ML20028F871
| ML20028F871 | |
| Person / Time | |
|---|---|
| Site: | Susquehanna  |
| Issue date: | 08/05/1981 |
| From: | Halapatz J Office of Nuclear Reactor Regulation |
| To: | Vollmer R Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20027A667 | List: |
| References | |
| NUDOCS 8302040546 | |
| Download: ML20028F871 (56) | |
Text
- -w.:=
O l
C ',
NUGG..' i.
- J :':,L u w.
939.,95.,.O ~*'
s'.w August 5,1981 MEMORANDUM FOR:
R. H. Vollmer, Director Division of Engineering FROM:
J. Halapatz Materials Engineering Branch SUDJECT:
DIFFERING PROFESSIONAL OPINION RELATED GENERICALLY TO SENSITIZATION OF BWR STAINLESS STEEL WELDMENTS 0
REFERENCE:
Memorandum, W. S. Hazelton to R. H. Vollmer, dated July 41, 1981, subject, " Differing Professional Opinion Related to~
Sensitization of BWR Stainless Steel Weldments"
,5 This merorandum is identified as a statement for record in accordance with NF.C Ap;:endix 4125. This memorandum addresses the reference memorandum:
Mr. Hazelton:
"Although most of the components he mentions are generally included in reactor ir.ternals, he also mentions installed large diameter pipe and concrete embedded flued heads.
I don't know which systems he is referring to, however.
My reviewers and I have some questions and concerns regarding how this ' issue' will be resolved."
Halapatz responds:
The lar;e diameter pipe and concrete embedded flued heads mentioned are those non:onforming materials for which licensees will claim undue hardship related to replacement thereof within the context of NUREG-0313. Rev.1 The Pennsylvania Power and Light Program for mitigation of IGSCC at Susquehanna, fcr ex. ple, see attachment 1, claims undue hardship related to such components.
Ti.e lar;s pipe is in the recirc and RHR systems. The flued heads are'in the reacto-water cleanup and core spray syst;.n. Given that these caterials are nonconforming, but will not be replaced, and with an undetermined accessibility for inservice inspection, it follows that netallurgical char:cteritation of these materials with respect to propensity to IGSCC is in order.
L31_
Mr. Hazelton:
"1.
First, what is the issue? We'vebeengoingaroundincirclesInthisone, but if it is to be resolved by a trip to San Jose, it appears to me that we'd better make sure that we know what it is that needs to be resolved, If the question is whether some welds are, or could be sensitized, a.
we can stipulate right now to that effect.
s b.
If the question is whether all GE supplied equipment met or does ceet Regulatory 1.44, we know tha t the answer is negative.
8302040546 820720 PDR ADOCK 05000387 P
\\
--::.w O
3
^
r' J
d
,.-- - - +_c
%. :! P * #
R. H. Vollmer
~
~
c.
If the question is whether components not made to 1.44, an::/or are sensitized, are likely to fail in service, looking over welding procedures will not resolve the issue."
Halapatz responds:
1.
The issue is that we really haven't metallurgically characteri:ed the materials of interest with respect to their propensity to IGS:.
Halapatz assures Mr. Hazelton that HaTapatz hasn't been going around in circles on this one. Halapatz has been to San Jose before; he takes no great delight in traveling on business to San Jose or elsewhere, should that be bothering Mr. Hazelton, but he thinks we simply must tne a a
good hard look to determine what materials are in place in i;TO' S and 2 what their propensity is to IGSCC based on puidelines and criteria
~
the NRC applies in the licensing process, i.e., we should anticipate rather than commiserate failures. We should then implement a well-defined surveillance program for these components identified as 1GSCC "
- prone, r..e sa.:~ c. D~, s. -: -
..,.w e *
- a.
It's the severity of sensitizationathat higher heat in;u: cevelops more severe sensitization and a greater propensity to IZ::.
This position is endorsed by R.G.1.44, attachment 2, via the fcilowing:
"All welding processes will result in some carbide preci;i r tion in the weld metal and in base metal heat-affected zone :f :::v.less steel welds, but significant sensitization does not nc-: '
result when typical welding procedures and material chemistry t :. sed and when no further heating of material occurs. However, t% re is evidence that atypical welding methods using very high r.e.1 input could result in stress corrosion cracking in the heat-rif::.ed zone of the weld. To avoid this, the welding procedures t a material chemistry (if necessary) should be controlied tc trcvent undue sensitization of the heat-affected zones of the welcments.-
Controls to prevent sensi:ization of the material durir.. 0 cint nr. include:
(1) avoidirr,alding practices that res.it -
- 3 ger.2 ration of high heat, s'., r.aintaining low r. eat in:n.
- r. trol-r lir; current, voltage, uni tr; vel speed, (3) limitir.; i-erpass temperature, (4) using stringer bead techniques and av icin excessive weaving, and (5) limiting the carbon level cf :d material where section thickness makes the material more prone to ss'nsitiza-tion.
.'l *.
2 "In addition, welding procedures should be qualified ty pasjdn'g3 suitable intergranular corrosion test in all cases where the procedure is used for welding stainless steel having a carcon t
l level greater than 0.03 percent.
The qualification test should be performed using base material having the maximum carbcr. content anticipated and the minimum and maximum thicknesses ar.:icipated.
4
l'.
.razw G
7
,s o.
)
A
. ~.a.f
- f...... :.. i m e r C
"As a minimum, the variabies tnat should be controllte in the qualification test are neat input, interpass temperature, and weld-ing techniques for specific section thicknesses. "
"6.
Welding practices and, if necessary, material composition should be controlled to avoid excessive sensitization of base metal heat-affected zone: ef weldments.
An intergranular corro-sion test, such as specified in subparagraph C.3. above, should be performed for each welding procedure to be used for welding material having a carbor content of great:r than 0,03 percent."
This position also is endorsed b," the current GE position on compliance with R.G.1.44 as developed within GESSAR Docket No. STN-50-447. As shown in '_
attachment 3, Shoreham, for exan le, committed for NSSS to the GE positioW#
in GESSAR, Docket STN-50-447, whico limited welding heat input to 110,000 joules /in. GE, however, later reconsidered, in 1980, its position as shown by the GESSAR II Docket 50-447 and limite,d welding heat input to 50,000 joules /in., in essential compliance with R.G.1.44.
The differing professional opir.ior, expressed is consistent with R.C.1.44 and the current GE position.
To the extent that it can be established, however, welding was performed on Shoreham components to the 110,000 joule /in limit.
On this basis, suspect welds having a greater propensity to IGSCC should exist. NUF.EG-0313, Rev. 1, invoked by the NRC, however, addresses only piping, but not reactor internals or structurals nor nonconformir.; raterials as cited for Susquehanna.
b.
Mr. Hazel ton's state:Ent that GE equipment is in place which does not meet R.G.1.44 is a major contradiction and serious challer;e to the credibility of MTEB SERr uhich address the matter.
The Shr.rekam SER, Docket 50-322, dated April 1931, attachment 4, for example, acknow-ledges that R.G.1.44 has been met, despite the farcical use of the expression " satisfy the ir.tcn: of R.G. 1.44."
The LaSalle FSAR con.i-tr: ::. the 110,000 joules /in. liri; indi-cates that the issue is gcneric (attachment 5).
Halapatz, R.G.1.44, and currently GE, share a common posijion; to c.
wit, that weld heat input should be controlled to lessen the severity of sensitization and hence propensity to IGSCC. "Looking over welding procedures" will contribute much information for'the metallurgical characterization of the materials of interest and c-their propensity to IGSCC. The review would simply assess. aterials within R.G. 1,44 guidelines.
\\
f
-GE%
i I
.m
-g
)
. )
.aw l ?,
c, n,7,,,
7 :.:f:- --
-A.
M:. Hazel ton:
"My reviewers and I believe tnat the lack of problems with IGSCC failures of BWR internals, together with the results of laboratory tests showing _ the rMcessity for maintaining stresses at or above yield to cause IGSCC in BWR environment, Justify our position that complete conformance to Regulatory Guide 1.44 is not required for us to make a finding that the applicable GDC's are met."
s Hzlapatz responds:
Halapatz is astounded by Mr. Hazelton's statement that there is a lack of,
problems with IGSCC of BWR internals. Mr. Hazelton and his reviewers are f.
directed to attachment 6, which is a compilation of nuclear power plant-
?
experience to which the NRC library suscribes. Mr. Hazelton will note 3
that the compilation describes IGSCC failures of reactor internals components,.
which have become commonplace. Based on Mr. Hazelton's premise then, given -
the number of failures which have occurred, stress levels at or above yield would appear to be the rule rather than exception, in reactor internals and structural s.
Mr. Hazelton's argumer.t relating to at or above yield stress levels seems to be specicus.
Based on FSAR information, NT0Ls sucn as Shoreham, reactor internals and structural materials will be the same, be welded using the same welding ccr,trols and enter service in the same metallurgical condition, experience the same service environment and endure the same stress levels in service, whatever they may be3as the reactor internals and structural materials applied in pre-NTOL plants. Therefore, they can be expected to crack.
Halapatz finds'it incredible that NT0Ls are 'being. licensed with full know-ledge that failures are likely to occur. The basic fact underlying his DP0 is that these potential failures should be identified and kept under surveillance.
"3.
Inspection programs on BUR internals are being carried out.
These include ~~
spray lines, feedwater sparcers, LPCI connectors, jet pump components, shroud head assembly (stand oi:as and steam separators), and thc eteam 4
dryer assembly. This program has discovered cracks in core spray lines and spargers, for example.
(Note that not all cracks were in peid HAZs.)
In addition to these components that are inspected fairly often, the CRD i
guide tubes and instrumentation piping are inspected every 10 years, 1
when it can be made accessible. We believe that this inspectioa program l
1s adequate, and is about the best that can be done."
- r
- .s*
1 i
Halapatz responds:
+-
In all candor, Mr. Hazelton's first statement, more properly should read,
" Inspections on some BWR internals are now being carried out.". Item 55 of l
1 u
,ory-g
--,--w e-e-p--
w--m-er
..n
)
)
(
. Vollmer 2~
s a'ttachment 6 is illustrative of how the problem is being handlet.
It would appear more factual to state that cracks discovered in core spra,e lines, spargers, etc., are forcing us to inspect some components. The fact that not all cracks were in weld HAZs leads one to conclude only that severely sensitized base material was applied improperly.
Given this failure history and given that the metallurgical histories and environmental and stress level experience of the materials applied in NTOLs will be the same as those of pre-NTOL materials which have failed, it is most baffling to read, as in the Shoreham'SER dated April 1981:
"4.5.2 Reactor Internal and Core Support Materials a
The applicant has met the requirements of General Design Criterion 1 and g Section 50.55a of 10 CFR Part 50 by assuring that the design, fabrication, and testing of the materials used in the reactor internals and core support structure are of high quality standards and adequate for' structural integrity.-
The controls imposed upon components constructed of austenitic stainless steel satisfy the intent of the recomendations of Regulatory Guide 1.31, ' Control of Ferrite Content in Stainless Steel Weld Metal' and Regulatory Guide 1.44,
' Control of the Use of Sensitized Stainless Steel.'
"The materials used for the construction of componer;;.s of the reactor internals and core support structures have been identified by specification and found to be in conformance with the requirements of NUREG-2000 of Section III and Parts A, B, and C of Section II of the ASME Code. As proven by extensive tests and satisfactory performance, the specified materials are compatible with the expected environment and corrosion is expected to be negligible.
"The controls imposed on the reactor coolant chemistry provide reasonable assurance that the reactor interr.als and core support structure will be adsquately protected during operation from conditions which coule lead to stress corrosion of the materials and loss of component integrity.
.."The caterial selection, fabrication practices, examination and testing procedures, and control practice performed in accordance to these recom2 mendations provide reasonable assurance that the materials usec for tfie l
reactor internals and core support structure will be in a metallurgical l
condition to preclude service deterioration. Conformance with tr.e, require-I ments of the ASME Code and the intent of the recommendations of the I
regulatory guides constitutes and acceptable basis for meeting the require-
~
ments of General Design Criterion 1 and Section 50.55a of 10 CFR Part. 50.,"
th epar? cc,armu G5 :
t l
It is~ noted that Section XI coverage of BWR reactor internals is in the? course of preparation.Section XI acceptance standards for BWR core support struc-tures also are in course of development. NRC guidelines relatinc to PSI and e
a
e ra -
o
')
s)
-s
..u R. H. Vollmer
/- @ '
ISI of reactor internals are,at this time, no better definec t..an the inspec-tion program identified by Mr. Hazelton. Halapatz declines to juoge as i
adequate an inspection program based on the accessibility of surfaces for inspection. On this basis, sensitive surfaces,which should ce inspected, may never be inspected. Attention is called, in this regard, to item 12 of attachment 6, which addresses cracks in the core support guid of Dresden 1.
Halapatz further declines to accept such a program, as described by Mr. Hazelton 4
as "about the best that can be done." There is much more that can be done!
It
- l follows, more logically, that the identifi4:ation of IGSCC prone reactor internals, core support structures and components nonconforming to NUREC-0313, Rev.1 should be inspected on the basis of mandatory accessibility of surfaces as prescribed inspection intervals core restrictive than those prescribed by i
Section XI.
t#
Mr. Hazelton:
g, 4
"4.
Our reviewers feel strongly that they should be included in any NRC investi-l gation into the validity of tneir reviews.
It is clear tnat we cannot afford to send a contingent of six or mort people to San Jose for this purpose, therefore, we urge that General Electric be given the oppor-tunity to come here to make a formal presentation on the subject.
I would be pleased to assist in preparing a proposed agenda, which I believe should cover much more than details of weld procedures."
Halapatz responds:
Halapatz feels even more strongly about the issue and is very strongly dedicated as his duty, in accordance with NRC Manual Chapter 4125-02, to the expression of his differing professional opinion.
At the risk of appearing irreverent, given the apparent sensitivity which has developed, Halapatz suggests that the NRC would be best served were his differing professional opinion be considered outside the jurisdiction of.
Mr. Hazelton and his reviewers and instead be submitted to an impartial peer m
revieu group for review, evaluatien and recommendations.
Hala st: is aware that the NRC includes in its staff > individuals, not here ofo*E identified in this matter, for whom he has professional respect for tneir knculedge7of the technical aspects of the issues, in addition to their professional, objectivity and integrity.
g p,,
h
- .. c '.
r
.1-g c s%*
q,,
a l
l
[ %
0 Hdla atz
'{ }
- ?
(
.atei s Engineering B'r.ach
)
I cc:
W. V. Johnston, DE S. S. Pawlicki, DE G. Johnson, DE
- u. Hazel ton, DE i
J. Thomas, NTEU
't M. Levy,liTEU e
\\
(QTD 1." ' #\\% %. Y
.c_
\\
~~
m PPaL 3
IW O NOf1H NtNIH $1R!ti, ALLENTO WN, P A.18101 PHONt (2156770 5131
...w--
NORM AN W. CURTi$
vece Presidene.Engineenng & Construct sn.Nucteet nosni June 30, 1981 f
Mr. A. Schwencer, Chief 3
Licensing Branch No. 2 Division of Licensing U.S. Nuclear Regulatory Commission Washington, D.C.
20555 SUSQUEHANNA STEAM ELECTRIC STATION RESPONSE 'IO URC GENERIC LETTER 81-03 AND NUREG 0313 " TECHNICAL REPORT ON MATERIAL EELECTION AN'D PROCESSING GUIDELINES FOR BWR COOLANT PRESSURE BOUNDARY PIPING" ER 100450 FILE 841-9,883 841-13 PLA-856 Attached please find PP&L's response to Generic letter 81-03 and NUREG 0313.
Very truly yours,
-]
(/)
/.
'l l,J. Jf-c.A~
^
g.
~
N. W. Curtis Vice President-Engineering & Construction-Nuclear c-r TEG/mks 1
Attachment h,
f.?
s.
I l
1
, P E N N S Y L V A NI A P O W E R & LIG H T COMPANY 9
h9 A s% _ m #*
M"'%
i d 'W 7 ::q:
a rut trrz..':::
- ::.~:
t SUSM*: HA::::A STE7J: ELrcP.Ic :*rJ.72 c::
Since 1975, when knowledge of intergranular stress corrosion cracking (IGSCC) in austenitic stainl'ess steel in boiling water reactor (BWR) piping was brought to its attention, Pennsylvania has undertaken an extensive program, Power & Light Company (PP&L) for the Susquehanna Steam Electric Station to effectively eliminate the possibility of IGSCC-caused downtime, and also to conform to the requirements of NUREG-0313, Rev. 1, and previous publications.
PP&L's program of IGSCC mitigation includes the following
' methods :'
Replacement of susceptible materials,where practical, with 1.
materials that conform to Section III of NUREG-0313.
O Recirculation system discharge valve bypass line - (4"f!-
A.
replaced with carbon-limited Type 304 SS (low carbon stainlessc,,
steel that meets the mechanical properties of regular grade Type.
304, and has a maximum carbon content of.03%).
f-Core Spray and Head Spray-replaced with carbon limited B.
Type 304.
C.
Reactor water clean-up system (RWCU) - replaced with Type 304L SS.
D.
Instrument piping and bottom drain - replaced with Type 304L SS.
Any further replacement of nonconforming material would result in what PP&L feels is an undue hardship, because it would involve replacement of already-installed large diameter (20" or larger) piping or flued heads that are imbedded in concrete.
.In view of the above reasons, and with consideration dor'.the_
very extensive effort by PP&L that resulted with the replacement -
of a considerable amount of nonconforming material, PP&L requests that further material replacement requirements not be imp'bsed (Generic Letter 81-03) because of severe undue hardship.,i 2.
Elimination of lines whose functions are no longer required.
\\ ;_
A.
Control Rod Drive Hydraulic Return.
r:P 3.
Use of low carbon corrosion resistant weld build-up (shop method) for field welds.
A.
Recirculation system risers.
)
e
.m, _ _ _ _... _ _,,
cr<vic:n.
t'. f Desigr. improvc.=ent]o cli::. int.t e f
A.
Rocirculation system inlet cafe ends (extensive randif ca-t tion to replace all safe ends'and'tEckmal sleeves).
5.
Solution heat treatment.
A.
Recirculation system risers (shop welds).
In addition to the mitigation methods listed above, PP&L has used several more, some of which are listed under Section V of NUREG-0313, as follows:
6.
Dissolved oxygen control During normal plant operation - PP&L has relocated the A.
control rod drive (CRD) pump intake from the condensate storage tank to the condensate makeup / reject line.
This results in using CRD water with the lowest oxygen concentration available (ess9n-l tially water of feedwater quality).
j l
B. :During all phases of oper'a' tion / shutdown except normal ?..
operation, oxygen levels are reduced by the use of a mechanica,1 The deaerator is expected to maintain an oxygen vacuum deaerator.
content of less than 250 PPB during start-up, hot standby and shut-down.
7.
Weldino Parameters A.
Block welding was prohibited.
Interpass temperature was limited to 350*F maximum.
B.
No preheat was used (in excess of a working range of C.
60*F to 150*F).
8.
Territe Control All of the weld metal and all of the type 304/316 castings This were.specified to have not less than 5% ferrite content.
level is generally recognized as sufficient to provide immunity from the initiation of IGSCC.
a e
,l Use of induction heating stress improvement (IHSI)
IHSI is a method of eliminating or substantially reducing i
residual tensile stresses on the inside surface of the pipin?g. These tensile stresses are primarily caused by welding and haveebeen PFEL is identified as one of the three major causes of IGSCC.
1 unit presently considering the use of IHSI for 108 welds on No.
that do not conform to Section III of NUREG-0313.
These welds are among a total of 110 welds of 4" diameter and larger that are j
(ISI) presently scheduled to receive augmented inservice inspection in 'accordance with Section IV.B.l.b of NUREG-0313.
I j
e
. ")
,]
e, The 2 0 welds are'ec ft:...:
, 4" 1.
Recirculation sistch~2I~scharge valve hyt rr 11n6 welds (HAZ, s on weldolots).
2.
Recirc. risers 12" welds (HAZ's on cwecpolets/
reducers).
3.
Core spray 12" welds (HAZ's on 316 SS flued heads).
4.
RWCU 4" welds (HAZ's on weldolets/ flued heads).
5.
Residual Heat Removal (RHR) system - 41 welds (20", 24").
6.
Balance of recirculation system - 45 welds (4",
22", 28").
For No. 2 unit, PP&L intends to use the IHSI process for 103 9ut of a total of 105 non-conforming welds.
All 105 welds are also presently scheduled for augmented ISI.
,1 v '.v The 105 welds are as follows:
1.
