plants of Westinghouse design with US fabrication. A relatively large number of these plants have been inspected, and very few indications have been found.
Outside of France, a total of 39 plants of Westinghouse design have been inspected. Of Rev.0 1-3 July 1997 od3709. doc:1b:07/14C
s approximately 1900 penetrations inspected, only 10 were reported to be cracked, amounting to a less than 0.6 percentage. Of the 39 plants,9 were manufactured in the USA, and for these plants approximately 310 penet ations were inspected with only one reported to be cracked.
Thus, for Westinghouse plants manufactured in the USA, only 0.3 percent of the penetrations t
have been found to be cracked.
l Root cause evaluations concluded that the cracks were caused by PWSCC of the Alloy 600 l
material. Electricite de France (EdF) and Westinghouse concluded that the following factors contributed to the Bugey Unit 3 PWSCC.
l l
Susceptible microstructure produced during manufacturing l
l Surface finish on the inside diameter surface of the penetration L
Stresses induced during welding, which caused ovalization of the penetration l
i 1
l l
l i
}
I Rev.0 1-4 July 1997 c:\\3709.occ:1b:07/14/97
... - ~_
t TABLE 1-1 WORLDWIDE VESSEL HEAD PENETRATION PWSCC INSPECTION RESULTS*
Number of Total No. of Number of Penetrations Rate of Plants Penetrations Penetrations With indication Country inspected in the plants inspected indications Detected" France 47 3225 3213 105 3.3%
Sweden 3
195 190 7
3.7%
Switzerland 2
72 72 2
2.8%
Japan 17 960 834 0
0 Belgium 7
435 435 0
0 Spain 5
325 102 0
0 Brazil 1
40 40 0
0 South Africa 1
63 63 0
0 South Korea 1
65 65 0
0 United States 5
314 217 1*"
0.5%
Total:
89 5694 5231 115 2.0%
Based on data available as of January 1996 (Europe) and July 1996 (U.S.).
" Ratio of number of penetrations with indications detected to number of penetrations inspected.
"* Oconee indications were not counted as cracks, because they had no measurable depth. Eddy current reinspection after one cycle did not indicate any growth l
l Rev.0 15 July 1997 o:U709.coc:1b:07/14/97
.. - -. -. - - - = -. -. - - - - - - -.. _--
l 1.4 WOG AND NUCLEAR INDUSTRY PROGRAMS
SUMMARY
A number of WOG programs were initiated to investigate the reactor vessel head penetration PWSCC issue. The key programs are summarized in Table 1-2. Additionally, selected utility programs have been responsible for the resolution of IGA due to sulfur species, and penetration attachment weld cracking. Domestically, the Babcock and Wilcox Owners Group (BWOG),
Combustion Engineering Owners Group (CEOG), Westinghouse Owners Group (WOG) and the Electric Power Research Institute (EPRI) agreed to combine their efforts as part of the Nuclear Energy Institute's (NEI) Alloy 600 CRDM Head Penetration Cracking Task Force. The purpose of the task force was to evaluate the issue and to recommend appropriate generic actions.
Through this effort, the Owners Groups (OGs) and EPRI have conducted the following tasks:
Perfor.ned safety analyses of vessel head penetration cracking
. Standardized flaw evaluation methods Developed flaw acceptance criteria Developed inspection methodologies to size indications in head penetrations e
Evaluated remedial measures and created probebilistic and economic decision making tools Evaluated leakage effects on the vessel head low alloy steel shell in addition, WOG has developed penetration repair techniques, plant inspection guidelines, and evaluated available leakage detection devices.
The NRC has evaluated the safety analyses and concluded that PWSCC of Alloy 600 head penetration is not an immediate safety concem (6].
Under the programs, research on PWSCC was conducted domestically and overseas, for j
example, as shown in Refs. 3,7,8,9 and 10. The studies focused on & aterial aspects and mechanics. Material aspects, thermomechanical processing effects, material properties, residual stresses, and microstructure were studied. A model of PWSCC susceptibility and cracking probability was developed [10).
l Finite element analyses were p6riormed to determine stresses in the penetrations. The finite element analyses performed included simulation of the whole spectrum of the mechanical fabrication sequences experienced by the RV head penetrations, such as the welding process, hydrotest, atraightening and service loads. The finite element simulations allowed the deterinination of the applied as well as the residual stresses in the penetrations under any given specific geomatrical, material, welding, temperature, and loading conditions. Based on the stress data, PWSCC initiation, crack propagation, and final failure were then evaluated. The analysis also fumished results for the time period required for the PWSCC to penetrate through the wall thickness of the penetration and the critical crack size above which instability would occur, initial crack growth behavior was assumed to be represented by the model developed by P. Scott [11].