Recirculation system discharge valve bypass line -
4-4" welds (HAZ's on weldolets).
2.
Recirc. risers 12" welds (HAZ's on sweepolets/
reducers).
3.
Core spray 12" welds (HAZ's on 316 SS flued heads).
4.
RWCU 4" welds (HAZ's on weldolets).
5.
Residual Heat Removal (RHR) system-4 6 welds (20", 24").
t 6.
Balance of recirculation system - 45 welds ( 4 '.',
22",
28").
The only other known welds that do not conform to Section III are numerous small (4 2" diameter) socket welds, used m'ostly on in-strument piping.
Because of the difficulty with reficcted signal ~
interpretation due to their small size, they are exempt from the type of inspection ( ultrasonic) that is meaningful for detecting the inside-surface originating IGSCC, and therefore, PP&L believes that augmented ISI is not required for these small welds.
The SSES leak detection system has been reviewed for h0 REG-0313 compliance and it is our determination that the design conforms to the requiremrnts of Section IV. B.l.a.
j{2F s
g
_.,__r_
\\
u ihvT.MkD. 8 b '
]
jyi e
May 1973
[p' to,,
- ~
U.S. ATOMIC ENERGY COMMISSION t REGULATORY GU DE o
DIRECTORATE OF REGULATORY STANDARDS REGULATORY GUIDE 1.44 CONTROL OF THE USE OF SENSITIZED STAINLESS STEEL A. INTRODUCTION AISI types 321 and 347) which also provide sorne protection against sensitization, since these materials are General Design Criteria 1 and 4 of Appendix A to not currently being selected for use in light water cooled 10 CFR Part 50, " General Design Criteria for Nuclear reactors.
Power Plants," require that components be designed, Process controls diould be exercised during ah fabncated, erected, and tested to quality standards stages of component manufacturing and henctor commensurate with the importance of the safety construction to minimize exposure of stainless' steel to functmr.s.to be performed and that they be designed t contaminants that could lead to stress corrosion accomodate the effects of and be compatible with the cracking. Since some degree of material contamination is environmental conditions associeted with normal inevitable during these operations, halogens and halogen operation, maintenance, testing and postulated accident bearing compounds (e.g., die lubricants, marking conditicar Appendix B to 10 CFR Part 50, " Quality compounds and masking tape) should be avoided to the Assurance Criteria for Nuclear Power Plants and Fuel d.egree practical.
Reprocessing Plants, requires that measures be established to assure meterials control and control of All cleaning sclutions, processing compounds, special processes such as welding and heat treating r r.d degressing agents, and othel foreign materials should be to assure performance of reliable testing programs. This completely removed at any stage of processing prior to
[
guide describes acceptable methods ofimplementing the any elevated temperature treatment and prior to
\\
above requirements with regard to control' of the hydrotests. Reasonable care should be taken to keep (l) application and processing of stainless steel to avoid fabrication and construction areas clean,(2) components severe sensitization that could lead to : tress corrosion protected and dry during storage and shipment, and (3) cracking. This guide applies to light-water cooled all crevices and small openings protected against r e s.c t ors. The Advisory Committee on Reactor contamination. Pickling of sensitized stainless steel Safeguards has been consulted concerning this guide and should be avoided. Special precautions should be taken has concurred in the regulatory position.
to avoid surface contamination with fluorides from welding rod coatings and fluxes. The quality of water B. DISCUSSION used for final cleaning or flushing of finished surfaces during installation should be in accordance with Regula-Control of the application and processing of tory Guide 1.37.
stainless steel to avoid severe sensitization is needed to f'
diminish the numerous occurrences of stress corrosion Solution heat treating and testing shoQd normally cracking in sensitized stainless steel components of be performed on starting material. Ilowher,in order to nuclear reactors. Test data demonstrate that sensitized assure the proper solution heat treatedbndition of the stainless steelis s'gnificantly more susceptible to stress surface areas of finished comporients, it may be corrosion cracking than non-sensitized (solution heat preferable to perform the solution heat treating and treated) stainless steel. Of specific concern in this guide testing operation at a later stag,e of component are the unstabilized austenitic stainless steels, which manufacturing, q
include American Iron and SteelInstitute (AISI) types
"'-'t 304 and 316 normally used for components of the Solution heat treating should inch.de.:ooling rates reactor coolant system and other safety related systems.
sufficiently rapid to prevent precipitation of carbides to a degree that the material is not susceptible to This guide does not cover stabilized stainless steels (e.g.,
USAEC REGUt.ATORY GU DES c.p
.e,ubl.h.d su dn me, b..nu.a d hv.mu i 6adic.aae.h e a.
0;;' " " J.'. ^'.C.d.,. i.c*";"7".: ""r.1.'.:.".; i ", *,;
c,
d m...
.~
~-
~ c
... u.
~.n.
c
- m. v.s..
e. d -
.~.
c.
c-
...h
. oc. ms.
.gd-p.
- d. gwerl e Alt n: Ch f.Pubhc P eng.st.tf.
u.t.ng.g ofic p.cb
- m..t po...l.d.cred.n s.
.d
..h.sh 6ff 3 9 n he vh. p..de.
l u.d in th. follo.v,g 3.n tur d eve n.:
- ***" ""." **. *CO.*.I .'.J'.*.**.* '.U',*,',
l*,.".'."l'.'. "**" "
- i. -
.a a e -.
M.". ':'.,0.!."' ","" ?
I: '=l"".'.T" o
%."":'.*.3.::" '*.l'::::' :::::'."'t ".*::""-
l '==,Wl:2 a r=""-
+,
I I
..s
?
2 intergranular stress corrosion. Water quere.hing(used for required for piping if its diameters are sufficiently small simple shapes such as bars and plates) shoud produce an (e.g., instrument lines and control rod drive hydraulU
/
t acceptable cooling rate. However, cooling by means systems) so that it could withstand a single failure other than water.luenching is acceptable only when the without an accompanying loss.of. coolant accident as enling rate is sufficiently rapid to prevent sensitization.
definedin 10 CFR Part 50, Appendix A.
4
'this determination is made by subjecting the material to a suitsbie in'trgranular corrosion test such as Practice E Stainless steel subjected to sensitizing temperatures of American dociety for Testing and Materials (ASTM)
(800 to 1500*F) during fabrication (except during A 262 70.8 welding) should be retested with a suitable intergranuLr corrosica test (such as ASTM A 262 70)to demonstrate Practice E of ASTM A 262 70, " Copper-Copper that the thermal treatment did not result in undue Sulfate-Sulfuric Acid Test," and the accompanying sensitization. Specimens for the retest should be screening test Practice A. " Oxalic Acid Etch Test," are subjected to a thermal treatment that duplicates the considered suitable tests for verifying non. susceptibility temperatures, number of cycles, holiing time at each j
of the material to intergranular stress corrosion.
cycle, and minimum heating and cooling rate in the 800 Although these accelerated tests use different to 1500*F range. If more than one cycle at only one environments than anticipated in reactors and do not temperature is to be used in production, one cycle with provide information relating directly to susceptibility to a holding time equivalent to the total time wah,ld be stress corrosion crackingin reactor environments, these acceptable for testing purposes.
(
tests do readily detect the presence of significant sensitization of the material, a condition which has been Under certain conditions material subjected to related with actual intergranular stress corrosion attack sensitizing temperatures (800 to 1500*F) during special in reactor environments. These specific tests are Processing may be acceptable for intended use'(e.g.,
identified here because they are the only known tests nitrided control rod drive material). These conditions endorsed by a consensus standard that includes should include, as a minimum, assurance that:
acceptance criteria (acceptable-non. acceptable basis)
- 1. The process is properly qualified and controlled for the material being tested. Altemate test methods to develop a consistent and uniform product, that can be qualifed are also acce; table.
Irrespective of heat of material and equipment used;and 2.
Adequate documentation exists that the Processed material will not develop intergranular stress Specimens for the intergranular corrosion tests from
.ts service life. Adequate corrosion during i material with carbon content greater than 0.03 percent documentation should include actual service experience should be tested in the solution heat. treated condition.
and/or test data in simulated environments and Specimens from material with carbon content of 0.03 Perating conditions. Service experience should inc!ude percent or less should be tested after a sensitizing treatment of one hout at 1250*F 2 25*F.
Positive evidence through destructive examination that intergranular stress corrosion did not occur.
1 Controls should be maintained on the chemistry of the reactor coolant and auxiliary systems fluids to which All welding processes will result in some carbide the material is exposed. Chloride and fluoride ion precipitation in the weld metal and in base metal concentrations should be specified to be less than 0.15 heat.affected zone of stainless steel we!.ds, but significant parts per million (ppm) at all times. Dissolved oxygen sensitization does not normally result when typical concentrations should be maintained below 0.10 ppm welding procedures and material chemistry are used and~
~
during periods when the material is at elevated when no further heating of material occurs However, temperatures., When the oxygen content regularly there is evidtnce that atypical welding diethods using exceeds this level, such as occurs in BWR reactor very high heat input could result in.itress corrosion
+
coolants during normal operation, sensitization of cracking in the heat.affected zone of the weld.To avoid material that is welded without subsequent solution heat this, the welding procedures and material chemistry (if treatment should be further controlled by limiting the necessaiy) should be controlled to', prevent undue carbon level in the material to 0.03 percent. Carbon level sensitization of the heat.affected zones of the control is not needed for weld metal and castings with weldments. Controls to prevent sensitization of the duplex structures since these product formswith normal material during welding may include:,(1) avoiding carbon levels have demonstrated adequate resistance to welding practices that result in the generation of high intergranular attack. Carbon level cor. trol may not be heat, (2) maintaining low heat input Sv controlling l
current, voltage, and trayal speed,(3) limiting interpass temperature, (4) using stringer bead techniques and ASTM Standard A 262 70," Recommended Practices for avoiding excessive weaving, and (5) limiting the carbon Detectins Susceptability to Intersranular Attack in Stainless Steels" may be obtained from ASTM,1916 Race Street, level of the material where section thickness makes the l
Philadelphia, Pennsylvania 19103.
material more prone to sensitization.
1.44 2 i
=.. m
=------,w-r, g
+ - -
---m.-g-
.--,-y,.-.
e-g-g-.-
-p-smme w--
+m 3
---m.--g+--=
---7---
v-----w cw y 9
~
8 In addition, miding procedures should be 4.
Material subjected to sensitizing temperature in qualified by passing a suitable intergranular corrosion the range of 800 to 1500*F,subsequer.: to solution heat test in all cases where the procedure is used for weldin; treating in accordance with subparagraph C.2. above and stainless steel having a carbon level grester than 0.03 testing in accordance with subparagraph C.3. above, percent. 'Ihe qualification test should be performed should be L Grade material; that is it should not have a using base material having the maximum carbon content carbon content greater than 0.03 percent. Exceptions anticipated and the minimum snd maximum thicknesses are:
anticipated.
(a) Material exposed to reactor coolant which has a controlled concentration of less than 0.10 ppm As a minimum, the variables tthat should be dissolved oxygen at all temperatures above 200*F during controlled in the qualification test are heat input, normal operation; or 1
interpass temperature, and welding techniques for (b) Material in the form of castings or weld, specific section thicknesses.
metal with a ferrite content of at least 5 percent;or (c) Piping in the solution annealed condition C. REGULATORY POSITION whose exposure to temperatures in the range of 800 to 1500*F has been limited to welding operations, provided Unstabilized, austenitic stainless steel of the AISI it is of sufficiently small diameter so that in the event of Type 3XX series used for components that are part of a credible postulated failure of the piping diding normal (1) the reactor coolant pressure boundary,(2) systems reactor operation, the reactor can be shuwdown and required for reactor shutdown,(3) systems required for cooled down in an orderly manner, assuming" makeup is emergency core cooling,and(4) reactor vesselinternals provided by the reactor coolant makeup systein only.
that are relied upon to permit adequate core cooling for 5.
Material subjected to sensitizing temperatures any mode of normal operation or under credible in the rany of 800 to 1500*F during heat-treating or postulated accident conditions should meet the processing other than welding, subsequent to solution following:
heat treating in accordance with subparagraph C.2.
1.
Material should be suitably cleaned and suitably above, and testing in accordance with subparagraph C.3.
protected against contaminants capable of causing stress above, should be retested in accordance with corrosion cracking during fabrication, shipment storage, subparagraph C.3. above, to demonstrate that is not construction, testing, and operation of components and susceptible to intergranular attack, except that retest is f
systems.
not required for:
(
2.
Material from which components and systems (a) Cast metal or weld metal with a ferrite 8
are to be fabricated should be solution heat treated to content of 5 percent or more; or l
produce a non. sensitized condition in the material.
(b) Material with a carbon content of 0.03 l
3.
Non. sensitization of the material
- should be percent or less that is subjected to temperatures in the venfied using ASTM A 262 70, " Recommended rany of 800 to 1500*F for less than one hour;or Practices for Detecting Susceptibility to Intergranular (c) Material exposed to special processing, Attack in Stainless Steel," Practices A or E, or another Provided the procemng is properly controlled to develop method that can be demonstrated to show a uniform product and provided that adequate non. sensitization in austenitic stainless steel. Test documentation exists of service experience and/or test Specimens should be selected from material subjected to data to demonstrate that the processing will not result in each different heat treatment practice and from each increased susceptibility to intergranular stress corrosion.
heat.
Specimens for the above retest should be taken from each heat of material and should be~ subjected to a 8 Welding procedure means procedures qualified in thermal treatment that is repretentative of the accordance to the rules of Section IX of the American Society of anticipated thermal conditions tjent the production Mechanimi Engineers Boiler and Pressure Vesset Code.
material will undergo.
6.
Welding practices and, f necessary, material
' Solution heat treated means heating to a suitable temperature, holding at that temperature long enough to cause composition should be controlled to avoid excessive aB arbides to enter into solution, and then cooling rapidly sensitization of base metal heat affected zones of enough to keep the carbon in solution.
weldments. An intergranular corrosion. test, such as sPecM M mWag@ C.1 Q, h d k
' Material of product forms with simple shpes not subject Performed for each welding procedirte to be used for to distortion during heat treatment such as plate, sheet,' bars, pipe, and tubes need not be tested provided the solution heat welding material having a carbon codtent of greater than treatment is followed by water quenching.
0.03 percent.
1.44-3
,,,,m--.-
-v-
M E ~ ' ' ' ' d:~) T b s.5'...'..'l.
.:;-1
'~
~
- _ _. J._.
3 control of the Ute of._f.enr.itized _ Stainlent r. t <. 1 ( f./7 ? )
3D-1,44 use of sensitized stainless steel co:rplies w2 t.h Control of the Regulatory Guide 1.44, with the following modification:
1.
Regulatory Position C.6:
In lieu of the intergranular alternate corrosion test specified in paragraph C.3, an test method may be used to demonstrate acceptable levels of sensitization for welding procedures.
The ASTM A708-74 standard is used to perform intergranular corrosion test in paragraph C.6 except that the radius the bend test specimen is as specified in ASME Code, I
ofSection IX with veld-base metal interf ace located at the centerline of the bend.
For the NSSS components, confortr.ance to this guide is in I 2.
as decurrented accordance with the generic BWR position I
in GESSAR (AEC Docket No. STN-50-447) Chapter 54 w
l __.
Ref erence Sections 4.5, S.2.3, and 5.2.5 l
?"
\\
(
s j
r
(
9, 2
'"*<j r-(
y 3
i
~
3D-10a/b Revision 14j December 1978
- I gum i
1 1
g s:ainless s < 3
\\
- . stances of SCC in.ornece r4n.
- :re:}icsscrealswhichve.:
zee.. ne instances of SCC of 3D-r A n have :s.:
have
.thert l
ent 3RRs is not justiffed.
I vith presently i=plemented contro s.
g Prcjection of the early SCC incidents to.curr a
discussion.
j This post: ion is supported by the following Improvements in tech-l Process Controls for 304 Stainless Stee.
since the early search and development 5.2.3.2.1.1 nology festered by extensive GE-APED re f the following caior processing CE EWP.'s have resulted in implementation o controls for 304 stainless steel.Turnace sensitized comp ineut is res:ricted :e 110,00 1) f t
d interpass te=perature of 350oF is require.
I k
Block velding is prohibited.
f cold work.
3)
~-
Restrictions are placed on ified to cinicize S
Fahrication and cleaning controls are spec 4)
^
sc 5) hibited.
Fickling of velded stainless steel is probeen veil de:enstrated, by abs Cent;;itan*s.
in' the 5)
The ef f ectiven,ess of these controls hasincident in "nor=al" 3WR serv cf a single stress corrosion cracking "n rcal" is used to distinguish (Nc:e:
5 yctre since they were implemented.
1 is f cn ahncr:al service such as chloride intrus c.
GE 3WR Service Exnerience in GE BWR service have been 3.2.3.2.1.2 General.' All known incidents of SCCThe SCC in and,
1)
- -v;s:igated and documented.
re (References 3 and 4).
have been 'su==arized in the open litera:u cracking incidents which i
Iy f ar the tijority of the stress ccrres en?WR service occurred at the first.
h ve :: curred vi:h velded cocponents in
~
s,..
ec:=crcial EWR constructed in the were designed and ecnstrue:ed United 5:a:es.
plants rust be recognized that the firs:
However, the state cf the art i
ble at the :ine.
vith the best technology availa since the present prheessing cen-1 did not approach the present-day techno cgyThe i=portance of the cer.tr I
ie We'1 css, :he
- rels were being developed at that t m. cell, recognized.Nevcr:
r *lcnt and vere no:
5.2.3.2.1.1 p
i, lined in Subsection with 301 s:ainless steel in this fire.t..
cvtr:11 service experience in :he carly plants has been excellent.
080274
(
2.
-::a
't f.-
e P
e
\\
7 GEET A7 g
I
.g..p.r.-..gp-l i*
. c, 5~
_,O '
L_ -
Ccvorcc Are Mcir ure Control for Low Fj ::7tr.,
)
Welding Electrodes (Contir.ued
,3.3.4 caled containers or ovens as unused electrodes are Eloctrodes are distributed from send of each work At the rcquired.
If any electrodes are raturned to the storage ovens.
or contaminated are discarded. for more than one shift, inadvertently left out of the ovensreconditioned in
- uct, with accordance l
thov are discarded or
- nufacturer instructions.
f=teniticSsinless k
Fabricaticn and,Proccssing cf pu.
4..
5.2.3.4
'r r
5:cels king
+.
Avcidance of Stress /Corresion Crac 5.2.3.4.1 sensitiration Avcidance of Significant l
5.2.3.4.1.1 GDCs 1 and 4, 1.44 addresses 10CFF.50, Appendix A, f
ion and prec-7.cgulatory Guide recuirements to cen:rci the applicatsensitization tha could and Appendi:. 3, vcid severe ensing of stainless steel to a Icad to stress / corrosion cracking.
purchased in the solution-All austenitic resinlesc steel was with applicable ASMI and
'i heat-treated condition in accordance e
a ASTA specifi:stiens.
ter.peratur e s v:rt;f i
- catine Cooling rates f rom solution heatrapid enough to prevent i,
ion.
Non; l
required ec lewas verified using ASTM A262, xas
.i sensitizaticn r,*
f t
methods.
minimize the pcssibilitf cf All wrought fl Material changes have been made to (IGSCC).
king intergranu'ir stress /corrosicn crache reacter ecolant,res'sure 5:sinicss steel in t 316 or 304L with 0.031 I
austeniti changed to low carben Type heundary :sr
.... T C.
l 1
m m
)
I.:.3.4.1,1 Avcidance of Significant f e.. :. :. : a :: :.- lC:..::.nued) raximum carbon content.
There is no piping which is service sensitive or nonconforming as defined in ::L F.EG-0313.
_ For =anual welds with the gas tungsten arc (GTAW) and shielded netal arc (SMAW) welding processes, the heat input was limited by weaving and welding technique restrictions.
Non-weaving (stringer bead) techniques were used where possible.
Enen required, weaving was controlled to meet the following bead width limits:
for GTAW, the lesser of five times the filler wira dia.eter or 7/16 inch: f or S!.AW, the lesser of fou'r times t.4 the electrode ccre wire diameter or 5/8 inch.
For machine, 9-
- autenatic, and n.anual welding with processes except GTAW and S!'AU, heat input was restricted to-50,000 joules per inch.
Inter: ass temperature was restricted tc.50'F for all stainless 2
steel welds.
Righ heat welding processes such as block welding and e'_ectreslag welding were not permitted.
All weld filler metal and castings were recuired by specificatien te have a minimum of
.u..._ _ _. _ _ e.
1 Whenever any wrcught austenitic stainless steel was heated to te..,: era res over 800*F by means other than welding or thermal cc::ing, the material was re-solutien heat treated.
s There :::::cis vere used to avoid severe se.nsi irati n and to_c ccr.:1 ith F.ec.ul at.crv. Guide 1.44, Cent:01 cf the !_'se cf Sensitized e
.c.. c _.. _ e.e e_ etc__,
.e 5
5 Fcr c:r.--itment, revision, and scope, see Secticn 1. S.
-a#4+
A e
-w 5.2.2.4.1.2 Process Controls to Minimi2.e Ixpcsure to Centahinants l
l Ex :stre tc contaninants capable of causing stress /ccrresica rracking cf austenitic stainless steel cc.._cnents.::s arcided by l
o. 4,.,.
8-9
l i'
{ c
,- (w:. t+.wC-m-NUREG 0420 l [
)
)
l
'E
. e
^
l i Safety Evaluation ?neport E re'a:ec. to t',e 03eration of ower Station, o
i S,orenam N uc. ear s Unit No.1 x
t E
Docket No. 50-322 e
2
.f +
Long Island Lighting Company y
>a E
s
=
g U.S. Nuclear Regulatory E Commission Office of Nuclear Reactor Regulation "ia z
April 1981 o
m uJ
.. as e e
Po
[z l'.'kAT.,I-s e
- s. Qe,y:
a
~ e" t
(
~
x n.,
uJ
~
4/3
$e, R
3e m
m
,1 e
We conclude that, with the exceptions noted above, the thermal-hydraulic design of the core conforms to the Commission's regulations and to applicable regulatory guides and staff technic'al positions as set forth in Standard Review Plan Sec-tion 4.4 and is, therefore, acceptable.
4.5 Reactor Materials 4.5.1 Control Rod System Structural Materials, The mechanical properties of structural materials selected by the applicant for the control rod system components of Shoreham that are exposed to the reactor coolant satisfy the criteria of Appendix I of Section III of the American Society of Mechanical Engineers Code and Parts A, B, and C of Section II of the Code, and conform with our position as stated in Section 3.5.1 of the Standard Review Plan that the yield strength of cold worked austenitic stainless steel should not. exceed 90,000 pounds per quare inch.
q d
The controls imposed upon the austenitic stainless steel of the mechanisms T satisfy the intent of the recommendations of Regulatory Guide 1.31, "Controis-.
of Ferrite Content in Stainless Steel Weld Metal," and _Regulatorv Guide 1.44, "Contml of the Use of Sensitized Stainless Steel." Faorication ano neat treat-ment practices performed in accordance witn these recommendations provide added assurance that stress corrosion cracking will not occur during the design life of the component.
The compatibility of all materials used in the control rod system, in contact with the reactor coolant, satisfies the criteria of Articles NB-2160 and NS-3120 of Section III of the Code.
Precipitation-hardening stain-less steels have been given tempering or aging treatments in accordance with our positions as stated in Section 4.5.1 of the Standard Review Plan.
Cleaning and cleanliness control are in accordance with Regulatory Guide 1.37, " Quality l
Assurance Requirements for Cleaning of Fluid Systems and Associated Components of Water-Cooled Nuclear Pcwer Plants."
Conformance with the codes, standards, and regulatory guides indicated above, conformance with our positions on the allowable maximum yield strength of cold worked austenitic stainless steel, and the tempering or aging temperatures of martensitic and precipitation-hardened stainless steel, constitute an acceptable basis for mee_ ting in part the requirements of General Design Criterion 26.
~
4.5.2 Reactor Internal and Core Support Materials j
The applicant has met the requirements of General Design Criterion Ibnd Section l
50.55a of 10 CFR Part 50 by assuring.that the design, fabrication, a'nd testing of the materials used in the reactor internals and core support structure are i
l of high quality standards and adequate for structural integrity.
The controls i
imposed upon components constructed of austenitic stainless stesi satisfy the intent of the recommendations of Regulatory Guide 1.31, " Control of Ferrite Content in Stainless Steel Weld Metal" and Regulatory Guide 1.44, "Codtrol of the_Use of Sensitized Stainless Steel."
The materials used for the construction of components of the reactor internals j
and core support structures have been ide:itified by specification and found to i
be in conformance with the requirements of NUREG-2000 of Section III and Parts A, B, and C of Section II fo the ASME Code.
As proven by extensive tests and
\\-
4-23 Lj
l
]
]
\\
satisfactory oerformance, the specified materials are compatible with the 1
expected environment and corrosion is expected to be negligible.
The controls imposed on the reactor coolant chemistry provide reasonable assurance that the reactor internals and core support structure will be adequately protected during operation from conditions which could lead to stress corrosion of the materials and loss of ccmponent integrity.
The material selection, fabrication practices, examination and testing procedures, and control practice performed in accordance to those recommendations provide reasonable assurance that the materials used for the reactor internals and core support structure will be in a metallurgical condition to preclude service Conformance with the requirements of the ASME Code and the intent deterioration.
of the recommendations of the regulatory guides constitutes an acceptable basis for meeting the requirements of General Design Criterion 1 and Section 50.55a of 10 CFR Part 50.
4.6 Functional Design of Reactivity Control Systems l4 4.6.1 General The control rod drive system and. recirculation flow control system are designed for reactivity control during power operation.
Reactivity is controlled in the event of fast transients by automatic rod insertion.
In the event the reactor cannot be shut down with the control rods, the operator can actuate the standby liquid control system which pumps a solution of sodium pentaborate into the primary system.
4.6.2 Control Rod System Each control rod is moved by a separate hydraulic control unit. A supply pump provides the hydraulic control units with water from the condensate storage tank for cooling the rods and for moving them into and out of the core, with a spare pump on standby. The pump also provides water to a scram accumulator in When neces-each hydraulic control unit to maintain the desired water inventory.
sary, the accumulator forces watar into the drive system to scram the control rod connected to that hydraulic control unit.
At lower pressures the volume of water in the scram accumulator is sufficient to scram the rod.
At higher -
pressures most of the water to scram is provided from the reactor vesseT.
A single failure in a hydraulic control unit would result in the failud of only one rod.
In addition, any single component may be removed from the4ontrol rod drive (CRD) system without disabling the protective system. The protection system has been designed to permit periodic functional testing during power operation with the capability to test indiv* dual scram channels and motion of individual control rods independently, thus complying with the requifements of Criterion 21 of the General Design Criteria.
0,.
Preoperational tests of the control rod drive hydraulic system will he' conducted
'1 to determine operability of the system.
Startup tests will be conducted over the range of temperatures and pressures from shutdown to operating conditions in order to determine compliance with applicable technical specifications.
Each rod that is partially or fully withdrawn during operation will be exercised one notch at least once each week.
Operable control rods are tasted for com-4-24
-__.y_.__-
3 0
Information to be reported will include all abnormalities ranging from mirn wear observed during normal inspection to complete failures, including failure to open or close and inadvertent operation.
We will require the applicant to participate in this program.
To reduce the effects of safety-relief valve discharge to the suppression pool, l
the applicant has changed the safety-relief valve discharge device from a ramshead to a quencher design.
The applicant has stated that the overpressure protection l
will not be affected by this change. From a transient standpoint, the safety-l relief valve discharge critical flow is the flow of interest. The applicant l
has stated that the change from a ramshead to a quencher does not affect the critical flow and, therefore, does not affect the overpressure calculations.
On this basis, we find the change is acceptable with regard to the overpressure l
protection function.
In addition, a startup test will be performed to demon-i t
strate expected safety-relief valve discharge flow.
l In summary we have reviewed the system design to prevent overpressurization of f
I E
We conclude that this system, conforms to the the reactor coolant system.
requirements of General Design Criterion 15 and the American Society of Mechani-However, this cal Engineers Boiler and Pressure Vessel code and is acceptable.
evaluation is subject to confirmation by the ODYN re-analyses discussed above.
5.2.6 Reactor Coolant Pressure Boundary Materials 5.2.6.1 Material Specifications and Compatibility with Reactor Coolant l
1 The materials used for construction of components of the reactor coolant pressure boundary, including the reactor vessel and its appurtenances, have been identified l
by specification and found to be in conformance with the requirements of Sec-tion III of the ASME Code.
i General.. corrosion of all materials except carbon and low alloy steel will be j
negligible._ For these materials, conservative corrosion allowances have been proyided fcn all exposed surfaces of carbon and low alloy steel in accordance
~
The external nonmetallic with the requirements of ection III of the ASME Code.
insillation to be used on austenitic stainless steel components conforms with the requirements of Regulatory Guide 1.36, " Nonmetallic Thermal Insulation for l
Austenitic Stainless Steels."
Further protection against corrosion problems will be provided by control of j
the chemical environment. The composition of the reactor coolant wil'1 be controlled and the proposed maximum contaminant levels have been shown by tests and service experience to be adequate to protect against corrosion and stress corrosion problems.
The controls imposed on reactor coolant chemistry are in conformance Eith the recommendations of Regulatory Guide 1.56, " Maintenance of Water Purity irn Boiling Water Reactors," and provide reasonably assurance that the reactor coolant pressure boundary components will be adequately protected during operation from conditions that could lead to stress corrosion of the materials and loss of structural integrity of a component.
The instrumentation and sampling provisions recommended in Regulatory Guide 1.56 for monitoring reactor coolant water chemistry provide adequate capability to 5-6
. "')
'l
/
ctions to bring the.
1 i
detect changes on a timely basis and to effect correct ve a t ess corrosion.
I.
cos1&nt chemistry within limits which will prevent s r ce with the recom-The use of materials of proven performance and the conforman constitutes an acceptable d 31 of the General Design j
b mendations of the regulatory guides mentioned a ovebas Criteria.
Stainless Steel Pipe Cracking steel piping at 5.2.6.2 l
In September 1974, cracking was experienced in the stain ess This was the first of a series oft o Dresden Nuclear Power Station, Unit No. 2.
l incidents of intergranular stress corrosion cracking tha l tion system bypass piping
/
As a result of these incidents, a special task i
affected zones in Type 304 stainless steel rec rcu a auses of the cracking.
systems and core spray lines.
in the staff technical group within the NRC was formed to investigate the cTh h
L report, " Investigation and Evaluation of CrackingPipin in the reactor coolant 7 October 1975.
Fu I
i The task group found that austenitic stainless steel pip ngtible to stre high residual pressure boundary of boiling water reactors is suscepThis was to welds.
t d zones adjacent to welds j
stresses, and some sensitization of metal adjacent cracking.
cracks were expected to be preser.t in the heat-affec e ensitization has not l
and cracks should not occur outside these zones whe recommendations for led.
The two bypass lines have been entirely The applicant is implementing a number of the task group s i ting weldolets will be capped identified areas of high susceptibility.
removed from the recirculation system and the ex s in contact with the coolan utilizing a corrosion resistant (304L) weld inlaytransition pieces will The ' core spray safe-ends and the stainless steel l
teel. The control rod drive hydraulic return line will be removed an vessel return nozzle will be capped.
i h regard to intergranular
~
Although the applicant has taken corrective actions w ts may be appro i
stress corrosion as indicated above, additional act ond Processing itting NUREG-0313, Re We have issued a generic letter to the applicant transmWe sion 1, " Technical Report on Material Selection an BWR Coolant Pressure Boundary Piping." applicant's re Fabrication and Processing of Ferritic Materials lant pressure 5.2.6.3 Fracture toughness of the ferritic materials in the reactor c Welding of all collip6nents t
boundary is discussed in Section 5.3.1 of this repor.
i This compliance with the Code provides re of ferritic steels was performed in accordance w Code, Sections III and IX. assurance that cracking of components l
during fabrication.
5-7
~
)
q areas con.,.y of steels in limited accesstion for Areas s
used for ferriticide 1.71, " Welder Qualificaand examinat assurance reasonable ill The fabrication practices iding proceduresne intent of Regulatory Gu recommendations providecoolant pressure boundary id Accessibility." with these in the reactor ility.
acceptable of restricted accessib constitutes an accordance ferritic stainless steels em:d in 4
l Design Criteria 1 and 1,
ions ed above
.tisf actory in locat i
regulatory guides ment on requirements of Generaof Austenitic Stainless Sjeel with the stain-armtnce of austenitice inch, in l
s for teeting the and Processin components boundary, noding 90,000 pounds per squarf the l
w 1
Fabrication
[.6.4 reactor coolant pressurea yield strength exceestated in Se steel hin the steel havewith our position f austenitic stainlessvessel and 4
as Technical "
.s reactor components constructed o
- ordance of Branch l Welding, boundary and for the recommendationstrol of Stainless Stee
" Regula-J-
an.
controls imposed upon reactor coolant pressure itized Stainless Steel, satisfy the intent of the s
for Cleaning of Fluid System
.J ulatory Guide 1.31, " Con ie sed in the
" Control of the Use of Sens Power Plants," Regulatory appurtenances ositien MTEB 5-1 on Reg Guide 1_.Mua ity Assurance Requirements of Water Cooled Nuclearof Limited Accessibility.
ts 44 tection ification for Areas tion procedures, and proreasonable te ulato
, 'Q ui e ts and Associated ComponenGuide 1.71, "Weider Qual tion practices, examinarecommendations provide
.ory coolant pressurea mettalurg reactor with thesesteel in theres) and incracking during Material selection, fabrica accordance austenitic stainlessfrom hot cracking (Microfissu procedures performed in constitutes stress corrosionmentioned abov 1
assurance that the susceptibility toregulatory guides boundary will be freewhich precludes g
meeting the requirements o and Testing with the
- c. ondition Conformance an acceptable basis for Inservice Ins ectionre Boundary, service.
which are part Reactor Coolant Pressure of Reactor Coolant Pressu Bounda and 14'.
components iodic inspection k tight..
32," Inspection 50, requires, in part, that be designed to permit perassess str i
S.2.7 General Design CriterionAppendix A of 10 CFR Part l
i and' to operation i 1 of the ASME Code Class of the reactorand testing of importanensure that no deleter o t areas i
with nes The design accordanc'e
. integrity.and weld heat-affected zo To at Shoreham. reactor coolant pressurinspection ind to fa vice periodically during seraccess.for inserviceMethods have been de 3
welds of the ts and Class 2 componen incorporates provisions for N
reactor l
de.
of the iled requirements forh he Section XI of the ASME Co areas of those r cooled nuclearconstru I
remote inspectionto inspection personne.
l R Part 50, defines the detai n programs for
/
a Based upon Section 50.55a(g),10 CFand inservice inspect oincluding support preservicepower facility components 5-8
/
.msnm*
m.-
--,,,,,,,,,n-
--,,-,-wy------,,.----
3 6 ENGINEERED SAFETY FEATURES The purpose of the various engineered safety features in a nuclear power plant is to provide a complete and consistent means of assuring that the public will be protected from excessive exposure to radioactive materials should a major accident occur in the plant.
Systems and components designated as engineered safety features are designed to be capable of performing their function of assuring safe shutdown of the reactor under the adverse conditions of the various postulated design basis accidents de:,cribed in Section 15 of this report.
They i
are designed to seismic Category I standards and they will function even with complete loss of offsite power. Components and systems are provided with suf-ficientredundancysothatasinglefailureofanycomponentorsystemwilj not result in the loss of the plant's capability to achieve and maintain a: safe shutdown of the reactor. The instrumentation and control system for each 6 engineered safety feature is designed to the same seismic, redundancy, and' quality requirements as the system it serves.
Instrumentation and contro1*
systems are discussed in Section 7 of this report.
ts 6.1' Engineered Safety Features Materials The mechanical properties of materials selected for the engineered safety features satisfy Appendix I to Section III of the American Society of Mechanical Engineers-(ASME) Code, or Parts A, B, and C of Section II of the ASME Code.
The controls imposed on the use and fabrication of the austenitic stainless steel of the systems satisfy the requirements of our position on Regulatory Guide 1.31, " Control of the Use of Sensitized Stainless Steel."
The controls placed on concentrations of leachable impurities in nonmetallic thermal insulation used on austenitic stainless steel components of the
' engineered safety features are in accordance with Regulatory Guide 1.36,
" Nonmetallic Thermal Insulation for Austenitic Stainless Steel."
1 Conformance with the ASME Code and regulatory guides mentioned above, and with, our positions on stainless steel, constitute an acceptable basis for meeting applicable requirements of Criteria 35, 38, and 41 of the General Defign Criteria and is acceptable.
??
l We have evaluated the protective coating system used inside the primary contain-ment and determined that the system meets the rccommendations of Regulatory Guide 1.54, " Quality Assurance Requirements for Protective Coatings Applied to Water-Cooled Nuclear Power Plants." The total amount of unqualified paints and organic materials inside the primary containment is estimated to ber negligible. We therefore, conclude that the applicant's protective coating system is acceptable.
6.2 Containment Systems The containment systems for the Shoreham Nuclear Station include a Mark II type containment structure as the primary containment, a secondary containment surrounding the primary containment and housing equipment essential to safe 6-1
.i.i----,
m-c L. gu.C-t m -i.:s r..
, 39 3
- 3
.IA T: ', 9 7,:
e REGULATORY Gt1TDE 1.44 Initial Issue:
Revision 0, May 1973 Current Issue:
Revision 0, May 1973 La salle C.P.
Issued:
September 10, 1973
)
CONTROL OF THE USE OF SENSITIZED STAINLESS STEEL Regulatory Guide 1.44 describes acceptable methods for the control of the application and processing of stainless steel to avoid severe sensitization that could lead to stress corrosion cracking.
The purpose of this guide is to address 10 CFR 50 Appendix A, i y
GDC's 1 and 4, and Appendix B requirements to control "the application and processing of stainless steel to avoid severe g,
i c--
sensitization that could lead to stress corrosion cracking. " The i
guide proposes that this should be done by limiting sensitizat' ion due to welding as measured by ASTM A262 Practice A or E, or a,;other-method that can.be demonstrated to show nonsensitization in austenitic stainless steels.
l Tests by General Electric indicate that the test specified by i
A262 A or E (Detecting Susceptibility to Intergranular Attack in stainless Steel) detects sensitization in a gross way, and that the tests do not provide a precise method of predicting susceptibility to stress corrosion cracking in the BWR environment.
All austenitic stainless steel for LSCS Units 1 and 2 was
. purchased in the solution heat treated condition in-accordance with applicable ASME and ASTM specifications.
Carbon content was limited to 0.08% maximum, and cooling rates from solution heat treating temperatures were required to be rapid enough to pryvent, sensitiza tion.
Welding heat input was restricted to 110,000 joulesperiNch maximum, and interpass temperature to 3500 F.
High heativelding processes such as block welding and electroslag welding,were n'ot permitted.
All weld filler metal and castings were required by specification to have a minimum of 5% ferrite.
As deposited austenitic welds were controlled to have at least 35 ferriEh content.
See Subsection 5. 2. 3. 4 for specific reference toF*
control of the use of sensitized stainless steel.
Additionally, grinding of field erection welds was prchibited on the reactor fluid side of Class 1 stainless steel pressure boundary pipe.
Whenever any wrought austenitic stainless steel was heated to l
temperatures over 8000 F by means other than welding or thermal cutting, the material was re-solutien heat treated.
B.1-51
-v--
,.-._..._.-.4
,y-
- IF-73AF A':r*::.:::::.R
,s
_)
}';
MAR 0is 1!78 These controls were used to avoid severe sensitization and, to
" Control of the comply with the intent of Regulatory Guide 1.44, Use of sensitized stainless steel."
We believe that we comply with the intent of this guide via incorporation of the alternate approach cited above.
L
.1
- ~ ~ ~
(,-
9-..
/
s
.g J_-
-=
-Y r
f I
44 B.1-52 e
.~
~'
- 1..~, t cu racu y e l
i
(
n_
M.S 1. FIANrs, Mineger
]
16:5t Ma valu e j
, l [.g{
Ce ne r al Inf o. cr.... )
l'
- P 0 fit d s
/
- r. O. Sne 2612 5.;
Denver, Colorado 80201 Phone (303) 825-2101 Eg g
. CENERAI, INFORMATION ON NUCLEAR PO WR EXPERIENCE (NPE)
,NPE, compiles and reports on the operating
- when it happened (if it was prior to the start emperiences of all large light water nuclear of commercial operations, the appropriate power plants in the U.S.A.
We report phase of startup testing, e.g., "during fuel emperiences of IMR's located in other coun.
loading"),
tries also, but so far not as much informa-tion is available on those plants.
- what happened, d
we concentrati on operating problems (equip-
- why, ment breakdowns, malfunctions, outages, etc.).