l i
Rev.0 16 July 1997 on3709.coc:1b:07/14/97
t Confirmatory crack growth laboratory testing was immediately begun to verify that this initial l
assessment was correct. The integrity model was structured to be applicable to all penetrations l
regardless of product form or vessel fabricator. Subsequent testing to obtain comparison daia in this area was initiated in 1996. The crack growth test results and preliminary crack initiation j
1 test results are discussed in Sections 2 and 3.
l t
i 4j Rev.0 1-7 July 1997 c:0709.coc:1b;07/14/97
TABLE 12
SUMMARY
OF KEY TASKS PERFORMED BY WOG Item Task Gescription Status 1
Root Cause of Cracking C
2 Key Material & Operation Parameters C
3 Elastic Finite Element Analysis:
C Residual / Operational 4
Elastic / Plastic Finite Element Analysis:
C Residual / Operational; 3 Locations 5
Crack Propagation / Acceptable Flaw Size C
Analysis 6
Penetration Leakage & Vessel Head C
Wastage Assessment 7
Safety Evaluation C
8 Plant Screening / Susceptibility Cnteria C
9 Material Microstructure Charactenstics C
10 Leakage Detection Methods Survey C
11 Evaluation of PWSCC Mitigation Methods O
12 Gnnding Effect on Residual Stresses C
13 Development / Evaluation of Repaired C
Configurations 14 OD Crack Assessment C
15 Crack Growth Data and Testing O
16 inspection Timing and Economic Decision C
Tools 17 Penetration Attachment Weld Safety C
Evaluation Report 18 Crack initiation Charactenzation Studies O
19 Residual Stress Measurements C
20 Development of PWSCC Susceptibility C
Ranking Models Key: C = Complete O = Ongoing.
Rev.0 1-8 July 1997 ON1709. doc:1b:07/14/97
2.0 DEVELOPMENT OF A CRACK GROWTH RATE MODEL FOR ALLOY 600 HEAD PENETRATIONS Crack growth rate testing has been underway since 1992 to characterize the behavior of head penetration materials. The " modified Scott model," as described below was initially used for safety eveluation calculations in the NRC submittals made in 1992 and 1993. The goal of this section of the report is to review the applicability of that model in light of the past five years of testing, during which over forty specimens have been tested representing 15 heats Alloy 600 of material. The original basis of the model will be reviewed, followed by all the available laboratory results, and finally a treatment of the available field results.
The effort to develop a reliable crack growth rate prediction model for Alloy 600 began in the Spring of 1992, when the Westinghouse, Combustion Engineering, and Babc.ock and Wilcox Owners Groups were developing a safety case.to support continued operation of plants. At the time there was no available crack growth rate data for head penetration materials, and only a few publications existed on growth rates of Alloy 600 in any product form.
The best available publication was found to be that of Peter Scott of Framatome, who had developed a growth rate modal for PWR steam generator materials [11). His model was based on a study of results obtained by McIlree and Smialowska (12] who had tested short steam generator tubes which had been flattened into thin compact specimens. His model is shown in Figure 2-1. Upon study of his paper there were several ambiguities, and several phone conversations were held to clarify his conclusions. These discussions indicatad that Reference 11 contains an error, in that no correction for cold work was applied to the McIlree/Smialowska data. The revision of the Peter Scott modelis presented below.
An equation was fitted to the data of Reference 12 for the resutts obtained in water chemistries that fell witnin the standard specification for PWR primary coolant. Results for chemistries outside the specification were not used. The following equation was fitted to the data for a temperature of 330*C:
= 2.8 x 1' d' (K-9)' A' m / sec O
at where K is in MPa[m]". This equation implies a threshold for cracking susceptibility, K
= 9 MPa[m)". Correction factors for other temperatures are shown in Table 2-1.
The next step described by Scott [11] in his paper was to correct these results for the effects of cold work. Based on work by Cassagne and Golpi [13), he concluded that dividing the above equation by a factor of 10 would be appropriate to account for the effects of cold work. This step was inadvertently omitted from Scott's paper, even though it was discussed. The revised crack growth model for 330*C then becomes:
da
= 2.8 x 1042 (K-9)'d' m / see This equation was verified by Scott in a phone call in July 1992.