Because of this our output tends to take on
',
- how it could have been prevented, 1
a very negative slant. Readers should keep in mind that the'information we supply is
- and how it was corrected.
not Indicative of overall plant experience.
One has only to glance through our "Availa-References bility" section to find that, on the whole, At the and of each problem description,$there will nucleaf power' plants are compiline envious records of reliable service.
be one or more lower case letters enclos$d by parentheses. These letters direct the reader to the reference sections located at the baek,of either Purposes of NPE volume, BWR-2 or PWR-2, and indicate the* s'o'urce(s)
To help our subscribers of the information.
C.; L
- stay current with the problems occurring The 1,oose 1.eef Filine System in the nuclear power industry, There are F baste volumes (some have more than one
- and to assist them in retrieving infor-binder):
nation on past problems.
- A thin Light Blue Volume is used for a number of miscellaneous purposes. It contains a key
[.
word index, an availability sutscary, a list of By doing this.we expect to help our s
abbreviations and filing instructions. In subscribers addition there is a section entitled " LATEST MONTH OPERATING EXPERIENCES" that can be used
- prevent outages, (if the subscriber wishes to do so) to file the
- and save engineering and technician latest information for volumes SWR.2 and PWR-2 manhours.
for one month. In this way readers can use any spare time they have during the month to look our System over the new material. At the end of the month the pages are filed in volumes BWR-2 and PWR-2.
Each month we systematically review a very large amount of literature (periodicals, Note's New subscribers should decide if technical papers, technical reports, they will use the " LATEST MONTB OPERATING correspondence between the plant owner and EIFERIENCES" feature and instruct the the NRC, etc.). During t.his review, we Person (s) responsible for the. filing,,
select only the information pertaining to accordingly.
k-'
operating problems. This information is condensed somewhat (but we try not to lose
- Volume BWR-1 (red) is indexed"by boiling water content) and categorised. At the end of reactor plant naen. It serves as a companion the month we send the information to ove vc.lume to the more important volu:me BWR-2.
subscribers and they file it in the loose Volume BWR-1 contains briet plant descriptions leaf filing system which we supply.
and operating histories for each plant. The i
descriptions are prepared and issued at roughly Problem Descriptions the same time that a,new plant " starts up" and becomes commercially available. The initial In writing the descriptions of the opera-versions of the histories are issued af ter the ting problems (flied in volumes BWR-2 and P ant has acquired significan :empe rience.
l PWR-2) we try to include (if the information is available):
- where the problem occurred, C.
e e
e===
Apr 81
,v
O m
}
119 t P.l u e Yalume
. Central Info. es gr[
p.2 I
- The key word index group = ali probicos 4th the
- Valume BVR-2 (red) is indexed by boiling water scactor plant system and/or component. This sa:e key words.
. is the most significant part of the NPE system
- If you know that a particular plant had a partf-(f:r BWR's) and is where the informaU on on the cular problem, refer to the cross-reference for opirating problem is filed. This volume also that plant.
contains cross-references (plant name vs.
priblem item number.).
Note Same problems may still be filed in the Latest M nth Operating Experiences" section.
- VJ1ume P W 1 (dark blue) is the pressurised water r: actor equivalent of volume BWR-1.
Abbreviations
' ~'
- V:1ume'pWR-2 (dark blue) is the pressurised water To save space and reading time we use a number of rsector equivaleet of volume sm2.
abbreviations. % " Light Blue Volume" contalas our abbreviation list. In constructing our list Pirnta Already in operation we started with the American National Standards Institute abbreviations (ANSI Y10.19 - 1969). We B:,sist11y we will be reporting the experiences had to modify some of them so that they agreed rith what we thought was more corcon practice and of rnesv: red during the previous month. However, course we had to add some to the list.
thout 2 dozen plants started operating before we g
did (May 72) and we felt that our compilation Pare Nssabering System j
would be, incomplete if we did not go back and W
repIrt on those old experiences. We did this The " Filing Instructions" in the " Light Blue Volume" durits our first year of existence.
contain an explanation of our page numbering system.
Basically what we are attempting to do is to keegt,
St ytne Current the reader informed as to exactly what section ha;,
is in. This*is especially helpful if the page is va publish a large quantity of information every temporarily stored in the " Latest Month Operating month. Because we realize that most people do
' ast h ve much spare reading time we have designed a tumber of " reading time severs" into our systems issue Date_
- tha two column format for easier readinge The last month in which a particular page was issued k
or revised is always shown in the lower right hand
+
- es underlined upper case headline for each opera-corner of the page.
ting prob 1cm (you can skip the details of code Letters in the Marrins prsblems you are not interested in),
We use the following code letters in the left fiand
- basic filing is by plant component or system.
. margins of the pages in volumes BVR-2 and PWR-2.
(if you are only interested in instrumentation tnd control, you can skip everything else; the N-Nw - - - - A problem that we learned p;ge numbering system provides the clues).
about for the first time during the past N
=onth.
- thi page nu bering system (at the top of en-h p;ge) aliayg tella you what type of reactor (BVR or Fwr.) and what component and/or system O - old - - - - A Prob 2e= that *= ara. ly as
, including for the first time but on 7ou are reading about, a result f ur a forts to thorou ly
- the "N" (for New), ?O" (for Old) and "R" (for r C 2pletely report on sign cant a
ravised) code numbersin the margins (you may only w
ba interested in new or revised information).
-u,--
.a R1trisval of Information R - nevised - - - - we =ar have =ncovered some additional information-Iluring the R
past month that concerns a previousiy ThIre are many reasons why a person may wish to reported problem, or we may be correcting cetsr the loose leaf filing system to retrieve an a mistake.
capirience. For example, you may be experiencing 2
a prsble:s and wish to know how someone else has
- [
rolvsd the same problem in the past. NPE has 3 Teedback fastures that allow guick entry into tI7 filing tystems E encourages feedback from subscribers (recom-
- The basic r ethod of filing is by BWR or FVR plant mandations, suggested improvements, experiences system and/or corsponent.
we may have missed, etc.) and will willingly act as an information processing point for the industry.
k li I
- -*o----.-.e Apr 81 l
l
l' e,' g ?
b p.1
. 's
- Ch C.)
' h N.
]~ j g
]
it FiF f IE!!,E VFut;
/
,o 6k in. *. t.tter ht,se. 24h it 1o9g, cotpleh l
r 4
. atteched.
Quad-Cities 1 - Sept 70 (constr r ti M. 4y'br
/p,
2th fittit M/
st'y/6in.Crubberhose. 12 ft 1onte comp ete l
A seoidering wood fire occurred on to),si he toway /*
ge fth **Chicarr" fitting on one end aSd a copper w
core plate. The fire started from the test lease 8 8 fitting on the other.
from a 1500 V quarts flood light which wh moun M-4 A 2 conductor welding cable. 24% f t long, above the temporary wood decking used on tdW attached to a TIC welding gun. These conductors core plate. This flood light ignited either the were covered with asbenos insulation which had insulation materiet on a welding cable coiled on the been repaired with plastic tape.
wood decking er the plywood material of the decking.
The fire spread across a polyethylene plastic covered The outfit was fished out using "J" books. Se wooden frame which was utilised during erection oper*
riser was inspected and the jet pump was retastelled i
attone to control air currests around an optical a*
into the reactor vessel.
lignment instrument.
R ne test progra. was deiered about 4 days. (ejstr.=o) 1 2
)
Materials consumed tactuded about sixteen 1 f t core plate covers, 40 ft of neoprene welding cable. 3 f t of plastic ventilation air duct and 80 f t2 of poly-3.
CUIDE Tutts CAMF iS05E - CAUSED FUgL DAMACE ethylene.
Tarapur 1 - Aug 71 Three unsuccessful attempts to extinguish the fire were made from below the core plate with CO2 and dry Both reactors at Tarapur are operating with only 284 chemicals. A fourth attempt from above, with domin.
Instead of their ultimate capability of 368 fuel as-etalised water was successful.
semblies. Therefore 20 peripheral control rod guide tube positions havo less than therfull comple=
The fire lef t a deposit of black soot on the interior sent of 4 fuel bundles. The fuel support positions surface of the vessel shroud in the area above the which are unoccupied by fuel are plugged with fuel fire. Alse, the bottom of the vessel contained a small amount of water and bits and pieces of debris, support plugs. The fuel supports are attached to on the lower core structure, black soot and combia.
the guide tubes by welds. No control rods or drives are installed in these 20 locations. To compensate tion products were deposited on the exposed metal for the lack of the normal downward force afforded surfaces.
by the fuel weight, the CRD thermal sleeves are re-T'.e interior of the RPV was cleaned with demineral.
strained by a hold down device, ne thermal sleeve tred water, acetone, demineralized water with 500 is engaged with the guide tube bayonet and then ppen srisodius phosphate, wire brushes and putty locked against rotation with a set screw which on-gages a slot in the CRD housing flange. A core sup-knives.
Fort pin which engages lugs on the guide tube, pre-Analysis of the occurrence, the debris, and measure.
vents guide tube rotation.
cients indicated that the tecperatures did not sen-sittse the lower core piste structure.
The set screw dich is installed af ter the thermal sleeve is positioned prevents the thermal sleeve Because the dry emical extinguisher contained from rotating. Tts set screw is turned into the potasslun chloride, they performed another cleaning thermal sleeve ring until the cup point bottoms in operatica of_ the vessel interior before performing the CRD housing slet. It la backed out turn to the systein bydro.
prevent shearing during relative thermal movement between the CKD housing and the thermal sleeve. The Corrective reasures included 1) fire retardant paint set screw is not locked. For CRD ' housings in which on wood on the lower core plate, 2) discontinuing no CRD's are installed nothing prgvents the screw the use of quarts la:ps, and 3) additional approved fr a backing m.
fire extinguishers located in the area and in the
,a
- I*
I*II ~
During a refueling outage. dedge was discovered.
Wo of the 20 guide tube assemblies had become dis-c nnected. Each of these assemblies have 2 fuel 2.
wet.DINC OUTF1T FOUND IN JET PINP RISE 1t bundles installed. Once disconnected these assee.
b tu were lif ted during reactor operation by the R
quad-cities 2-APr 72 (startup testing) h draulic forces created by incoming reactor water fl v.
ey r se a distance of =*3-f t until restrained Jet pump flow comparison checks were being performed as a final check prior to installation of the steam by the underside of the shroud g steam separator
- - Calculations con-separator, steam dryer, and reactor vessel head.
1, fra hat if not rutrained, it is possible for guide,
This test requires that both reactor recirculation eup uena g
het bundles to rise pumps be running at the same speed with the recircu-
'I***
lation header crosstle valves closed. The indicated flow rates on 2 of the jet purps were about 507, of The motion of these 2 cobponents caused damage to the other 13 jet pue:ps. One jet pump was removed themselves and to adjacent structural components and the obstruction was found to be a heliarc welding Holes were worn through the fuel channels of 2 neigh-outfit consisting of:
boring fuel eierents and a s,:all section of the clad-W 9 *=*=. 8mm e e="*
Nov 77 e
4
. -n-
It Int.
S.
w= m yt e s. -- ~.. - - s u vrT p.w ow Dresden 2 - Startup d m ef 3 fuel rods was also worn tnrogn. In addt-t ten, the motion of the lif ted control r:t guide De vibration test progra: showed that th'e jet pump
(
t i.b a t broke 2 in-core guide tube stabiliser rods.
riser brace experienced high vibration whenever a All da9 age was confined to the imediate vicinity mismatch in pump speed between the recirculation of the 2 lifted guide tubes.
pumps occurred. This could occur whea one recircu-1stion pump was operating at partial flow while the Untreking of the 2 asseethlies appears to have been other one is at full flow. It could also occur fol-ceutsd by a f ailure of the set screws due to improper loving a pump trip when an attempt is made to start instellation and/or fatigue failure. Lif ting of the the inoperable ping. The hydraulic forces can cause 2 guide tubes was caused by the eventual random rota-an increase la jet pump vibration, throwing an addi.
tisa of the themat sleeves and the subsequent un-tional load on the mouncing bracket. Under auch a locking of the guide tube combined with a lack of condition half of the jet pumps would be operating tufficient downward force to withstand the upward slightly above their normal rating, forcing backflow hydtculic forces on the guide tubes. The upward through the remaining jet pumps, which have insufft-ferca is due to the pressure drop occurring across cient driving flow. The increased vibration occurs ths core plate. It has been calculated that install-in the 1cw flow pumps and occurs at the point which ins 2 or more fuel bundles in each assenbty provides the flow is at the point of reversing. An interlock a sufficient downward force' to overcome any lif ting systen was installed thats forcss which are present during normal operations.
-automatically prevents restart of an idle recircula-The design of plants,in the U.S. differs significantly tion pump when the operating pump exceeds 65% of frss that at Tarapur. Therefore no potential for a design speed, an alarm sounds if such a restart is s
S comperable incident exists.
(er bev) attempted.
-prevents racirculation pump operation (positive interlocks on pump speed controls) in the vibr'a-~
4 VELDINC PURCE DAM LEFT Its LPCI - BLOCKS JET tion regions.
i.s.
-sounds alarms if changes occur and the undesired mp regions are approached.
Irresden 3 - Nar 71 (after initial criticality)
De jet pump brackets were strengthened in all plants subsequent to Dresden 2 (Millstone 1, Fukushima I, Iuring functional tests of the Recirculation System, 4
etc.) The same interlocks were installed at Quad a low flow indication was observed on a jet pu=p.
Cities 1 & 2.
(bs, fr, ga) hg Nomal flow indications were recorded previously.
~
Chscks revealed no probleas with the flow transmitter (.
or indicators. The reactor head and vessel internals were then remove,d to gain access to the jet pu=ps.
6.
PEER ASYNMETRY - CORE INLET TEMPERAUME The vessel water level was lowered to~ 2/3 core DIFFERENCES height. Radiation levels were ~ 35 cr/hr. The cause fer reduced flow was found to be a weldin;; purge das Dresden 2 - Spring 70 (plant startup) ladpd in the transition casting of the jet pu p at a point whara the flow is divided from the riser inlet Differences between TIP and LPRM readings indicated to 2 noaales for two jet pumps. n e purge da: vas that a power asyurnetty problem existed. The asym-led;sd again.: r..e casting and blocked ~ 3/4 of the metry ratio was as high as 1:1.16.
flow to.the jet pump. Upon removal, the rurge das During a July,i70 shutdown some low sensitivity '
was revealed to be a standard plywood and rubber LPRM's were replaced. A gama scan confirme! the ds=,16 in. In dia, cons nly used in construction o
hallarc welding operatic s.
Due to the st=e and char-asymetry and indicated a ratto of ~ 1:1.07.';
n seteristics of the dem, it must have originated from Various tests and. analyses were conducted to rule out ths 16 in. LPCI discharge piping. De LPCI system is the following as causes
]
the enly source of 16 in pipe that discharges direct-8 ly into the recirculation piping.
.- fuel, control rod or control curtain effects.
a The dam consisted of a sandwich of four 3/8 in. thick
- flow non-uniformity or flow variations between plywood semi-circular segments and 2 circular 1/8 in.
thick red rubber sheets held together by h - 20 x lt in.
"*i stava bolts. One half was broken in two, the other was e nc u at t cause was non-waiform tem-half was badly fouled and the 2 halves were held to-Perature of the core flow at the core int and that gathsr by the normally used chain. Various portions this could be corrected by properly distr uting the of the intemals were inspected, vacuumed, and flushed, feedwater flow around the downcomer anm.lus.
Missing from the dem and not recogered were 30 in.2 D e feedwater sparger on Dresden 3 was modified of, rubber in small pieces,16 in. of wood splinterse (enlarging and plugging selected holes) and tests 4 nut screw combinations,1 washer and 1 nut.
LFCI run in 1971 support the nop-unifom temperature and HPCI testable check valves were disassembled but theory, and the idea that the situation can be cor-no foreign inaterials were found. All fuel asse=blies rected by properly distributing the flow around the g
vare rc=oved, game scanned and inspected. Additional downcomer annulus. However, as of Feb 72, the re-chseks on various systems and coeponents were c.ede.
sults were not considered conclusive because of er-(fs) rors and uncertainties in the incore instrumentation data.
l
- -= e m-- - -.
yoy yy 1.
..l
s ut, p... )
II. P * *: t "
3
- p. 3
,)
i Feedwater sparger modificatter. w*rt also made at rotated aneut i. '. ann was vesting be neen N..
D 4,ad Cities I and 2 and other p. ants.
and t5. veus t ull because sf s, failure of the to a d e asse-bly. The Se:a Lalt,t.erbly and inu st They believe that the asyrretrien do not present a washer were cilssing. the retairer and connectis..
safety probige.
(fr. sk. al' hk) 1/2 in.. holt icapture the besa and belt dating servicing) were detached, and one keeper on one of the lower restrainer clamp bolts was etssing.
7.
WEUT110M SOURCE ENCApSUIATION FtlLLitt A underwatcr search (with TV) located and recovered Big Rock Pt.
- Apr 72 all missing parts with the exception of the 2.25 in. dia insert washcr. They feel that the most we avuttiary neutron sources hed see. Installed for likely resting place would be in the vessel se the about 1 yr. They are entloony & tyllium sources with bottori of the outer annulus. thering the process of as inner encapsulation of SS and an outer encapsu-raising the beam bolt retainer, it was inadvertently letten of a strconius alley. A visual inspection of dropped into the opening in the support shelf lef t the sources revealed that the outer encaosulation by the removed jet pump No.17 eimer and it settled had failed la the eres of the lower end-cap weld on into the bottom of the reactor lower plenue. It both sources. The failed area had the appearance was sighted but they were not able to recover it, I.ater, during repair operations, portions of a characteristic of typical hydride attack.
welding outfit were inadvertritly dropped. All parts were recovered except for a small 1 la. 0D The lower end cap was recoved from one of the sources
"'*Prene bushing and half of its compaoloe 13/8 in.
and the inner SS encapsulation was visually examined.
OD compression cap. Analyses indicated that none of No abnormalities were detected. It was concluded the lost items would pose a safety pretes, that the cause of the feiture was due to internal contaminattoa.
There was no damage to the riser portse.and seat.
the restrainer gate mount or the ellP* joint seat A new outer encapsulation was provided using strcalloy o the diffuser. Both of the bene bolt, heeper tubes with the upper end caps wclded to the tubing.
loc wel s w m to n.
Und m ater W emees of The SS encapsulation was placed in this tubing and the other pumps resealed:
a lower end cap inserted. Mechanical forces were used to crimp the tubing to the lower end cap.
-3 of 35 beaci keeper lock welds appeared to be Because of the activity of the sources. At was in-possible to quickly perform the repair in a moisture restrainer gate keeper lock welds free environment. Th w holes were provided at the appeared to be marginal.
top and the bottom of the sheathing to pemit flow
-1 of 19 restrainer gate ved;es was icproperly of water and steam in the annulus to eliminate the po sit tor.ed.
effects of having moisture trapped inside the
-1 lower restrainer claep bolt was loose by strcalloy tubing at the lower end cap. Because of several turns but the keeper was soundly weldcd this they planned to monitor for Mtimonv. In their g, p,,,,
g water cheelstry progree.
(13)
A netallurgical exam was perfor-ed and a special 300 f t-lb counter-clockwise torque test was applied 8.
JET PUMP H01.D DOWN ASSEMBLY FAlltlRE to all beai bolts. The keeper lock welds on 2 purg*.
sheared during this test indicating snadequate beam Quad-cities 2 - Aug 72 preload.
A fretre fump trip occurred while at [57. power.
As a result of exams, analyses and tests. they 6uring recovery, and while at 607. power. various concluded that the f ailure occurred because the in,trunentation showed abnomal readings which led hold down beac bolt keeper was inadequately field to the conclusion that the No.17 jet pump had welded and/or field preload was nel app!!ed.to failed.
the hold down beam.
Either of tMete could have 1ed to loosening and the dI*aesembly of the pump.
The jet pump assembly provides for re= oval of the
,g Theinlet-nimerismechanicallyclamped*/
inlet-mixer section if inservice inspection la The loss of the retainer andel/2 in. bolt fre the top of the mixer was probabLO unrelated. The bolt desired.
to the riser vita a remotely operable beam bolt is lockwelded to the retainert however, there was arrangement which transaits a 23.000 lb downward very little veld material. Norcel vibration claeping force. Thla is - 3x the upward force probably rotated the bolt out and allond the bolt and ensures mintoal leakage at the spherical as.d retainer tn fall free. Coopective a:tions (on sealing joint at the end of the riser. A lock all pumps) in:luded:
I welded keeper piece over the bolt prevents jap rotation of the bolt. there'~ maintaining the bes*
-correcting all observed installation deficiencies.
load. A pair of lock welds ar,e provided to pre"
-refurbishing puep No.17 vent rotation. The lower discharte end of the
-breaking all bene bolt keeper welds and re-miser is held Iderally hv i restr !rce bracket applying require <i team and tes. bnlt preloads.
which cust be opened for r.t::cr remvat. It is
.reveld all bes: bolt keepers locked in the closed position by a pair of bolts
-recheck that the veldE were a,pplied to cect and they also have lock welded keepers. The swing
- torque testing and that the hea3 bolts do not ing gate portion of this bracket is latched to a
=ove when torqued in both directions (this alsn 8
fixed frene so that even if the bolts were loose checked pretensioning), and the swinging nember would have to be raised
-reconfirn that all beam bolt keeper wlds were 0.3 in, for disengageoent.
sound af ter the torque tent.
N ne dur.eton of the shutdawn was 55 dars. (e< qv.,i>
Following shutdown and head removal, the inspection revealed that pump No. 17 had lifted 3 to 6 in..
- - ~ " * - -
- Jan 74
.~
.,---r
t
- .I~4 M.* tst Int.
)
eni of the th r..-
8.;t which encare vessel I. 4 mountir.3 Lasc.et
- r.? dr}. %1ste in the sparr -
bracket allow fer retitiva t etion t,ct ween the ve...
f*tEL BRACE tHPptD ON & DAMAGES TM FUEL L
3 cui0t Wedges bearing er alnnt the vessel wall are used to preload the sparsers and prevent vibration. The peach go'ttom 3 - May 72 (construction) amcent of spring. or preload. is enough to resist the force of the N flow and any calculated transients During erecties of the structural steel in the without allowing lif ting from the installed positten.
reactor buildles, a dissonal brace (8 x 6 x The calculated flow load for 10$% rated flow (360'y) 9/16 in. angle, 13 ft 10-3/4 in. less, weighing to 1316 poueda and would produce a 0.194 is. dis-
~ 360 lb) slipped from its theker and dropped from placerent. The reactor assembly drawing specifies an estimated beight of 30 ft. The steel erection a 3/1 16 preload. The man aparger stress.
sibcontracter was evidently following accepted at the uncties of the header and Tee box la rigging preceduree.' The angle brace pierced the calculated for 1057 rated power at 10.970 poi and thipping container of the top fuel guide. The 13.380 pet for S/16 and 3/8 ta. displacement. The czact extent of damage could not be determined at origleal spargers were built by stearns-Rogers to CE tha time because the reactor building crane was drawings and specs.
not yet operational. In July they were able to uncrate the guide 'and make -dimensional checks.
The 4 spargers were reaoved from the reactor vesset Asc s which make up the ce!!s of the guide were and were emanined on the reactor refueling floor.
j bint and pushed out of posittoa. The unit was The moderate radiation field frera the sparsers shipped off site to a location where the special prevented any extended close visual observation. All aptics required for a thorough examination were 4 spargers showed extensive cracking at the, $
avellable. Steps were taken to ptotect equipment connections between tee box-to-headers and tee bou--
st= rad in the reactor building at Elev. 234 with to-thermal sleeve. The E header of the NE and the Noltnkins ad care-lias. ae be== wa= real ced.
s header of the ra: sparger were hanging on by s.orin (nr.asi) ligaments at the bottom of the heeders with open h through-wall cracks extending around the remainduF
,1 G.
JET PPfP CRACKS - CARgURIZATION of the joint. Open cracking was visible on 3 sparsers
/
at some of the tee box joints. FT revealed additional
/
Dresdea 2 & 3 - early 1969 (construction) cracking in all 4 sparsers. The cracks tended to be singular at steple joint configurations such as j
An investigation of castings used in the original the machined undercut on the tee box near the ther=al I jst pump assemblies revealed cracks in the tran.
sleeve joint, and multiple at complex joint con-sitica ptece,180' elbow, noaste. coupling and figurations such as the rear corners of the tee box.
[,
blange. The cracks werTcaused by surf ace The SE aparger showed the least amount of header.
N joint cracking. All cracks were on the external e
carburitation. The casting supplier had been.
employing the shell-mold process and used car.
surfaces of the sparsers except for the 2 open cracks bonaceous materials in the binder and wash.
La the NE and W sparger.
The foundry process was modified'and replacement castings were purchased and installed.
(oe)
There was no eridence of general cracking at any other point on the spargers. A few isolated cracks were detected in fillet welds on the lifting lugs J
- 11. N spARcER CRACKS and brackets, and one small crack was located near the middle of the South header of the SE sparger.
M111stene 1-- Oct 72 Other isol.ted cracks were located at work-hardened areas such as sheared piste edges, drilled hole Th2 reactor internals were inspected as part of the surfaces and a ding mark on the East header of the investigations that, sere performed after the chloride SE,sparger.
intrusion incident (see VI. C. 3).
Visual exam-S. --
inations revealed failures of the N spargers. An The end brackets of all spargers except the'3r extensive metallographic and analytical icvestigation showed semicircular wear patterns, on theCvessel vzs perforced to determine the cause of the failures.
side edges of the slots, caused by the nMrided $$
It was concluded that the failures occurred prior pin bearing on the annealed 55 bar. In?some cases t
to the chloride escursion as the result of high-cycle octagonal wear patterns from the vesses bracket fatigue following a condition of inadequate cold were observed. The pin keepers and the pin handles spring. New spargers of codified design were installed also showed wear pgtterns caused by high, frequency with a carefully controlled procedure to ensure vibration of the pins. One su pin keeper,was that proper cold spring was applied to all sparsers.
entirely broken off.
Metallographic analysis coupled with the elegiance Thsre are 4 sparsers located in the upper portion of the vessel. Their primary purpose is to distri.
of wear and long-term vibration of the sposger bute the reactor W. although the reactor
- cleanup coeponents indicated that the major failure mechanism Isop also discharges back through the R N line (SE was fatigue and that the failures were transgrinular in nature.
end,$V spargers). A series of 1 1/2 in. holes.
i spsced at equal intervals alnns each sparger branch.
discharge the water radially into the flow passing The most probal1c cause was 1. proper installation.
This resulted in inadequate preload, and subsequent from the stea= separators down into the jet jucps.
failure by eechanical fatigue. The extensive dssage and oxides indicated a considerable period of Alth:vgh access to reactor internals does not require operation af ter the failure, whir.h confirmed that it, the spargers are designed to be removable. The the cracking was not a result of the chloride tharmal sleeve, welded to the sparger at assembly, incident.
is a slip-fit in the vessel noaste. Brackets on the gg h E
g.
I..
t-r. !
- 11. Et actM D'
bear cf the t* e s..,et.
's el" la. 5 s p..c t r butt e stn/ t..
nitridad pin t:.th Ft..
. Q ut a r al cycit.
'" b allswa ennsidirnhie..d lu *
.cf rs!sttre ravemtr.t.
b.r c?
t.
, vme i rumf t ism a..
- u st the pralsad p'aint, inc l udin..
brerksts esd sting symr tracket sh e et thzt theu cert t-members provided t he rear raint.
t owe e 4 r vm weld distortion.
Lack of assembly preload, together with the thermal-W new ut str9 lo e ade using f. in. *che dule fifs sleeve gap produced by the assembly procedure, was pipe fratre r than schedule 40) for the sparger sufficient to explain the severe vibration which was
. len. Thi. allows a higher preload to be imposed evident from wear patteros on 3 of the 4 spargers.
with a given r an stren. The Lee hem (divideri was reduced from I in, to 0.5 in. wall thicknt ss, gasically, a sparger was set in place with 1/2 in.
which Provides some flexibility and reduction esf spacers at the point where preload was applied. The stress in the attachment weld of the sparser Itg.
brackets were edjusted to allow the piiss to engage the reactor sparger support brackets and then welded.
A special computer progrura is used to calculate The sparger was removed and wedges, whose thickness sparser stra spes. Thust are at 12 points, mainly provide for proper preload, were welded to the in the pin.nd bracket aros, and in the Lee box sparger. Hydraulic jacks were used, at final install-and attachment zone. The new Installation plans ation, to apring the sparger enough to insert the call for a 0.5n10.030 preload which will result pins. Procedures called for recording of the gap, in a prcloed of 530 lbs.* Under normal operating between the sparger and the vessel wall, during conditions, the man stress produced by this pre-fit-up and at final Installation, to allow preload load is 17,4M psi at the sparger leg attachm nt to to be checked.
the tee box.
A review of the installation data indicated that.
Additional specs cc.ntrol at least 3 problem areas if the gaps were recordeJ! properly, the preloads in which the original sparger was weaket (1) the did not meet the design intent. The 5V sparger, material as purchased in the solution agnealed Norret 43, lodicated a ** negative" cold springt condition, with a man hardness of Rockwell 3963 1.e., the floal gap 9/32 in. less the the inital (2) the parts are :.ot heated for forming) and must gap. Recorded cold spring values on the other be solution-heat treated If processed Isr.a manner spargers varied froa 1/32 to 9/1(a in.
Inducing cold works and (3) welding is done by the gas tungsten or gas metal-arc process with cor.trols The cost likely cause for improper preload was on heat ir.put and ferrite content of the weld believed to be the veld distortion. When velding metal. This process avoids potential fluoride the cold spring wedges on the pipe, experience conto-ination from veld-flux fumes. In addition, had shown that weld distortion tends to increase the cutting fluids as well as other materials used the chordal length of the sparger. This can in processing. and the cleaning process are control-reduce the cold spring if not corrected. The led to prevent contamination by thlorides, fluorides, problem was first discovered at he Vermont Yankee and ether potentially harmful materials.
plant when, prior to operation of the plant, the spargers were removed for codifications to improve The new spargers were installed using a modification fV distribution. It was discovered that the pins of the initial proccdure and recording detailed were eatlly removable, despite the supposed pre-measurem nts for subsequent engineering evaluation.
load. u was necessary to add draw beads to restore The spargers were shipped to the plant site with the dest;n preload. Isever plants with the wedge the therral sleeve welded in place but not cut design use a teplate, or its equivalent, to to proper length. The thermal sleeve was insert-insure that the sparser can be returned to its ed in place on the FV nosale and measured in erfainal ehepe after the wedges are velded in order to cut to proper length.
place. 1here were no draw beads on the'Hillstone a
sparter, 30 it was assured that this effect was The in-vessel'sparger support brackets were machin-over-looked.
ed to provide a flat upp'er surface removing the portion of eaterial that was partially
- worn away To expedite the sparger replacement program, the by vibration of the original spargers.'
sparger design being built in C shops for 218 in.
4 vessels was modified to fit the 224 in. Millstone The sparpr was again lowcred into place and the vessel. This allowed the incorporation of several thermal sleeve inserted in the Jsonale. A measuring j
design Inprovements c:ade since the original Hillstone pin was utilized to deterritne tisc proper position-i design. Considerable improvenent had been made in ing of the end brackets on thelsparger. This the distribution of IV to the various jet pumps.
positioning applied a measured amount of preload The latest criteria called for the flow to be to the sparger. The sparger was again renoved equal 12.51, to each segment associated with s jet from the vessel and the end brackets welded in g
i pump. A aeries of tests had bee:. run on flow p
,c,,
distribution, and a comput+r progn as developed to
'7 obtain the correct hole patterns to achieve untform The sparger was reinstalled and tisdenis forced l
flow
- into position with a spreader bar-so the pins could be insialled. The gap between.the sparger Sparger distortion during site installation, which and the vessel was neasured and increased by occurred during the velding of the pre' load wedges, turning tbc Jcck bolts to provide the required required considerable rework in the field. This final value of old sprirt. The jack olts may have been the cause of inadequate preload in wcre then tack welded in place.
the original itillstone installation. The new design provides a shop-veldtd saddle, which mounts licasurenente were obtained as specified during the Installatlon.
(pt.ga.th)
- -.= e.-
e gg j
gem-m.
.-m.
- = =. = -
2-
e s
s
,s.)
I!. b $ctor Int.
Teledyne. Southw st kesearth and Ct were all in.
volved in the investigative program.
f e4'
\\
17.
s marn 1N conf stirpoRT cPtt Underwater TV revealed that both the NW and Nt sparsers were cracked in the junction bem and Isre sden 1 - 1961 & 62 front plate, subsequent close-up emanination (in vessel) revealed loose jack belt / vessel contact.
During 1961 inspections 8 minor cracas or irregu=
loose end bracket plas, and changes la sparger larities were found at the wlds of tne core support prenimity to the vessel watt on all but the SE i
grid. Although these cracks did not af fect the sparger. Tests indicated that only the SE sparser structural integrity of the reattor internals they had maintained the cold as-installed preload contt:; red thee to ensure that propagation had not (a4000 lb). The SE sparger is the entry point for
)
neutred. During the refueling outage that started clean up system return flow. While the SW Nif, and
)
in how 62 a berescope was used to monitor 99 of the e M sparsers showed a resistance to movement, some j
volds previously inspected. Including the previously exhibited any significant preload.
nstsd cracks. and 44 additional welds. They had to novs 29 fuel assemblies to perform these inspections.
purther inspection revealed:
The condition of the cracks had not changed. One
-Some wear (~1/16 in.) in sparger pin slots and m ee soall track was found in'the additional area inspscted.
(re) on the pins.
-Wear on top of RpV brackets which support sparger end brackets.
-Batter definition of cracking on NW and NE sparggts.
13.
N SPARCER CRACKS - HICH CYCit FATICtII
-A small (~1 la. len5th) interrupted linear FT m
Indication on the top side of the SW sparger.
(,
Mittstone 1 - Apr 73
-A small (~1 7/A in, length) interrupted linear y*
"***I"""
.d'.
pl&naed shutdown and inspection (as part of the
-Thermal sleeve /noeste support pad wear.
e sit water intrusion recovery survel11ance program.
"" I *
- "*# S**
see VI.C.3) revealed new cracking in 2 of the 4 t
sparger.
new IV sparsers (see 11.11). Subsequent esame disclosed that the 3rd and 4th spargers had pr Although analysis of the failure was incomplete, crack indications. This 2nd set of spargers. No.2 it was postulated that the fatture was caused by stesign replaced the originally installed which a less of preload due to the effect of superposition' ses found to be severely cracked. Based on the of thermal strain in excess of elastic 11mLi die
/
anitysis which followed the chloride intrusion
~
(
to a radial temperature gradient, upon a prestressed incidsnt, it was concluded that the cracks in the
/
section of pipe under-fixed bending deflection.
spargers had occurred prior to the chloride transient g
g
,g gg e a result of high cycle fatigue following a by hydraulle inputs af ter loss of preload, caused co.dition of inadequate cold preload. The fatture failure by fatigue resulting in the observed cracks of the 2nd set of spargers. in the absence of in the sparger tee-boxes and wear on end pins, ch!cride, substantiated the conclusion regarding the
- brackets, and thermal sleeves. However. It was initial failure. However It raised questions noted that one of the sparsers had maintained its concerning the completeness of the definition of preload, but had a minor PT crack Indication.
sparger operating condition, used as a basis for the Millstone design.
A cold flow test was performed with the fully loaded reactor in the col'd shutdown condition with the reactor pressure vessel head removed. The N system was operated over the full operating range of flows.
1%o tests were performed: the 1st with the shrged.
head removed and without cleanup systeci flowr^the 2nd with shroud head installed and with etwenup system flow. All 4 N spargers were instrarmented with varying combinations of acceleror:eters, atrain
[
gases, and dynamic pressure transducers. JAlso included was one dynamic pressure transducer located
.)
in each of the 4 fW lines outside the vessel.
,7 N
1 1
l l
The St. SW. and NE spargers were recoved, f$atrument-l
)
ed. and rainstalled with no change in preload.',The
{
fs11ed NW sparger (sent to Vallecitos for meEft-ad lurgical exams) was replaced with a new instrumented sparser of the same design and installed with a
///
preload of 1700 lb. The as-tested set of sparsers
/
(
/
then represented the range of observed failures.
/
in terms of degree of preload and cracking.
The SW sparger was observed to vibrate the most, I
being the only sparger to exhibit responses at
(
discrete frequencies. V!brations were observed
)
in the vertical plane at 20 and 40 Ha in a mode 2
we= e - -. -
Jugy 73 g
r
l 5
whc re u selero^cte r: 6t ??
te e e d s;strar arm
.s were in phuc. Vibratlur. lave:
) the radial
}
s.'
/
,dirsctiin were comptrable tu levens in ths vertical direction.
Th. re vt sa d desgn prc.vidsde The pressure spectrum for the SW sparger arm and
-$olid, low-stress attachment in bc,th ve r s i..I tee-box contained numerous sharp peaks which and horizontal planes independent af t*~
f lo-included structural resonances determined by shaker ibility of the sparger structure.
)
test. In contrast, t' a SE pressure spectrum
-Elimination of preload requirements.
i e ntained broad-band low-level response with smaller
-Increased natural frequency of the sparrer
]
l peaks at frequencies generally above structural structure.
l resonance) determined by shaker test. Pressure
-Reduced possible high stresses resulting f ron-spectra for the face plate was similar to the tee-high amplitude vibrations (vessel axial and boic for the SE, NE, and NW spargers. The NE and radial directions).
NW indicated very little discrete frequency content
-Removal of the sparger as a primary load carrying although there were small blips at blade-passing member (except for moreal operating loads carried frequency which were also observed on the SE and by a greatly reduced moment are fixed at the SW FW line transducers.
vessel).
As mentioned abose the' change involved the weldinr.
preliminary analysis revealed no significant of brackets to the cladding of the reactor visua l differences between signal levels in the tests with and without the shroud head / separators in-interior. The bracket veldment was done af ter 2 stalled. While neither of the failures experienced layers of weld-clad were provided to build up the posed any threat to public safety nor any detectable clad thickness to preclude the bracket veld ent heat-af fected-rone encroach:sent into base metal.
anomalies in plant performance, cracked sparsers was an undesirable and Abnormal operating condition.
Millstone estimated that it would regtre 75 to To gain an ucierstanding of the problem an extensive 100 man-ren (i.e., 30 to 4h welders each receiving additional program of analysis and testing was up to 2.5 rem) for the velders to complete the job.
init16ted to define the failure mechantxa. A 3rd
,13 set of apargers. No.3 design, was fabricated to a new design which was based on information available Testing and analysis of the No.3 design depended at the time, in order to return the plant to service in part upon results of analysis and testing of the while analysis ef forts continued. The new spargers No.2 design directed at defining details of stcarty-were instrumented with dp transducers, strain sages state hydraulic input which resulted in the chu rved and RTD's and a hot flow testing prograra was planned, failures. Further, definitions of steady-state and transient thermal-hydraulic inputs tu the The new spargers were of a fundamentally different Millstone spargers were required to cer:plete ths support design and this was expected to reduce the analysis. Hot flow testing data was to be rccordsd possibility of fatigue f ailure. This was achieved at prescribed intervals during nornal rcxtor by effecting a substantial reduction in operating startup and during pcwcr increase to rated to.sd.
and initial installation cean stresses by adding a Provision was made for continuous recording of up primary load carrying member which provided essen-to IS sensor readings sieultaneously for IEited tially rigid attachment of the sparger t#the time periods. Major objectives of the tests to !.s vessel. 'Ihle assentially removed the sparger as conducted during reactor startup and initial pi.rlod a primary load carrying cember and eliminated the of operation weret dependence of structural response on preloading.
As a result, bcth the inttially installed and
-To incasure the amplitude of possible sparger operating mean stresses were markedly reduced vibratton under service conditions.
(initially installed from near yield down to
-107. yields operating stresses no longer include
-To identify possible dif ferences in the a :plitude preload and as a result are ~25i of yield) based and frequcncy of pressure oscillations as co.p.. rcd on then current definition of applied loads.
to cold test data. and identif# driving functic.s for sparger responses in tje IV syste. and in !Je Applied load restraint for each sparger was provided the spargers.
E by a bracket welded to both LLa RPV wall and the
. 7 sparger header adjacent to the N nozzle. This attachment provided stability to both horizontal
-To diagnose possible therjnal effects which cuy affect sparger perfortance. These ceasur a nts and vertical planes. Midway betueen the bracket included through-wall temperature gradients and the ends of the sparger headers, horizontal imposed on the sparger during startup and durin-and vertical (one side only i.e.,
horizontally normal operations, and theis plitude of possibic outward and vertically upward) amplitude limiting water temperature variationTin the sic!nity of dampers were attached to the RPV wall. Clearance the sparger, due to uneven 5)gu mixing.
between the sparger and dampers was set to a low A
value at assembly.