Rev.0 21 July 1997 c:\\3709. doc:1b:07h4/97
~
~.-
4 Scott further corrected this model for the effects of temperature, but his correction was not used in the model employed. Instead, an independent temperature correction was developed based on service experience. This correction uses an activation energy of 32.4 kCal/ mole, which gives a smaller temperature correction than that used by Scott (44 kcal/ mole), and will be discussed in more detail below, Scott's crack growth model for 330*C was independently obtained by B. Woodman of ABB-CE
[14), who went back to the original data base, and had a smaller correction for cold work. His oquation was of a slightly different form:
b = 0.2 exp [A + B in {In (K-C)})
I dt i
Where A = -25.942 B = 3.595 C = the threshold for cracking This equation is nearly identical with Peter Scott's original model uncorrected for cold work.
This work provided an independent verification of Scott's work. A further verification of the modified Scott model used here was provided by some operational crack growth rates collected by Hunt, et at [15).
The final verification of Peter Scott's model will come from actual data from head penetration rnaterials in service, as will be discussed in detail below. To date 15 heats have been tested in carefully controlled PWR environment. One heat dia not crack, and of the fourteen heats where cracking was observed, the growth ratos observed in twelve were bounded by the Scott model.
Two heats cracked at a faster growth rate, and the explanation for this behavior is being investigated.
A compilation was made of the laboratory data obtained to date in the Westinghouse laboratory tests at 325*C, and the results are in Figure'2-3. Notice that much of the data is far below the Scott model, and a few data points are above the model. These results represent 14 heats of head penetration materials.
The effect of temperature on crack growth rate was first studied by compiling all the available crack growth rate data, for both laboratory and field cracking of Alloy 600. This information is summarized in Figure 2-2, where the open symbols are for steam generator tube materials, and the solid symbols are for head penetration materials. The results are presented in a simple format, with crack growth plotted as a function of temperature. The effect of stress intensity factor variation has been ignored in this presentation, and this doubtless adds to the scatter in the data. The remarkable result is a consistent temperature effect over a temperature range from 288*C to 370*C, more than covering the temperature range of PWR plant operation although there is a wide scatter band in the figure. The work done originally in 1992 results in a calculated activation energy of 32.4 Kcal/ mole, which has been used to adjust the base crack growth law to account for different operating temperatures.
j l
Rev.0 2-2 July 1997 OM709.coc:10:07/14/97
F A series of crack growth tests is in progress under carefully controlled conditions to study the temperature effect for head penetration materials, and the results obtained to-date are shown in Figure 2-2. Sufficient results are available to report preliminary findings. The tests were performed with an applied stress intensity factor of 23 Ksi.E (25.3 MPa[m)"), periodic unload / reload parameters of a hold time of one hour and a water chemistry of 1200 ppm B + 2 ppm Li + 25 cc/kg H,. The results are consistent with the previous steam generator and head penetration material work. In the case of heat 69, the three results in the middle of the temperature range,309'C,327'C and 341*C have the same trend as the scatter band, almost exactly, while the high temperature and low temperature results are both lower than would be predicted by the activation energy, as shown in Figure 2-2. The results for heat 20 show a similar behavior, with the results at 325*C and 340*C also within the scatter band and nearly parallel to the heat 69 specimens, but at a lower crack growth rate, as shown in Figure 2-2.
The effects of several different water chemistries have been investigated in a closely controlled series of tests, on two different heats of archive material. Results showed that there is no measurable effect of Boron and Lithium on crack growth.
The key test of the laboratory crack growth data is its comparison to field data. Crack growth from actual head penetrations has been plotted on Figure 2-2 as solid points. The solid circles are from Swedish and French plants and the solid stars are from a US plant.
Figure 2-4 shows a summary of the inservice cracking experience in the head penetrations of French plants, prepared by Amzallag (16), compared with the Westinghouse laboratory data, corrected for temperature. This figure shows excellent agreement between lab and field data, further supporting the applicability of the lab data.
1 Therefore it can be seen that the laboratory data is well represented by the Scott model corrected for temperature using an activt: tion energy of 32.4 kcal/ mole. Also the laboratory l
results are consistent with the crack growth rates measured on actual installed penetrations.
l Therefore the use of the modified Scott model in the safety evaluations and other evaluations of I
head penetration integrity is still justifiable, in light of both laboratory and field data obtained to date.
l i
i l
I 4
l Rev.0 2-3 July 1997 c:0709.coc:1b:07/14/97 I
l
\\
\\
TABLE 2-1 TEMPERATURE CORRECTION FACTORS FOR CRACK GROWTH: ALLOY 600 j
i Temperature Correction Factor (CF)
Coefficient (Co) 330C 1.0 2.8 x 10
325 0.798 2.23 x 10
320 0.634 1.78 x 10
310 0.396 1.11 x 10
l l
300 0.243 7.14 x 10
290 0.147 4.12 x 10
b = Co (K - 9 )' m / s dt where K is in MPa[m)"
l t
l Rev.0 24 July 1997 c:0709.coc:t' C. H/97
i crack 48.drw 1E-09
.......................e.....u.......