-To provide a comparison of thermal ef fects in the hV sparger and the SE sparger that are influt n.cd The welded attachment to the RPV eliminated the by cicanup system flow '.n the SE sparger.
Jack-bolts used for contact with the RPV in the y
previous design. Otherwise, with the execption of instrumentation attachment. the revised design was unchanged.
,Dee.
eger-e.g
.- )
tb pr M r te a' 1"
- w. *.
s.: < G*. tw :. vit t t l
Jt. p.escpr tat.
)
a-n. t.
te v;than rie liwite for etM s;
- p. N
.s t a ruter crot tu. 1ns j results alsa lA M A Raf. sy ef farsd (ddtMansi it.f arratian en thart I cyclin;: ef f ects which may have bu t..
prsblems seconbry or cuatributing; cause of sow previm.s'
/
t
-It. could be concluded from results of the cold sparger failures. These ef fects were associat<-
tests (mentioned above) that 1,oth of t%s loose with the mixir.g of coolant flows at dissimita:
temperatures. Analyses showed that the No. 3 cpirgers (NE and SW) were prone to f1Na induced vibration of the type which involved interaction design spargers would withstand this thermal cvt1-ing, in combination with vibration levels at <
cf the vibrations with the flow. There were tus strong indications the vibration was self-excited, 337. p6wer, for > lyr without risk of failure.
to the very shallow nature of cyclic temperature rtther them forced by pressure fluctuations le variations in the metal (only a few mills) which
' De 3 features (mounting bracket, midpoint bumpers contributed most to the fatigue usage facter in the feed lines.
and end shima) of design No.3 were expected to evaluation, it was concluded the even if crack work individually or, is combination to seduce the initiation were to besia, it would not propagate to a gross sparger f ailure without high vibratory prabability of sparger failure.
- Dey installed dp sensors between all 4 vertical stresses.
riaers that feed the sparsers, specifically for 71bration data from the sparger instrumentation was operating surveillance purposes. Significant obtained for a period of several weeks following op readings would be cause to shut down.
the completion of the planned test program.
This
-Hetallurgical exams of tFe SV and SE sparsers data confirmed that the low-level vibration at SC" revealed their failure mode was high cycle power was not increasing with time.
(sy) fatigue.
Data obtained from dynamic pressure sensors provided Raf. aca described the methods.used to install the evidence that the dominant mode of sparger vibration new spargers. Radiation measurements showed the was not a result of pressure fluctuations in the IV vassel walla in the sparser area would contribute This evidence was consistent with thas.
most of the dose rate during subsequent repairs.
system.
Thi vessel walls were hydrolasered resulting in 1 obtained from cold flow vibration tests of thelNo. 2 reehr contact with the vessel wall and 400-500 design, which were conducted during the reacpr aresa/hr in the general area. At this time, the outage prior to installation of the No. 3 design It was concluded from these tests that dicision was made to proceed with the sparger repair spargers.
sporger vibration was flow induced and self-excited, work with the fully loaded core in place. An aluminus work platform was installed in the vessel rather than forced by line pressure oscialations.
ta cleanup the sparger area, PT and remove the Vibration characteristics of the No. 3 dealgn were A wooden platform was constructed above in some respects similar to those observed in cold th3 aluminum platform to prevent any subsequent flow testing, and it'was believed that similar spergers.
tools or equipment used from f alling into the mechanisms were operative.
The wooden platform was made of 14 sections of 3/4 in, plywood, "pir shaped", and held in place Efforts to better define the nature of the flow-csre.
induced vibration niechanisms were continuing, and by a 6 in, steel ring attached to the aluminum more work was necessary to develop and confirm a platform. De vessel clad in the sparger area solution to the problem. power was being held at m
was them UT tested and cleaned with' the hydrolaser.
80*,. and this would probably continue until the Ta provide for reducing the radiation level for the next planned refueling outage (Sept 74). A plan vsssel repair work, extensive use was made of lead sheet and blankets hung on the ID of the vessel vall was developed for off-site engineering activities from the reactor vessel studs. Lead blankets were steed at providing a finst resolution. The plan also installed on top of the wooden platfern,at the tentatively includedt reactG vessel ?D. f.xposure was 266 R to 285 men *
-The creation of a full scale test facility to be completed in Jan 74.
Thz areas where clad weld build up would occur
-The performance of tests in the facility.on design vsre layed out. The areas were PT tested for No. 3 in Feb, on a prototype of a design No. 4 in cracks and then ground from special clean rooms P. arch and April, and on a final desigac$o. 4 in attached to the vessel well. Velding on the designofthermalstressfixIpectsfor u
vessel clad began shortly af terward. As each pad vald buildup progressed, the new weld was ground, desip No. 4 7
-The development of applicable depign No. 3 removal the 1
ee rino eac arger was then trued up. De 4 feedwater spargers were then end fabrication of an instrumentation placed into the nossles for fit up of the brackets package for design No. 4.
design of mechanical and strr'se fix aspects of to be installed from the clad pad to the sparger.
h
-1 e e
Th3se brackets were then welded to the new pads, the design No. 4 prototype, fabricabion and produc-Tinal fit up of the sparge.rs took place with tion f inal design No. 4 replacemy spargers to shin measurements being made in the vessel. The sparsers were then removed from the vessel and the R * * *' " '" ^"' *
~(***** "'* *'*"*
snd brackets welded in place. The sparsers were Instrumented in order to follow their operation N see ite:= tt 16 f or addiciaa=1 inrar==tica-during reactor operation. Then all 4 new spargers ware installed in the vessel.
(acn)
Startup occurred in mid-July af ter a 3 mo shutdown.
The hot flow test results showed that the amplitude of sparser vibration was high at 1007. power (this I
e qn -
s ~
July SO l
a e.
,m, meen-amantme-.i
_M m....,
.w.
.._._n.,,-
p. 'I p.v-1.
en PT ecespttnce innpection
]
,littens.
ir operations, g,
h ut.:er service
. - N. 7 3 A, r ing wu to be casrcas'ad thrsughout ths repa Nine Mile It.
bly.
t Fla-t nanagement considered that the dryer imnoae tracks were discovered in the steam dryer assemne assembly is mounte dryer no significant safety risks, however, the h next asser.bly was to be reinspected during t e the steam separator assembly.
eal sides of the wet stesa plenum and provides a s d
acy scheduled refueling cutage to verify the a equ between the wet steam plenum and the dry steamflowing out the top an of the modifications.
(xb.agv) dryer Moisture is removed by in: pinging on theha and found vanes, flowing down through collecting trougdrain tubes to the reactor wa Iuring the Spring 74 refueling outage they r
They additional cracks in steam dryer chambers.ficient welds.
were caused by are strikes and insuf (syh) annulus below the steam separators.
The affected areas were replaced.
The initial discovery resulted from a visual exam-ination which reve f
red 3
This crack IMADE00 ATE STEAM SEPARATORS & DRYERS, in the top longitudinal weld joining the No.
diffuser to its vertical support plate.
24 in, then f
15, extended along the weld centerline for ~
1971-72 I
l support VGrgassen -
curved downward about 3 in. In the verticaGE was requested to investigate th The cyclone separators are arranged in an annu a l
durine and coordinate repair procedure.
configuration so they can be lef t in placeD e chev plate.
refueling operations. Commissioning tests '
coved for refueling.
3
.011 in. be installed at equal intervals along affecte the struct-f weldments of all 8 diffusers to rein orcefatigue failure, l Nov 7@ ute and reduce the possibility of wm e==-- % +=. I t . ~. ,,.m I . :er Int. '; E aleeve into the nsa:Is. Tha sp:rger and inter:- 7 af ths nazzle could br O psetad far cracking car: ths prsesdure. CL w:s ..ng to rsconennd this Le with n'a protsed an the spargt; and ths rstatning done ecn all plants of this d.rsign, including oper:- ( pins at the ends of the s pra-arn were essentially ting plants. lessa except for the reactin; force against the hvdraul'- flow. The only variable introduced was CE recomended this procedure be accomplished prior ths change in the average radial gap between the g g g nasale and thermal sleeve (i.e. leakage area varied). preparation for fuel loading. The 4 spargers were Bissd on this, CE recomended the following: 1) removed and found to have diametral gaps of 6 to 13 mils. The pipes were expanded to a negative opsr: ting Plants - Perform inservice inspections clearance of 1 to 4 mils to assure tight fit and during refueling outages to determine the conditions reinstalled. The operation took < 1 wk. of spargers: 2) Near Operating Plants - By expanding the (parser thermal sleeve, the radial gap between ths noasle and thermal slaeve could be reduced to a point where instability muld not occur. The 18. FAft2RE OF BIADE CUIDE YOKE sptrger should be removed ?*.s impanded, as necessary, ts give the proper radial gap co eliminate flow induced vibrations; and 3) Future Plants - For those Arnold - Feb 74 (prior to initial fuel loading) plcats where vessel and thermal sleeve fabrication and installation schedules allow, the thermal Ihring operator training, action was undertaken to sleeve was to be welded to the nozzle safe end, the lift 2 blade guides from a storage location in the thirmal sleeve design was similar to the advanced dry fuel pool. When the 2 blade guides, connected wald:d design for use on BVR/6, no change was by a yoke at the top, had been lifted about 3 in., rsquired. (akb) the yoke handle broke. The blade guides fell back into the storage location and a piece of the handle R felt to the fioo, sf *.be fuel, pool. All piecds were retrieved. Pe yoke is cast atuait.ua. Ja y I'r of the handles of the blade guides connected by the failed yoke indicated a crack in one corner' of one handle. ,All remaining blade guides were removed for individual tests and exams. All blade guide handles and yoke handles were loaded to 1.5 times static load and then Irf examined for indications of cracks. One s-blade guide handle and one yoke handle were rejected des to indications of porosity. Only guides and yokes with tested and accepted handles were returned to the vessel. (eno) =
- 19., JET PUMP FAILURES Quad-Cities 1 - Apr 74 t
A scheduled jet pump inspection was performed dur-ing refueling operations. This was carried out with the aid of underwater TV and a video tape., unit, ne inspection included i -Inspection of the tack welds on the beam. bolt-keepers and a check of beam pocket fits 7 -Observation of the beam bolt assembif while 17. EntSSIVE W SPARCER CLEARANCg applying 200 f t-lb of torque to thfbeam bolt in both directions. ~ Cooper - Jan 74 (prior to initial fuel -Inspection of the restrainer gate and wedge positions ~ loading) and restrainer bolt tack welds. - As part of the CE evaluation of cracking in W The following discrepancies were notedt spargers at Hillstone (see II.13), it was discovered -While applying up to 200 f t-lb to the'* beam bolts, thet there was excessive clearance (about 11 mils) movement was detected on 3 bolt assemQes. (10, between the thermal sleeve (which is a single P pe 13, and 16). They failed the test. i attached to the W sparger) and the reactor vessel 3 restrainer gate bolts and keepers N re found to nozzle into which the sparger themal sleeve fits. be completely sheared and missing, 2 of these were Vhin the radial cleatance exceeds about 6 mils a from pump 5; one was from pump 6. vibration occurs that contributes to crackitts of -30 of 37 remaining restrainer bolt keeper welds l. the sparger and nozzle wall. CE developed a fix were either missing or cracked'. k which consisted of 1) re cring the spargers from -The restrainer gate wedge on jet pump 5 was missing, ths nozzles (done from inside the vessel), 2) -The restrainer gate on pump 18 showed signs of measuring the diametral gap between thermal sleeve . wear in the vicinity of the wedge indicating and norsle, 3) expanding the therust sleeve to make possible vertical movement of the wedge prior to a tighter fit, and 4) re-insertion of the thermal the inspection. e---~*--* Feb 77 9 t .A tubssqusnt inspectf97 cf the inlet rissr brecc3 frrpumps5andtishow,*tubrrce]vssaalwilds ]
- p. 1:
to b3 intret. v ./ t: point w.s located as a severed instru ent 1*. ( The following parts were nissing: 3 restrainer gate where the line was attached for suppart ta en -
- t..
bolt heads, 3 restrainer gate bolt head keepers.1 It appeared the failure could have bect c au.. : r; wedge, and 1 wedge spring. Atte:spts to locate and excessive vibration of pump 6 due to the tailed retrieve the missing parts retrieved the following restraints. taox) 1 restrainer gate bolt head. 2 restrainer gate bolt head keepers; 1 w dge in 3 pieces (wedge, wedge blade, and wedge apring): and 1 additional wedge Ref. aqr provided additional information. ne restrainer gate wedge was missing on pucp 6 (not 5). bale. Several of the original restrainer gate wedges fit in place too deeply and did not meet the installa-An extensive plan was developed by the station and ti n t lerance sPecified. The wedges for the new CE to repair the discrepancies. All 20 beam bolts E* * *****
- **'E"*
- ** E I
were to be completely detensioned and then retension-
- E*#
- E" E* * " # * ***
ed, ne procedure was one prniously used at Quad- 'S Cities for jet pump repairs but modified from the tensioning procedure used for the original installa-n e restrainer gate bolt keepers were also sli htly F tion. It included observing the beam bolt for redesigned, ne new design eliminated a shoulder rotatios while being tensioned. The retensioned W lip on the edges of the keeper. Those edges beam bolts were then to be revelded with 2 tacks cat ;ed interference and da= age to the tip of the Per bolt. welding electrodes used to tack the keeper in place. n e lip served no useful purpose and wa.. cli=inated All 20 restrainer gata assechlies were to be replaced to aid in the welding operation. with new assemblies. n ese were to be installed, properly tensioned and restrained, and the keepers Riser braces on all 10 jet pu=p inlet risers were welded with one tack per bolt (total of 40 bolts), inspected with u.iderwater TV. All brace *to vessel wall welds were intact. [ The riser braces on all inlet risers (1 riser /2 jet pumps) were to be completely ~ inspected by using y Jet pump 5 was lif ted ~ 2 in, from its nomal underwater TV. This primarily checked the brace position. The seating surfaces at both the inlet and outlet were inspected with TV and were found to vessel well welds for integrity. in nor=al condition. Pump 6 was lifted frcm its nomal position and raised co=pletely out of the Pumps 5 and 6 were to be removed and raised to a + water for inspection. Both the inlet and outlet position for more cocplete viewing. A complete seating surfaces were nomal. Slight wear on the inspection was to be perfomed. / order of a few.001 in, was ebserved en the belly band in the vicinity of the contact area with the Attempts were to be made to locate and retrieve E** (*9# the missing parts. This included 2 restrainer gate bolt heads and one restrainer gate bolt head.. keeper. N Ref. asy provided additional infor=> tion-Several Additionally, an attempt was to be made to determine design changes were made to the replacement parts the source of the extra wedge bale. to facilitate remote repairs while restoring the pump assemblies to their critinal functional re-All parts retrieved from the vessel were sent tc CE quirements. The basic design remained unchanged. for exams to determine the possible causes or The design changes were incorporated primarily in reasons.for the failures. assecbly features and techniques. Even though there were significant failures of f'l Restrainer gate asse:blies fcr all 20 purps were restraints and fasteners, there were no jet pump replaced with new assemblies. The entire assemblies failures or loss of integrity. Particularly there vere replaced rather than just the failed bolts in were no failures of any jet pt.np holddown compoc ents. crJer to minimize radiation exposur#te persennel. Dus there were no unsafe cenditions existing uuring h restrainer gate vedges for the tci. assta5!!es previous periods of operation. were slightly thicker to assure,,the installation tolerance was maintained. All newly installed The cause of the failed restrainer gate bolt keeper vedges met the specified insta11atien tolcranca. Redesigned restrainer gate bolt keepers were tack welds and the apparent lack of preload on the installed on all 40 restrainer gate bolts. Two tack bears bolts was attributed to constructbn and in-welds per gate bolt keeper vere applied instesJ of stallation deficiencies. De apparent cause of the one veld specified for the os,1ginal assc:bly, to the failed restrainer bolts and the missing and provide a greater margin of integrity. possibly moved restrainer gate wedge was probably excessive vibration as a result of the deficient AstandardrestrainergaterequireYspecialcodifi-installation. The problems discovered were similar cation in order to facilitate inna11: tion on Pump to those discovered during inspections related to 5. The standard gate was cut in 2 at a point 45' the failure described in II.S. It appeared that from the center line of the wedge on the side nearer all jet pump problems experien:ed at Quad-Cities the hinge pin. A spacer block was thtn velded be-could probably be traced to faulty craft installa-tween the 2 parts in order to provide en of fset of tion and worlananship during initial Imta11ation. 1-3/8 in. The offset was necessary to ccepensate for cne side of the jet pu:p hold dem bracket which A previous failure,of flow instru=entation for pump was bent. Because of the very close proximity of the 7 on 11 nit 1 appeared to have a reasonable explanation hinge pin to the inlet mixer of the jet pump, a new and could be related to the failures. nat failure e e.- - e.. s(pt 74 l [ l l VC. 13 - 2
- n
..ver11 cf th* A t t e m
- 11. Rsec trr Int.
i v-tensin,cd te ,et the s.
- p. 12 V
,. r.. ts. { rceut e ents. Fine.!!y.'the fa!!" - \\
- 'r t r t c In.d de sign re r!uireeents v..f d; sign bolt and keeper w6te provided le order to snow er field installation of te bolt and keeper with the (asy i cvrid interference of the I n vol v e c..
ciner. Na ttandard restrainer sate was also modified for An extension arm was installation on Pussp 6. The added to the gate on the hinge pin side. extension arm had a locator screw assembled in it to replace the locator block and adjusting screw that were broken off. The extension was ecsembled to a new desi5n gate bolt and the exis'.ing hinge pin, and was secured by a new design The nut was prevented from rotating by a nut. k;eper which was tack welded to the extension crie. The bolt was kept from rotating via a Right-Hand The locator and 1.ef t-Hand thread design feature. (crew was adjusted to mate with the belly band on the inlet mixer assembly and uck welded in position. N The 2nd 1=cator = crew aa ru=e 6 had been found with a broken tack weld and was not properly positioned This screv was returned to la its locator block. its proper location as indicated by the original e Pu=p 7 tack weld and a new tack weld was applied. was found with one locator screw missing from its i E A new locator screw was designed C sounting block. to facilitate remote assembly and installed, . )[ cdjusted and tack welded in position. Nie== bait recaiaera " r* 'o==d =1*=tas 'r== th-These beam bolt assemblies of Pumps 10,15 and 16. f*, retainers were not replaced since their only (* function is to prevent the beam bolt assembly from falling from the inlet-mixer during assembly or Any future disassembly of these inlet disassembly. mixers will require the beam bolt assemblies be The beas bolts for removed as separate assemblies. c11 20 jet pumps were retensioned using the procedure - developed during the jet pump repairs to t' nit 2 in This required a minimum bolt rotation at a 1972. torque value of 50 ft-lb which provided assurance Following re-the specified tension was applied. tension (ng of the beam bolts, the beam bolt keepers vite sec.ed with 4 tack welds / keeper instead of. the eer Qinal 2 tack welds / keeper to pro-vide a greater tragin of integrity. fallbre cf the flow instr a nt line for Puep 7 NThe was identified as a severed line during this outage. Because of the enginnring and mock-up required to ...f. develop the necessary tools to repair the instri. ment j fix for the severed line was not available line, the It was planned to operate during 7 for that outage. the subsequent cycle with the instrument line as is now, and to try to repair it during their next refueling outage. 1 N The folievins 1=ose part= re==ined la the ve==ett ( Retainer for beam bolt ("L" Shaped piece 1/8 in. g thick, 2 in, wide, one section 2 in. long, one 8 section 3 in. Iong), 2 Bolts for above retainer (3/4 in. Iong x I/2 in. dia), small piece masking tape (4 x 2 in.), l.PRN spring reel guide piece (1 in. dia x 2 in. long), short piece of welding rod (6 in, long), sleeve from refueling shield-cylinder (3/4 in. ( The proba-dia gate (cattle chute),1/2 in, long). bility of flow blockage f rom the pieces was considered small. N The co"*t"'1*" the saa2ysis or the sospectionswas the f ailures were f atigue failures and repairs The vibration was caused by excessive vibration. se sept 74 c) Q(f M. Uhn s 2:. Cy- " " N CER O AS. A?m CM !!. Pen tor Int. F*~ r n LE. ctAtaDtc c4Acc: p. 13 Millstone 1 - Sept 74 - refueling shuto m The exams and analyses indicated the cracking in the vessel shell to nozzle was. limited to the SS See itses 11. 11.- 13 & 16 for details on previour cladding sad resulted from high cycle thermal f atigue information. Tuo sets of N spargers had failed in attributable to excessive bypass flow around the service. The failure of the original Design 1 previous N sparger thermal sleeves. It was believed sparger set was found during the inspection whict this bypass N flow caused rapid temperature fluctua. followed the chloride intrusion.ncident of Sept 72. tions in the affected area, a phenomena which was 8 At that time it was concluded that failure was due expected to be minimised by the interference fit of to high cycle fatigue as a result of inadequate W W sparger Design 4 (installed in Fall 74). The sparger preload. The replacement sparsers, Design interference fit was based upon measurements of the
- 2. were essentially of the original design but were dia of the thermal sleeves and the nozzle bores.
Installed with a much higher cold preload. After his procedure was expected to assure continuous the replacements were operated ~ 6 wk, inspections contact around the circumference of the thera,1 again showed fatigue failures. sleeve, except during intermittent periods of very low N temperature (}00*F) experienced during startup The 3rd set of spargers. Design 3, was installed in Exams prior to assembly did not reveal points of June 73 and was in service until late Aug 74, severe local discontinuities. This operating cond!. These apargers featured a fundamentally different tion, associated with reduced N flow, produces an support bracket attached to the vessel well near the average radial gap of < 0.003 in. As a result of N nosale. Instrumentation installed on the spargers these design improvements virtually no leakage showed excessive vibration when the plant was oper-through the gap was expected during: full power sted at > 85*. power.. Power was subsequently limited operation. The anticipated leakage,at low power to 80L to avoid excessive sparger vibration. levels was not expected to cause significant annular fluid temperature fluctuations in the During the scheduled refueling outage which began thicker regions of the nozzles, because the small in late Aug 74, an inspection of the spargers revealed amount of leakage flow would be healed as it passed that the 2 uninstrumented sparsers and associated through the annulus. support brackets were badly cracked. i r of the bracket cracks was extended to include the cladding cf A fracture mechanics analysis based upon ASME the N nozzle corner radii due to the proximity of Section XI was performed to determine the permissible the support brackets. Several linear indications flaw depth and in no case did the detected cracks were noted in the nozzle area. Therefore, PT of the equal or exceed the permissible flaw. entire clad area adjacent to all N nozzles was performed revealing a total of 23 indications. Con. In order to understand the n~ature and cause of the current with these exams a clad defect removal original N sparger vibration problem, a full scale procedure was prepared. test facility was built in San Jose. A test program was conducted which aided in isolating the cause of In conjunction with the removal of the clad defects, vibration. Design 4, evolved from the informatico a metallurgical boat sample was extracted from one obtained from the test facility. A set of Design of the cracks for analysis by CE. In addition, 4 spargers was installed in Millstone during the Teledyne performed an analytical evaluation of the Tall 74 outage. Following a brief period of con-cracks and ITT Crinnell reviewed the NDT crack re-firmatory test data sequisition during plant startup, c: oval and crack eeasurement procedures and data the unit was to be re6urned to full power operation. fcL A ;uacy. De: cold flow testing in the full scale test facility Th y shrity of the cracks were ~ 1/8 in. In depth, identified the cause of sparger vibration to be the largest being ~ 7/16 in. deep and 2 in. long on excessive clearance between the-sparger thermal the NE W nozzle. The majority of the clad cracks sleeve and the vessel safe-end. For a given pressure were oriented radially outward from the longitudinal drop across the sparger therejape critical flow axis of the W no:zle on the reactor vessel shell to area (or radial clearance for a given size of norrie corner radius. ther.a1 sleeve) which, if etceeded, permits sparger vibrations. The results oFthe test program Indi- ~ All cracks were cechanically removed by grinding in cated the failures of Designs 1 and 2 and the s=all increments followed by PT. The crack lengths excessive vibration of Design 3 were caused by diminished gradually with the depth of grinding, excessive clearance between the themal sleeve and typical of a fatigue type crack. Crack depths were the vessel safe-end. measured by spanning the crack surface with a straight edge from which depth micrometer readings The Design 4 interference fft., concept was proven by were taken. Additionally, wax impressions of 2 cold flow testing in the test' facility, and was cracks were taken from which depth measurements expected to perform satisfactAfily. Verification as well as crack profiles could be determined. of this design was to be accomplished during initial The final ground areas were blended to a 3 to 1 plant startup using data from an Instrumentation taper to eliminate any surface discontinuity, and package installed on the spargers. Based on the a final t'T was performed to ascertain that the cracks test results, the salient design features of the - had been completely removed. In conjunction with Design 4 spargers were.: the grinding restrictions used to assure that the
- 1) Interference fit between the sparger themal Iow-alloy steel base metal was not touched, one of sleeve,and vessel safe end This was based on the deeper ground areas was etched revealing only several tests showing there was no vibration SS clad cetal.
under any flow conditions when the thermal sleeve was tightiv fitted to the safe end. All 4 W nozzles were UT examined from the ext rior
- 2) Forged-welwed tee between the thermal sleeve and of the reactor vessel during the outage and no sparger headers: This reduces peak stress levels I
reportahic indications were detected. ( %* 9 wa== **-. Nov 7 ,y l l P -; m m l ... he r tor im j A res;ew of tha tesett j essura \\essel !tre
- f. pert j
- r. 14 was cc,ndactrd, th3 conclusten being tha: t he :...mn :
l of base r etal removed did not detract fr-tt. l in the tee by a f.tter of 4 due to smaller stress original design requirements. A fractur. rwch-nics ( concentratie... *:ds is due to the use of full analysis indicated the calculated critical fla l penetration we Ic:.. core uniform sections, and depth was not approached. Removal of the cracks l large radil at the junction of the header pipes assured that discontinuities, capable of preparation. I and the thercal sleeve. did not exist at that time, ne cause of the cracking, the planned surveillsace and the corrective action
- 3) Larger flow distribution exit areat his was associated with N sparger Design 4 remained the incorporated to lower the pressure drop at rated same. I=plemeotation of those measures was expected flow from 17 psi for the first 3 designs to 10 ps!
to assare a safe return to full power operation. l for Design 4, which increases the stability margin. (bbr)
- 4) Sparger installed preload of ~ 2600 lbs This was incorporated to iohibit sparger radial motion due to W flow. This was accomplished by installing the sparger with a preload that exceeds max expected W flow forces.
Erch Design 4 sparger was provided with vibration instrumentation. Only one sparger (SW) was to be fully instrumented with numerous strain gauges and RTD's. Due to the i=portance of detecting local temperature fluctuations in the region of the nozzle to vessel shell radius. RTD's we.re installed on all 4 N nozzles prior to plant start-up. Data taken from thrse sensors was to be conitored and recorded in a m [ ctnner similar to the vibration monitoring planned far the hot flow test program which would corres- [ pond with the plant's ascent to power following the cutage. ~ IJT of all 4 N nozzle corners from the reactor vessel extarior were planned af ter ~ 6 mo of operation. All indications Jetected at that time were to be satis-fcctorily dispositioned prior to return to power. During their next refueling outage, UT and PT was pituned on the accessible nozzle corner cladding of all 4 W nozzles conc. ent with an exam of the spargers. 11T, identica: to that performed at the 6 ma interval was also to be conducted during that outage. (axp) Ref (bbr) provided additional information. Originally thsy concluded that they had 23 cracks, none of which penetrated into the base netal. A revaluation of the duply ground aress was planned and conducted sub-sequeut to tM al reload operations. This was necessary tu -.a all deeply grcund areas had not bun chemica!!y etched and the apparent clad thickness was substantially in excess of that specified in the original den 16n 0%u ents. Following vessel flooding end subsequent te the fuel reload operations, the R21. N srincta caxcxs water within the reac::: vessel cavity was drained to a level just belo. the W spargers and the vessel R Mahleberg - Aug 73 and su=:rer 14,74 - refueling work platform was insta!!ed. A visual exam of the g; shutdown previously ground areas adjacent to the 4 W nozzles revsaled 3 areas of rust (ferric oxide). The 3 areas n ey inspected their spargers after-the discosery of wsre on the nozzle radii of the SE, NW and NE nozzles. cracks at Millstone. Some indications of very small This represented exposure of the low-alloy steel cracks were found at that time but continuation of base metal by the crack removal operations. operation was still conside. red practical. A close visual inspection with the aid of a stone operated The reference also stated that the correct number of chair showed larger cracks in 2 of the7 spargers in detseted cracks was 20. the su=mer of 1974 In a close collabgestion between CETSCO and BW the replacement progras was very Subsequent to the visual exams, PT was performed on carefully prepared. It involved among many things the the NE N nozzle grind area, t!T of clad thickriess was fabrication of a shleided platform and shielding performed at ~ 35 locations adjacent to the deeply curtains for the vessel walls in order to reduce the ground areas, 2 replicas of each rusted area were strong radiation (1 R/hr) in the working area. ncds and depth ceasurerents of the base metal pene-ne 4 W spar,ers were fabricated before the outage tration were perforned, in addition Teledyne e parforned an analytical evaluation of this base (they had to be modified slightly). The whole k natal penetration and ITT Crinnell exanined the ruated replacement work was performed by a tes:s of 27 sreas and advised with respect to replica formation specialists of Pannesmann and BG (the owner) and procedures. CETSCO supervisory personnel. m- - e --== - e. =- Aug 76 (It. '. t it..) In 54pt ar.d U:t h, a preparatten ~ogram for m. rsplacing the m rpts vaa carris at by CETSCO (.orrec tive rcp. - I sere pe r s..- e.: utir u in,. (nd Muhlsbar; persenn=1. It involvsd p1tnning the specially disignid to:,le. A1: in " tai t t :. vel necessary operations, reworking the previously associated r.uts were re oved !:. - tr+ appor t to a kts. ; ordered new spargers, f abrication of a shielded There were no lost or unacceunt d int part:.. I;...! c platform and shield curtains for the vessel walls tools used were a ec :rercial hvdr ulle. teni t cutter and employment of some 25 skilled men. and nut cracker codalled for the app!! cation. The sparger was then recentered unin; a cont-ercial Shutting down the plant, opening the vessel and hydraulic jack with special fixtures built at the installing the platform and shielding was completed plant so the sparger could be jacked against 2 of within 2 1/2 days. Work started in Oct with the the sparger support brackets. The force required removal of the old spargers. One of them had a to move the sparger back in place was calculated to circumferential crack. Indications of slight be < 2000 lb. cracking were also observed on the 2 sparsers which were still sound in the sunner inspection. Af ter the sparger was recentered, the N line inside of the drywell was examined by PT. No relevant The installation of the new sparsers involved a indications were _found. The N nozzle was also careful adjustment of the thermal sleeves to the examined by radiography. This showed that the thermal die of the W nossles. The gap between nozzle sleeve which attaches to the sparger was in its proper and sleeve was determined as a critical parameter position. The end of the thermal sleeve appeared to for the possibility of vibrations. Detectors for be normal in all respects. sound emission were installed at the outside of the nozzles. Redesigned sparger restraints were installed at each support bracket. These restraints were sub-As planned the whole operation was finished after stantially stronger in design and clamped tightly 6 days. The accumulated exposure of all people around the sparger. They were not directly attached involved was 44 man-rem. 1he duration of the to the sparger support brackets as wese the original R *h"td "a "== ~ 2 t/2 "*- (b=b. bra.6 s.htk) U-bolts. The nuts which bolt the restraints to the 9 sparger were tightened to 40 ft Ib anfsecured by staking to prevent them from working loose. Legs 22. VIBRATION - W SPARCER U-BOLT FAILURES from the restraints were designed to tngage the wall bracket to limit any sparger movement away from the Humboldt Bay 3 - Nov 74 - shutdown nozzle due to hydraulic forces, while allowing the sparger to cove as a result of thermal growth. The reactor N sparger consists of a continuous section of oval cross section pipe located around The new restraint design was not expected to be the inside of the vessel at an elevation below the susceptible to the same types of vibration problems top of the core chic:ney. The sparger is supplied postulated for the original U-bolt design because the with wathr from one inlet nozzle. It has 1/2 in. new restraints were held in firm contact with the holes near the bottom on both sides to distribute sparger. The new restraints were fabricated in the the IV uniformly in the vessel. The sparger is plant from solution annealed ASTM A-276 Type 304 SS supported by 8 gusset type brackets around the ID of bar and flat stock, the vessel wall and was originally designed.to be ' held in place with U-bolts mounted on each bracket. Twenty-three days elapsed from the time the sparger The U-bolts, which were fabricated from 1/2 in. problem was first discovered until the corrective dia=eter SA-276' Type 304 austenetic SS, fitted around repairs were completed. All of the repair work the sparger and were held in place with nuts on was performed by plant personnel working from the both sides of the support bracket. The nuts were refueling platform on the reactor extension tank to tack welded to the bracket. minimize radiation exposure. The total accumulated - ~ exposure was ~ 6.5 man-rem. All special tools and During a refueling outage, it was noted that the the new restraints were built by plant personnel. sparger had moved 1 1/2 to 2 in. sway from the nozzle Engineering support was provided by CE. and some of the U-bolts were broken. A close exam showed that 6 of the l'-bolts had broken through on During the 1975 and 1976 refuelifa,n* mintenance one or both legs. Tvc bolts were sent to CE for outages a th: rough visual inspection of the W metallographic exams. These bolts were broken sparger,allsupportbracketsjands11spargerre-through on both legs just above the nut on the top straints was to be made. In addition, each restraint of the support bracket. was to be checked to assure it was still clamped tightly to the sparger. During the 76 outage one The spargsr thermal sleeve still extended ~ 22.6 in. of the restraintr was to be removed and replaced by i into the nozzle. The sleeve had an interference fit a new restraint of similar design. The r moved and the nozzle ID is uniform so there would not have restraint was to be subjected 4o a thereugh exam. been an increase in W flow around the sleeve. 9 (bes) j JF CE's exam procedures consisted of visual, FT, scanning i electron microscopy, metallograph'y and hardness. The exams showed the failures were caused by high cycle fatigue. It was postulated that the U-bolts vibrated in resonance with input from a source such as the N pump or rectre flow. It was further postulated that .e the sparger itself was not vibrating or that the vibration amplitude was very low. - e - - e. -- - Feb 77 .- ) s s Vel. %"r-2 II. betar Int. f-1,--i fvery simi.' to the iSan'r*
- p. It
- '.. : s t ew ). 4 ) inspu the !*
Le nt:: trter ni:xt scheduled refueling rmt ',, inspect 23. W SPARCER CPX r* (nacluding FT of the nozzle blerd f.m '. :S [ tiresden 3 W spargers during its nemt : W aled ( Dresden 2 - Dec 74 = refueling shutdos. refueling outage. W sparger (304 SS, Sched 40) inspection was per. The salient design features o' the Destge 4 W fErined in response to a CE reconunendation. The sparger were: impaction consistec. sf using underwater TV to cxamine areas including: 1) welds of 6 in schedule
- 1) Interference fit between the sparger th6rsal 40 header pipes to the 11 in, junction box pipe, sleeve and vessel safe end.
- 2) Forged-welded tee between the thernal sleeve engrgement with clavis ends of each sparger and and the sparger headers.
- 5) elaLed W nossle blend radii area on the racetor vessel. D e inspection revealed 2 cracks.
- 3) Different sine and location of exit holes in The first crack was located on the upper part of the sparger.
the right side header pipe to junction box weld cree of the south-west quadrant sparger. The
- 4) Schedule 80 rather than schedule 40 (304 SS) crack appeared to be relatively straight, extending spargers.
(bfn) l 90* around the pipe circumference. He 2nd crack w:s located on the upper part of the left side N nere were 200 men used to replace the heIder pipe to junction box weld area of the north-spargers. (cyc) ecst quadrant header. This crack appeared to be a little more jagged than the first crack, near the weld area and extended ~ 200* around the pipe y; circumference. r L' Ths bearing bars (preload spacer), on the right i sids of the south-east quadrant sparger and on the right side of the north-west quadrant sparger were est in contact with the vessel wall. No other discrepancies were ' observed. Bassd on the underwater TV inspection, the predoninate failure mode appeared to be fatigue cracking. De vidio tapes of the sparger inspection were viewed by /" \\* CE personnel and concensus of opinion was the cracks locked very similar to the sparger cracks at Millstone 1 (see II.20). The cracked spargers at Millstone vare mistallurgically examined and deternined to be trtnsgranular (fatigue cracks). Aa viewed through TV, the bearing bar on each of 2 epargers appeared to be out of contact with the vessel well by a slight amount. Although there vara severa1Jesde explanations for this, the actual cause ns isot knotas. This condition had been f observed at uther :"JR's and its presence did not altsr the conclurion that the sparger czacks were 4_,. crussd by fatigue due to flow induced vibrations. y,w Numsrous full-scale cold-flow test had been conducted at a CE test facility on W spargers of several diffsrent configurations to determine the cause of .y vibration. Rose tests showed that unstable flow-incuced vibration occurred as a function of the fol- / loving variables -Ths op between the sparger inlet and discharge. -Ths average radial gap, which permits leakage flev between the inside of the thermal sleeve snd the outside of the W nozzle. -Ths amount of damping present in the system, particularly at the thermal sleeve-to-nozzle l interface. Ds program for corrective action was to 1) PT i ths accessible portion of the nozzle blend radius of each N nozzle, 2) further inspect the old Dresden 2 spargers and determine the actual radial grp when they were renoved 3) replace all 4 Dresden ) 2 spargers that outage with new spargers of the i e e,= * - - ' ~ Feb 77 y e --,----.-,n m., .-,-n, ,,-n
- t - Int.
U..
- .. l
~ '" m ) -1 ( - h t, '$ - retving An extensiv3 inspsetton of.all t te t I ht p ps durint. its first refueling outage an Apr N. revealed a l ar gte j 3 v. A schecLlr.' b. t pu.p inspection was perfntt include d numbcr of jet pump discrepanc ia.. - Treu s the aid et t'enrtvater TV and video tape. Ifpr. e - bea:n bolt torque test f ailures, she are <t restrainer pletion af inspections on all 20 jex pumps. tr: gate bolts and keepers, missing and cracked restrainer, following discrepancies were notedt 1) The he.e Fate bolt keeper tack welds, a missing restrainer bolt retainer clips and 1/2 in. cap screws were fc ec gate wedge and indications of wear in the vicinity to be missint on jet pueos 7 and 8; 2) The restrasner of the wedge indicating possible vertical movement adjusting screw on the shroud side of jet pum 15 of the wedge prior to the inspections. was missingt and 3) Restrainer gate keepers on jet pumpa 1, 6, 12, 15 and 18 had their velds in plact, All of the jet pump problems which had occurred at but not fused to the restrainer gate. Quad-Cities were attributed to f aulty craf t installa-tion and worlananship during the initial construction The apparent cause of the missing retainer clips, (bhp) of both units. the 1/2 in, cap screws, and the adjusting screw was attributed to equipment failures caused by At the end af the outage a neutron dosimeter tuba, installation deficiencies. Based on the parts a neutron dosimeter, an adjusting stop screw and a being inadeouately installed, the vibrational grinding burr were still missing and assumed as forces present during nor nal operation could have (but) caused those components to become dislodged fro
- remaining in the vessel.
their normal positions. Iksring jet pu.p repairs to Unit 2 in 1972 (see 11.8), the beam bolts and keepers were the primary con-ponents repaired. The repairs necessary to Unit 1 4
- during its past refueling outage in 1974 revealed that further defective installations (see 11.19) 1 existed. It was felt that their then present con-I ponents were originally installed in a deficient manoer and the possibility of their failure was 25.