... s..
330aC
.4...4......
. 4..
..... >...l.
..4...4..
E
>.... +...
..<....s.....4...
j E.10 s.
.. g..
.,.... >....,... o.... $..
..........4...4..
..a..
..a.......
6.
..,...>...3..
7 4
s 6
. 4..
Z
.....L.........t.....z.....a...
..e....6.....b.
..g....i...4....4..
s g
g.
.....b....)..
t O
......1..
- ....A.... a...-
...l...
W O
- ""?'.'
1E.11........
.2
...i....6.*..
,*=**t***.=.,*"
k.... >..
..g.....q.....(..
>... ).
... 6 s.
s.
....q..
4. ;...;..
6...;..
g f
,.4....4....q....,.....,....,............
.4.....<.....
......>....)...y... >..
. g... q.
.g 4
...4...
... e.
p..
..)........
..q..,......
..q.....(..
..s....s.,
..q..
.. a.... g.... a........e.........
.I 1E 12 O
20 40 60 80 100 K - MPa SQRT(m)
Figure 2-1 Scott Model for PWSCC of Alloy 600 at 330*C, as modified from Reference 11 Rev.0 25 July 1997 c:\\3709. doc:1b:07/14/97
TEMPERATURE, DEG. C 372
'n 333 3 15 2ss ze2 c*
ioco.o G.
c*
~
c0.0 g
d W
R w
v xn h
M N
N oA s
e 3
X*
N 1, N i
z e
g g
x xi i s
x N
w x
g u
NA iN N
O A
A \\_
d
\\
E Y
E m
m
<m o
\\
U E
ir o
~
m o
x uw r>
N M
N s
A 0.i n
O_i 0.00t55 0.00 6 0.00166 0.00170 0.00175 0.00180 RECIPROCAL TEMPERATURE,1/DEG. K w
% e C50A Fed Data CADM Las Tests A Heat 20 9 Heat 69 An etnes are S/G Tube lab 4 F*d Data Figure 2-2 Comparison of Temperature Effects Results with Other Laboratory and Field Data Rev.0 2-6 July 1997 0:0709.coc:10:07/14/97
crack 47.drw 1 E-09...
330*C 32.5 'C cf) p.
e..
E
... W r.
LU 1E 10.... -............ _ -
>==
.y....
......e................
4 g
Z
.,5.g.
.........eu..
+e-e i
,s.. g..
o o
o O
.......................u.........
E e
C i
o-W 1E 11
.......g........... '.
S..
g
]..............................
g g
4,...........
, r....
l 1E 12 O
20 40 60 80 100 K - MPa SQRT(m)
Figure 2-3 Summary of Available Westinghouse Laboratory Data for Alloy 600 Head Penetrations at 325'C Rev.0 27 July 1997 c:0709.coc:1b:07/14/97
S Comparison of Field & Laboratory Data 10 6
1 C'J :
ie
- _... ~
m e
iem *-
o e
i..
uI e, e a
-4 i
e.
i.e o e' E
00'8' S I % *W'e' l
.5 0.1
=
q-...-
- .==
n s
4 g
4 g
O i
0.01
- e FIELD oWestinghouse Lab *.___..
~
_t 0.001 0
10 20 30 40 50 60 70 80 Ki(Mpa m")
eigure 2-4 Comparison of French Field Data and Westinghouse Laboratory Data (W_ results reduced to 290*C using Q = 130 KJ/ mole) [16]
I 1
l Rev.0 2-8 July 1997 c:0709.coc:1b:07/14/97
t 3.0 WESTINGHOUSE CRACK INITIATION MODEL DEVELOPMENT AND CRACK INITIATION TESTING 3.1 CRACK INITIATION MODEL Westinghouse advanced an Alloy 600 PWSCC initiation model for primary components in l
Pressurized Water Reactors (10). Briefly, the model incorporates three contributing factors for l
the prediction of crack initiation time; namely, material condition, stress, and temperature.
l These are discussed below.
l Material Condition and Microstructure As reported by several authors [17,18,19,20, and 21), the Alloy 600 microstructure is a function of the thermomechanical history of the material heat as well as its carbon content.
Alloy 600 material heats subjected to mill annealing at low temperatures, i.e., 926*C or less, i
exhibit a fine grained microstructure with heavy transgranular carbide precipitation and little or no carbides precipitate on the grain boundaries. Such a microstructure is reported to be more susceptible to PWSCC. On the other hand, a high temperature mill-anneal (>1000'C) tends to put more carbon into solution, increases grain size, produces grain boundary chromium carbide precipitation and renders the material more resistant to resist PWSCC. Norring, et. al. [22), did not find a correlation between the total content of carbon and the crack initiation time, but they observed good correlation between the amount of grain boundary carbides and crack initiation time. The fact tnat grain boundary precipitation is beneficial to PWSCC has been reported by many researchers [23). Norring, et. al., [22), showed that the crack initiation time varied directly (linearty) with grain boundary carbidos. Their data suggested that when the grain boundary carbide coverage is increased by a factor of 3, the crack initiation time also increased by a similar factor (from 4,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> to 12,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />). Bandy and Van Rooyen [24), pointed out that i
in addition to grain boundary carbide coverage, other features relating to processing history l
variables such as carbon concentration gradients, substructural features, grain size distribution, cold work, intragranular carbide distribution and the grain boundary segregates all play an important role in the cracking behavior of the Alloy 600 material.
When considering the influence of microstructure on the PWSCC susceptibility for the purpose of the current evaluation, to enable comparison d heats fabricated at different vendor shops, the thermomechanical processing history effect is separated from the grain boundary carbide coverage effects. In general, the influence of the grain boundary carbides is known and the coverage (G) can be easily measured directly from the microstructure. The influence of other structural features due to processing history cannot be assessed directly. These processing effects are represented in the torrent treatment by a single parameter (A) characteristic of the i
fabrication shop (vendor). This approach provides a means of comparing the PWSCC l
susceptibilities of Alloy 600 material heats from different vendor shops although they may
{
contain similar grain boundary carbide contents.
i Rev.0 3-1 July 1997 c:\\3709. doc:1b:07/14/97
i l
l l
Influence of Stress i
Steady state tensile stress in the component, either due to residual and/or applied loads, has a strong influence on the PWSCC.
Bandy and Van Rooyen [24), reported that the time to failure varied inversely as the fourth power of applied stress in both annealed and coldworked specimens. They also reported data to support that coldwork reduces the resistance to PWSCC. The effective stress at a given Alloy 600 location is a function of the fabrication steps and their sequence, the yield stress of the material, and the service stress. In general, the local residual stresses resulting from fabrication can play a more significant role than the service stresses themselves.
Temperature Effects Several investigators [17,24), examined the role of temperature on PWSCC. It is well established from these results that the crack initiation time decreases exponentially with temperature and that they are related through an Arrhenius equation expressed as a function of the activation energy of the process. The experimental results confirm that Alloy 600 PWSCC is a thermally activated process and the activation energy for the process varies approximately between 50 to 55 kcal per mole. An activation energy value of 55 kcal/ mole is consistently applied throughout the current assessments, for crack initiation. A different value,32.4 applies for crack growth as was discussed in Section 2.
3.2 THE WESTINGHOUSE CRACK INmATION MODEL l
l Consistent with the contributing factors discussed above, the crack initiation time (t,) or the rate of crack initiation (1/t,) is proportional:
l l
1/t, a (Stress)"
a e*'"
a inverse of the grain boundary ' arbide coverage factor, (1/G) c c" e
- so that 1/t, et G
Since the nature of the vendor thermomechanical process 5g is also a significant contributing factor, one can say that for a given fabrication process a
omt 8 8 1/ t, = A (3-1)
G i-Rev.0 3-2 July 1997 o-0709.coc:1b:07/14/97
1 The proportionality constant "A" can be chosen to represent the procusing conditions j
representative of a given manufacturing process or manufacturer, and could include parameters such as yield strength as part of the expression.
i "A" can be assessed for a given heat by substituting the parameters of a se:vice component with a known cracking history for the heat of material. "A" will then represent the processing condition (or the vendor) by the definition we have just established.
l The parameters in the above rate equation (3-1) are described below:
i A
is a constant, relating to the processing, and fabrication conditions of the material G
is the grain boundary carbide coverage factor o
is the effective tensile stress (resulting from applied and residual stresses) n is the stress exponent having a value ranging from 3.5 to 4.5 for Alloy 600 in primary water O
is the activation energy for the crack initiation process and has an approximate value of 55 kcal/ mole R
is the gas constant (1.987 cal / mole degrees K) l T
is the absolute temperature in degrees K, and t,
is the time to initiate cracking.
i 3.3 CRACK INITIATION TESTING Westinghouse currently has an ongoing autoclave test program to establish the PWSCC crack initiation behavior of archive Alloy 600 RV head penetration material heats from a variety of fabricators representative of microstructures of'RV head penetrations that are currently in service. The objectives of the Program are:
To determine the effect of penetration microstructure and material type (vendor) on the relative susceptibility to cracking.
To define a (natorial index (A) to assist in plant maintenance planning.
The program is sponsored by EPRI and the CE, E, and B&W owners groups. The accelerated testing is conducted under dense steam with hydrogen at 400*C and utilizes full size ring samples fabricated from RV head penetration tubing from different vendor shops. A listing of l
vendor shops representing the ring samples employed in the testing is provided in Table 3-1.
l Rev.0 3-3 July 1997 o:\\3709.coc:1b:07/14/97
l i
To provide reference benchmarking, samples from steam generator rolled transition tubing and Alloy 690 penetration material are also included in the test matrix. Penetration material specimens with known crack growth behavior measurements from previous test programs are included for comparison with other data.
l I
This environment has been shown to provide adequate acceleration (up to 500x) to provide results within the test period. This will be verified using the specimens from heats that have been tested previously. Test samples under the doped steam test will be inspected at 25,50, l
100,200,400, 800,1400 and 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br />. Inspection will include visual, metallographic and destructive examinations.
l L
The ID surfaces of the ring samples are strained by controlled cyclic ovalization to simulate the j-residual hoop stress,es in the plant. The stresses are quantified based on the ovalization. The i
final cycle of ovalization is calibrated to induce a 2mm difference in measured inside diameter.
This corresponds to the upper 95% of the measured ovality in the outermost penetrations in service. The cyclic straining procedure of the full ring samples is illustrated by the loading curve shown in Figure 3-1.
l The testing is, conducted under two phases. The first phase involves a cumulative exposure of I
up to 800 hours0.00926 days <br />0.222 hours <br />0.00132 weeks <br />3.044e-4 months <br /> in six exposure intervals. Periodic inspections are performed at 25,50,100, 200,400 and COO cumulative hours of exposure. The second phase testing involves the exposure of specimens for a cumulative exposure of up to 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> with an interim inspection at 1400 hours0.0162 days <br />0.389 hours <br />0.00231 weeks <br />5.327e-4 months <br />. Currently, with the Phase I testing completed, the preliminary test results indicate clest trends in the initiation behavior. Out of the six heats of material tested.
l two of the heats consistently showed higher susceptibility to cracking; the worst heat being the heat that also showed the highest crack growth rate under the crack growth test program discussed in Section 2. Further useful trends in cracking behavior are expected at the end of the 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> exposure. The overall results of the program are expected to provide useful information for plant maintenance planning.
I i
3 h
j Rev.0 34 July 1997 j
c:U709.co::1tx07/14/97
'l
.l
. _ _ = _ _
t TABLE 3-1 MATERIAL HEATS EMPLOYED IN THE ALLOY 600 RYHP CRACK INITIATION TESTS S No.
Heat No.
Supplier Fabricator As Prod. Size 1
93510 B&W B&W
%" (6 pcs) 2 93510-R B&W B&W
%"(6 pes) 3 91069 B&W B&W
%* (6 pes) 4 93511 B&W B&W
%" (6 pes) 5 WF675 B&W Creusot Loire 3-5/8"(1 pc) 6 WF151 Sizewell Creusot Loire 3-%"(1 pc) 7 M-7817-1 (EO-CE Standard Steel 4-1/8"(1 pc) 6943#2) 8 R13-4 (NX64209)
CE Huntington 4-1/8'(1 pc) 9 NX810175 Huntington 6* (1 pc) 10 NX34C3-68 Huntington 6* (1 pc) 11 R177 Vattentall Sanvik 6"(1 pc) l t
i f
Rev.0 3-5 July 1997 c:U709. doc:10:c7/14/97
i Active and Residual Strains during Residual Stress introduction 40000 fn_
--a,--
---.---p.---H I
g 30000 f.
~~,
f l
~
.E 20000
---/---g
,x-
.t t
l
\\
/
g' jQQQQ
. 3
..,.7...,..
o s
f E
0l
-t-
\\.-
l l
h i
- =,
s
./
T g
W#~,g---
10000
%g, -
J i
-20000 t
0 0.25 0.25 0.75 0.75 1.25 1.25 1.75 1.75 relax relax relax relax Loading Cycle f
t
- Load (pounds) --s- -Strain (ue) 1290/6:00 t.
Strain (ue) 3:00/9:00
^
i i
Figure 3-1 Residual Stresses and Strains induced During Controlled Cyclic Ovalization of RV Penetration Ring Samples f
f i
[
Rev.0 3-6 July 1997
[
t o11709. doc:1b:07/14/97 f
t
l
4.0 REFERENCES
[1]
WCAP-13565, Rev.1, " Alloy 600 Reactor Vessel Adapter Tube Cracking Safety Evaluation," February 1993 (Proprietary).
[2]
F. Hedin and P. Gasquet, " Alloy 600 Reactor Vessel Head Penetration Cracking: An industrial Challenge," 12* SMIRT Post Conference, August 23-25,1993, Paris, France.
[3]
Rao, G. V., and Wright, D. A., " Evaluation and Resolution of the Primary Water Stress Corrosion Cracking (PWSCC) incidents of Alloy 600 Primary System Pressure Boundary Penetrations in Pressurized Water Reactors," Proceedings of Fontevraud 11 Symposium on " Contribution of Materials investigation to the Resolution of Problems Encountered in PWR Plants," Royal Abbey of Fontevraud, France, September 10-14,1990.
[4]
A. S. O'Neill and J. F. Hall, Combustion Engineering, " Literature Survey of Cracking of Alloy 600 Components in PWR Plants," Report prepared for EPRI, January 1990.
[5]
U.S. NRC Information Notice No. 90-10 " Primary Water Stress Corrosion Cracking (PWSCC) of inconel 600, February 23,1990.
[6]
NRC letter from William T. Russell to William Rasin of NUMARC (now NEI),
November 19,1993.
[7]
Pichon, C, Boudot, R., Bonhamour, C., and Golpi, A., " Residual Life Assessment of French PWR Vessel Head Penetrations through Metallurgical Analysis." Service Exoerience and Reliability Imorovement: Nuclear. Fossil. and Petrochemical Plants. PVP Vol. 288, ASME,1994.
[8]
1.agerstrom, J., Wilson, B., Persson, B., Bamford, W.H., and Bevilacqua, B, " Experiences with Detection and Disposition of Indications in Head Penetrations of Swedish Plants,
" Services Exoerience and Reliability Imorovement: Nuclear. Fossil. and Petrochemical BAD 11, PVP Vol.288, ASME,1994, pages 29 to 40.
[9]
Bamford, W. H., Fyfitch, S., Cyboron, R. D., Ammirato, F., Schreim, M., and Pathania,
(
R., "An integrated Industry Approach to the lasue of Head Penetration Cracking for the USA,: Services Experience and Reliability Imorovement: Nuclear. Fossil. and Petrochemical Plants. PVP-Vol 288, ASME,1994, pages 11 to 19.
[10]
Rao, G. V., " Methodologies to Assess PWSCC Susceptibility of Primary Alloy 600 Components in PWRs," Proceedings, Sixth Intemational Conference on Environmental Degradation of Materials in Nuclear Power Systems, NACE, August 1993.
[11]
Scott, P. M., "An Analysis of Primary Water Stress Corrosion Cra:: king in PWR Steam l
Generators," in Proceedings, Specialists Meeting on Operating Experience With Steam j
l Generators, Brussels Belgium, September 1991, pages 5, 6.
3
?
Rev.0 4-1 July 1997 eM709. doc:1b:07M4/97
[12)
Mc liree, A. R., Rebak, R. B., Smialowska, S., ' Relationship of Stress Intensity to Crack Growth Rate of Alloy 600 in Primary Water," Proceedings Intemational Symposium Fontevraud 11, Volume 1, p. 258-267, September 10-14,1990.
[13]
Cassagne, T., Golpi, A.," Measurements of Crack Propagation Rates on Alloy 600 Tubes in PWR Primary Water," in Proceeding of the 5th Intemational Symposium on Environmental Degradation of Materials in Nuclear Power Systems-Water Reactors,"
August 25-29,1991, Monterey, Califomia.
l
[14)
Personal Communication, Brian Woodman, Combustion Engineering, October 1993.
[15)
Hunt, S. L and Gorman, J., " Crack Predictions and Acceptance Criteria for Alloy 600 Head Penetrations" in Proceeoings of the 1992 EPRI Workshop on PWSCC of Alloy 600 j
in PWRs, December 1-3,1992, Orlando Fl (published in 1993).
[16)
Personal communication - C. Amzallag to W. Bamford, Feb. 26,1997.
l
[17)
G. Economy, F. W. Pement, Corrosion /89 Paper 493.
l
[18]
H. Tass et. Al., " Relation Between Microstructural Features and Tube Cracking Observed l
on Tube Samples of Doel 2 Steam Generator." EPRI Steam Generator Owners' Group, SCC Contractors Workshop San Diego, CA, March 1985.
[19)
A. A. Stein, " Development of Microstructural Correlation and a Tubing Specification for l
Alloy 600." Paper presented at EPRI Steam Generator Owner Group SCC Contractors Workshop, San Diego, CA, March 1985.
[20)
A. R. McIlree, "Results of Reannealing Studies of Trojan, Doel 2, Ringhals 2 and 3, Ginna and Indian Point 3 Steam Generator Tubing" Paper as in Ref.18.
[21)
G. P. Airey, " Optimization of Metallurgical Variables to improve Corrosion Resistance of Inconel Alloy 600," Palo Alto, CA, Electric Power Research institute, EPRI NP 3051, July 1983.
l
[22]
"intergranular Stress Corrosion Cracking in Steam Generator Tubing, Testing of Alloy 690 and Alloy 600 Tubes," Norring, Engstrom and Norberg, in Third intemational Symposium on Environmental Degradation of Materials in Nuclear Power Systems -
Water Reactors - Proceedings, The Metallurgical Society,1988
[23)
Z. Szklarska Smialowska," Factors influencing IGSCC of Alloy 600 in Primary and l
Secondary Waters of PWR Steam Generators" Proceedings of the 5*ntemational Symposium on " Environmental Degradation of Materials in Nuclear Power Systems '
i Water Reactors." Nace Meeting. Edited by D. Cubicciotti, August 1989, p. 6-1.
[24]
R. Bandy and D. Van Rooyen," Stress Corrosion Cracking of inconel Alloy 600 in High Temperature Water - An Update" Corrosion, Vol. 40, No. 8, page 425 (1984).
l I
i l
Rev.0 4-2 July 1997 l
OATr09.coc:1b:07HN97 I
[25)
Letter, J. A. Begley (APTECH) to B. A. Bishop, " Review of the Westinghouse Structural Reliability, Model for PWSCC of RV Head Penetrations," June 23,1997.
[26)
" Evaluation of Leaking Alloy 600 Nozzles and Remaining Life Prediction for Similar Nozzles in PWR Primary System Application," Hall, Magee, Woodman and Melton,in Service Experience and Reliabilitylmprovement, ASME PVP-Vol. 288,1994 l
l
[27]
"The Status of Laboratory Evaluations in 400*C Steam of the Stress Corrosion of Alloy 600 Steam Generator Tubing," Gold, Fictcher and Jacko in Proceedings of 2nd Intemational Topical Meeting on Nuclear Power Plant Thermal Hydraulics and Operations,1986
[28]
WCAP-13525, Rev.1, RV Closure Head Penetration Alloy 600 PWSCC (Phase 2), Ball et al., December 1992 (Class 2)
[29)
WCAP-13493, Reactor Vessel Closure Head Penetration Key Parameters Comparison, Duran, Kim and Pezze, September 1992 (Class 2) l l
l
[30]
G. V. Rao, " Development of Surface Replication Technology for the field assessment of Alloy 600 micro-structures in Primary Loop Penetrations," WCAP-13746, Westinghouse Class 2 report June,1993.
[31)
G. V. Rao and T. R. Leax, Microstructural Correlations with Material Certification Data in Several Commercial Heats of Alloy 600 Rcactor Vessel Head Penetration Materials -
WCAP-13876, Rev.1,1997.
[32)
WCAP 13929, Rev. 2, Crack Growth and Mocrastructural Characterization of Alioy 600 Head Penetration Materials, Bamford, Foster and Rao, November 1996 (Class 2C)
[33]
Newman, J.C. Jr. And Raju, l.S. " Stress intensity Factors for intemal Surface Cracks in Cylindrical Pressure Vessels" Transactions ASME. Joumal of Pressure Vessel Technology, Volume 102,1980, pp. 342-346.
[34)
WCAP-14572, Westinghouse Owners Group Application of Risk-Based Methods to Piping inservice inspection Topical Report, pp E-1 to E-6, March 1906 (Clacs 3).
[35)
Risk-Based Inspection - Develcpment of Guidelines, Volume 1, General Document, ASME Research Task Force on Risk-Based inspection Guidelines Report CRTD-Vol. 20-1 (or NUREG/GR-005, Vol.1), American Society of Mechanical Engineers,1991
[36)
NUREGlCR-5864, Theoreticaland User's Manualforpc-PRAISE, A Probabilistic Fracture Mechanics Computer Code for Piping Reliability Analysis, Harris and Dedhia, July 1992 i
l I
i Rev.0 4-3 July 1997 CCU700. doc:1b:07M4/97
.