IV SPARCER CRACKS not known at the time of the original Unit 2 repairs. The discovery of their new failures was coisistent Quad-Cities 2 - Feb 75 - re fueling shutdown with the deficiencies noted during the Unit 1 out-PT of the 4 W Sparger Junction Boxes revealed indi-age in 1974. cations of cracking. The test was being performed The missing beam bolt retainer clips and 1/2 in. upon request of CE because of concern over the occurrence of cracking in several other W Spargers cap screws of pumps 7 and 8 were not replaced. of the design utilized in Quad-Citits. SPeclat These pieces sere important only during the initial attention had been given to the welds at the inter-construction of the jet pump beam assembly or daring their removal. Af?er insrs11ation of the beam bolt, sections of the juretion boxes and the distribution the function of those pieces was completed. pipes. The pT tests revealed indications of several cracks in the heat affected zones on the piping side A replacement stop (adjusting) screw was installed of the welds on each of the 4 spargers. Following and tack welded to its holding clamp on pump 18. removal of the old sparsers, further PT exams were conducted on the FW nozzle cladding with the following As an alternative to removal of the jet pump throat, the veldinc tool used to tack weld the stop screw resultst 1) On the 60' norale, 3 porosity indica-tions were found, 2) On the 150' nozzle. one fino was .odi~'f. to perform its intended function. crack,. 3 linear indications, I crater and scattered a porosity indications were found, 3) On the'240' Alth n::- !. was postulated that the restrainer gate nozzle, one linear indication was found, and 4) keep =i Tn pu p 16 was loosened during the repair On the 330' nostle, 8 linear indications were found. of the'stop screw, the tack welds of all jet pr p All relevant indications were removed _by grinding restrainct gates were reinspected as a precautionary As a result of the faulty tack welds die-and retested satisfactorily. F reasurt. / covered on the jet pump restrainer gate keenrs of pe ps 1, 5,12,15, and 18, the restrainer tatr The failures were attributedfo fatigue to the sparsers bolts were retorqued, and new tack welds were placed caused by flow induced vibration, and compounded by and verified. A similar inspection of the 20 jet stresses induced by the thermal gradients inherent purp beam bolt keepers was performed with satis-between the FW piping and reactor vessel internals. There have been 36 startup/ shutdown cycles. I.eakage factory results. between the sparger and the W notric contributed. As of late Feb 75 there were 11 objects (dostreter, significantly to vibration of.the sparger assecbly and screws, clips, vire, vrench, rag) loose or nissing also imposed thermal stresses 'on the nozzle. In the reactor vessel. Search and retrieval atte pts p The failures were similar to tJiose at M111stsne and were continuing. Dresden 2 (see 11.23, 32 and 33). Spargers of a new The problens discovered on this inspection of Unit 2 design were installed in Quad-Cities 2. They utilized were sicilar but icss severe than those discovered an interference fit to eliminate IcaLage and thus follocing a jet pump f ailure in t' nit 2 in Aug 72. reduce the vibration. The old spargers wcre to be At that ti :c, a jet pump assembly had beconc dis-further inspected to determine the ac tu I radial gap that existed during operation. The spargers for tinit 1 lodgtd fro-its normal position and rotated in the vessel. Extensive inspections and repairs verc pet. were to be inspected during its next scheduled re-forned on all l' nit 2 jet pumps and they operated f ueling outage. The newly installed spargers for ( Unit 2 were to be inspected during its next scheduled satisfactorily thereafter. refueling outage in Sept 76. (bho,cmx) c o e ~ a. Av,76 l l
- u '. '.*. j l
t. c:m ryonss regler 7 Arn.lf I 1*.a.e : (; eta notts. Plua,ging L. j hnle: w - <-l
- p. I' slightly affect steady-state ener u.
2. candttio'is. The plugs mentioned M. y re u :11ar to those pseviously installed in rita rin and Vt r ont g Yankee (see I.12). CE determined th..e the natural i frequency of the LPRM tubes was 2.5 lir. Anong the permanent fixes CE was considering were clam to 1 hold the TIP tubes stable, diffuser or hubbler ) pipes to be inserted in the holes in the core plate, drilling 2 holes in the tie plates (transition pieces) at the bottom of the TA's, etc. Fukushima 2 During Dec 74, an unusual traversing in-core probe (TIP) trace was noted. Evaluation of this trace and subsequent diagnostic tests led to the conclusion there was a high probability that a neighboring channel had developed a hole. The reactor showed 11 to 121 peak-to-peak noise in Feb 75. Prior to the reactor shutdown, a 2nd location developed similar noise characteristics.on its TIP traces. Inspection of the channels surrounding those locations at a subsequent outage confirmed the hypothesis. In order to assess the probable cause and amount of 26. INSTRUMENT Vf BRATION - CHAWNEL BOX DAMACE damage, a detailed inspection plan was develo* ped and irept emented. A total of 197 channels received a Fukushima 2 - Spring 75 detailed visual inspection. 1he results of Mais Cooper - Apr 75 inspeccion are given below. [ mhlet, erg - 1975 Pilgrim - 1975 Est. Wear (mits) Brunswick 2 - 1975 Type of Number Through Wall Fit 2 Patrick - 1975 Arnold - 1975 Incore Inspected Wear 30-70 10-30 <10 Peach Bottom 2 & 3 - 1975 1.FRM 124 3 39 50 32 Hatch 1 - 1975 source 20 8 9 3 R srowns recry 1 4 2 - 1975 sRM 16 1. 6 c Vermont Yankee '1975 1RM 32 3 11 18 (* 5 None 5 CE reported that TIP systems had provided indications of possible vibrations at Fukushima 2 and at Cooper. Total I97 3 31 76 62 Prelicinary exa s by CE indicated that the coolant flow through byp**s holes in the piste supporting ths reactor cor.ould be inducing movement of the Cooper instrument compwn..ts sufficient to contact the ctruers of the channel boxes, causing wear and Special instrument checks indicated possible vibrations crtcht"g* 5.uA war and cracking were observed in and they reduced power to < $0% in late Apr 75. Fukushima 2;^ CLael boxes are structures around Essentially no indications of " noise" or anomalous bundles of i--t rods that help smooth the flow of readings were detected at a 507. power, 44". flow condi-watsr past r ,tr.J1es and through the core. tion. They elected to run at 501. throuth,the sucrner load period until oct 75 and then shutdown for a Sicitar bel..Ivr -but'of lesser reagnitude than noted fix. Their refueling was not scheduled an'til 1976. in Fukushi=a 2 was identified in monitoring results In July the NRC permitted them to incr' ease power to at Cooper. Cenar decided to reduce power and flow 607, and reduce flow to 407.. to < 50*.. The :.?.2 directed 10 other BWR's (Arnold, d Vstnont Yanke*, Erowns Ferry 1 and 2, peach Bottom T 2 tnd 3. Pilgrim 1. Brunswick 2, Hatch 1 and Fits-Arnold patrick) which had bypass holes in the core plate to review results from TIP monitoring performed TIP traces indicated 3 possible locations that had within the last 3 eo to identJfy any anomalous be-higher than average noise levels. Only one location hivist, and to provide the NRC with the results of had noise that was not clearly less ihan reporting th2 review. A screwhat similar problem is described criteria establishe6 in Apr 75 by theNRC. The TIP in 1.12. (bjv} traces at Arnold contained less noise Ahan those taken at Fukushima 2. Tha problen appeared to be generic to the BWR-4 class of reactors only. The BVR-3 class did not have lower In mid-May the NRC ordered them to reduce, power to care support plate orifices. R 5 0 ~ ( * ' c a ' t " ' ' ' 5 5: '"' " ' "' " ""' observed). The unit was shutdown in early June for Tha bypass holes were dril' led in the core plates to an inspection. The shutdown was preceded by a 72 he maintain the total bypass flow at ~ 107, of the total test run at power levels and core flo. rates up to ( core flow because finger springs (FA's were equipped 10G2. They planned to replace all channels which I with finger sprir.gs on lower core tie plates to provide experienced > 10 mils wear in the lower half or a crntrolled flow leakage path between the lower tie > 21 mils wear in the upper half. These limits were plate and the fuel channel) reduced the flow in consistent with the allowable on-site handling scratches or wear that any channel could experience 1 1 July 80 a -., - -. = a = '. l. P. -. durine w M fuel hsndling opsrations ane allo. -p
- 11. k. m t <.: li.t.
of additiona! duty. In pection of the first 4 ch e nt!
- g. l 'a bones (adjacent to thimble location with high noin.
mentioned above) showed unacceptable wear in the corners various types of incore in=trw ent? was the same as ( adjacent to the instrument thimble. ihe wear was as at other reactors. That i,. the total distribution much as 40 m!!s in depth with an additional 8 to 10 was bimodal and separable on the basis of bypass atts in several notches in the general wear region. holes being present or absent. Of the instruments The width of the wear region was as much as 3/8 to having bypass holes present. the 3 hole locations 1/2 in. The wear was not as serious as that at led the other locations for the cagnitude of channel Fukushima 2. Altogether, about 1/2 of the channels wear. CT had the following 4 inspection result adjacent to instrument rods showed wear. categories rejectable, probably rejectable, probably acceptable and acceptable. This number of categories They plugged the bypass holes and replaced 119 anticipated selective channel placement in the core channels. One FA was dropped during the handling if the availability of channels became lia,iting, operations. The length of the outage was ~ 6 wk. Nine channels had " double wear" and 6 of them were, Plugging the channels reduced the plant capacity by definitely rejectable. " Double wear" was a light - 151. The NRC then authorised them to operate up surface wear which occurred on the channel box ,to 851 power. In-core instrumentation and accelero-flats adjacent to the worn corner and beyond the wear meter data obtained after startup indicated there where channel material had been removed. At least was no impacting of instrument tubes against channels. one channel had thru-wall wear. Modifications were authorized in Mar 76 which approved the drilling of the FA lower tie plates of Types 2 and 3 FAs to provide bypass flow. Holes in the lower core support Peach Bottom 2 & 3 plate orginally provided for bypass flow were to be plugged to eliminate in-core instrument, tube vibraticns. Unit 3 vss derated to 301 power in early June following Units 1 and 2 were not authorized for operation untfi indications of possible TIP vibrations. Unit 2 had a later safety evaluation was cocpletedf/ CE felt shutdown for local leak rate testing on May 17. After that drilling new holes in the lower tia plate assembly it returned to the line in June it proceeded to would rearrange the coolant flow through the core 1001 power with continuous monitoring for vibra-sufficiently enough to allow operation at 1001 tions. It was restricted to 50% power af ter July 6. power and still prevent instrument tube vibratf on. Both units were allowed to so to 60% power, 401 flow in late July. Plans were to operate the units at rated power for short periods as part of the fuel Vermont Yankee preconditioning program. Accelerometers were installed on TIP detector drive cables. Prelimin-This unit was also able to operate at 80-85% (see ary results indicated there was no contact at < Hatch 1 above). They shutdown.for 540 hr in Aug 60% flow. PECO then asked for permission to operate and plugged holes. There was some wear. up to 501 flow and 3.62 MWt bundle power. In-core fixes were, performed on Unit 2 in Nov 75 and on Unit 3 in Jan 76. Fit: Patrick underwent in-core fixes in Feb 76 and Brunswick 2 plugged bypass holes in Mar 76. Pilgrim had already plugged bypass holes. H_suh I (buj,buk cne.cnf,cng,cnh cni) n is unit was able to operate at ~ 80-851 power while Anttoring for variations in TIP flux signals. Nsuhteber In letc Aug they proceeded to 100% recire flow (~ 951&os n) to obtain TIP traces. The TIP values of The fuel channels were affected during the first 2 the ratio of noise band width to signal amplitude cycles by vibrating poison curtains. - This led to a did not.=ceed 6% over any 10 in. of radial core temporary power reductica and subsequent outage 1ength and they planned to resume their 100% startup (~ 461 hr) to remove sete cf the.cartains and plug testing program to prepare the unit for cor=ercial bypass flow holes. Af ter re opl of the-last operations. (Holding the variation in flux signal curtains during the su rer 74 new problem aroset channels we'gtfueling outage, a below 61 allayed b'AC concerns about possible chafing re damaged due to and wear from vibration.) During a Nov outage, 192 vibrating incore detecter tubes. This phenomena fuel channels which were adjacent to the 48 in-core was hidden **behind the curtains" during the first j tubes were found to have no through wall perforations. 2 yr because the poison curtains absorted most of the However, 125 channels were rejected and removed from energy induced by the leakage flow through bypass service due to significant wear. New channels of holes in the botto= plate. Inspection of all affee-the same design replaced those rejected. There were ted charnels during the su : er 74 outage showed 64 channels identified as exhibiting minor wear excessive wear (not through holegl on 21 channels. and they were relocated in the core in such a They were replaced and the new cfEle was started manner that worn areas would not be adjacent to in-until a final fix could be determined. core tubes. Three channels were found to have no visible wetr and were to be returned to their original NIn early spring 76, preparations uere made for a locations in the reactor core. D e bypass flow holes final solution. It censisted of plugging the bypass in the lower core plate were also plugged during holes in the core plate and drilling 2 small holes the outage. Continued monitoring and inspection for in the fuel bottom pletes. The drilling technique vibration induced damage was planned. used was a electro-erosien-technique. Tests had been performed at CE and at W hleberg and the results were j promising. ne re=aining 160 fuel elements were Browns Ferry 1 drilled in this manner during the June 76 refueling outage. Af ter some cinor start-up problems, all 320 Inspection of all 248 channel boxes adjacent to incore instrurentation was completed in early Aug. The magnitude of the wear distributed among the 77 -g ,r g.y., _,.-,,,,,,.e., -g,,_ -m_.,,w.-m,., +s--,-- y1 .s. s e 4 (
- aM I" o f ths W nb-
.r:er-wn* planned hl. I II. heactor ht. rer lept 7 t,. Thsy hed a tra startup/shutdtwn
- p. 20
}cvelesa*ofNav 76. About - -" were used in the sonrrer rentacement. ( tm.e -.- .c_x,cyc) ( h21ss were eroded within 16 days (the 80 new element. altsady had the holes. 28. IMPROPER TACK WELDIN:" - TyrrN k'ZLDS 04 JET PUMP _S N5a==d i= = ta= =* **======== dar2=s =*ar* =P in July Dresden 3 - July 75 - refueling shutdown showed that the vibrations were much smaller than in the preceding cycle and that probably no hitting of !.During restrainer keeper lift terting of the jet thm channels was occurring. The TIP traces also pumps it was noted that several tack welds had j indicated a substantial improvement. A fuel channel parted on the outboard clamp bolt keepers for Mos. ' lacpection was planned for the 1977 summer outage. 6 and 17 jet pumps. A visual exa= by underwater j (hrv,cyq) TV had revealed no visible defects on either jet pump keepert however, during the lift test, which Ses item II.35 for additional indormation. ascertains the resilience of keeper and weld, the welds broke. The apparent cause of both keeper failures was poor tack welding. Properly made, the welds would normally hold the keeper securely to the jet pump restrainer. i Tailure of both inboard and outboard clamp bolt keepers could conceivably allow the clamp bolts 27. W SPARCER. NOZZLE CRACKS e Dresden 3 - May 75 - refueling shutdown . to back out of the jet pump wedge and restrainer assembly. A double failure of this nature occurring during reactor operation would result in increased vibration of the jet pump and might ultimatel f's11p Spicial visual and PT inspections rexealed cracks leak to separation of the jet pu=p body at the in the F% spargers on 5 of the 8 header pipe to junction box welds and on 1 of 4 thermal sleeve to fitting between the mixer and diffuser sections. junction box welds. As with the other plants, the crtcking was believed to be due to the desi$n of the The tack weld failures were induced by the artificial rpirgers which rendered them susceptible to flow-vibration of the lift testing. The jet pumps were induced vibration, securely fastened by means of the inboard clamp bolts, both of which were tested satisfacorily. Thi program for corrective action was to 1) PT in any event, the routine daily surveillance of jet examine the accessible portion of the nozzle blend pump flow characteristics would detect any such g rsdius of each W nozzle, 2) replace all 4 spargers discrepancy and the reactor would be shut down within (' that outage with new spargers to the new design 24 hr. (vary similar to, the " Design 4" sparsers), and 3) inspect the W spargers during their next scheduled The clamp bolt keepers were revelded ar.d successfully refueling outage. The Pr of the W nozzle blend retestad. Deze jet pumps were to be inspected redii resulted in a total of ~ 150 linear indcations during the next major refueling outage to verify the bsing found. The indications were oriented in the . integrity of the tack welds. (bse) direction of the nozzle bore centerline. The max length was 1 1/2 in., with avg length ~ 3/8 in. All linear inJh ations were removed by grinding. 29. CRACKED W SPARCER Tha max cavity depth from the clad surf ace was 3/S in. g 33 max pengt4 ; tion into base metal was 1/8 in. Oskarshamn 1 - Aug 74 Crinding waGne in 1/16 in. increments and af ter (. ~ etch incree-at, PT was conducted to verify whether The W sparger was made in one single piece. Cracks ths indication had been removed. The cross - were discovered during the su=ar 1974 shistdown. De sectional area of reinforcement removed at the replacement sparger was made in segment's7 The work vorst grind location was.144 in2 in the radioactive environment was very ti=e consuming. A special lead container from which the work could be Tha salient design features of the " Design 4" , conducted was constructed. This extended an outage sparger were: 1) interference fit between the sparger &~by ~ 4 1/2 me. See item II.38 for additional thtraal sleeve and vessel safe end 2) Torged-velded information. (brv,' bto) tee between the thermal sleeve and the sparger headers,
- 3) different size and location of exit holes in the sparger, and 4) schedule 80 rather than schedule 40 (304 SS) spargers.
N n i first design feature was based on several tests showing there was no vibration under any flow condi-l tians when the thermal sleeve was tightly fitted to th2 safe end (small radial gap). n e 2nd feature reduced peak stress levels in the tee by a factor of 4 due to smaller stress concentrations. This was dus to the use of full penetration welds, more uni. fera sections, and large radii et the junction of the ( hstder pipes and the thermal sleeve. ni 3rd feature was incorporated to lower the pressure drap at rated flow from 16 psi for the first designs to 11 psi for the new designs which increased the ~ stability margin. .~ m o - a e= Jan 77 M. F-b 2 3, . r. n
- m tr e. -
- . *4a:ter int.
\\ PM U.:.*_ .a.1:.; Evaluation of the 'results of th. In vection produced r!!,i t? I - Feb 76 - refueling shut.N.- several conclusions l Based on the reconrnendations of CE NDT of "- reacter vessel W spargers and nozzle inr.c-b l e r.* .CE was concerned that the tack weld failure could possibly indicate a vibration problem that a visual radi-was perforised to verify their functior..;. check that had excluded the restrainer wedges could interrity as part of the NSSS. VT of the P.' spars ers was perfomed and no anomalies were observed. not adequately evaluate. FT of the accessible areas of the reactor vessel W -Yankee engineering supported CE's position and nozzle inner blend radii was then observed. Therefore, strongly reconnended a completion of the exam in-the W sparsers were removed from the reactor vessel cluding the mechanical tension test. and as-built dimensions were obtained so that new -The potential hazard presented, in and of themselves, spargers could be fabricated. Concurrent with by the 2 small screws, should they become loose in fabrication of the new W nozzles, all of the fuel was the vessel, was not significant. removed from the reactor vessel, and PT was performed on the inner blend radii of all of the reactor vessel ,A CE engineer experienced with the exam and the jet FV nozzles. This r. vested a total of 78 indications pump failure at another facility was engaged to on the W nozzles 1 aer blend radii surf ace which supervise the 2nd inspection. The inspection was extended from a 12 in. reference circle around each completely repeated on jet pumps No 1, 6 and 15 nozzle into the nozzle to ~ 2.5 in. past the back and the remaining portions of the inspection were edge of the thetical sleeve centering tab. Fracture completed on the others. Nothing else of an unusual machanica analysis was performed and it was de-nature was detected during the inspection. De termined that the most critical region was the W suspected cracks could not be identified. (car) nozzle corner. Corrective action consiste 1 of removing all 32. ADDITIONAL W N0Z21.E CRACKS unacceptable indications by grinding. This repair i work was performed in accordance with approved Millstone 1 - Sept 75 - refueling / shutdown F procedures and engineering instructions under the technical direction of CE personnel. As.left grinding See item II. 20 for previous information. Iuring repairs were determined by use of etching, dye marking, reexams of the nozzle corners similar FT indications and colding to provide an accurate sizing of all were found. The number of cracks found in the 2nd repair grinding which penetrated the carbon steel exam (lyr after the first cracks discovered had base metal material for use in an engineerins been ground out) was greater than the number found evaluation. The max grinding penetration into the in the original n am. The depth of the cracks was reactor vessel base metal was 0.5 in. about the same. The max depth of the cracks found was ~.5 in. D e max depth of penetration into To prevent recurrence, new W spargers were installed base metal was ~.25 in. Less than 50 of the cracks which incorporated an interference fit with thereal found had penetrated the base metal. The cracks were expansien characteristics similar to the surro = ding difficult to detect, even using PT because they were nozzle to reduce the W sparger thermal sleeve bypass easily plugged by oxide. UT from the external nozzle flow, therey, minimizing flow induced vibrations circumference had not yet shown to be a viable and ther:;'. cycling. alternative. The W spargers had been inspected during the 1074 The cause of the cracking was still believed to be refueling outage and no anomalous indications were ' temperature fluctuations in the nczzle area. Because oburg.M of leakage around the thermal sleeve, the blend radius area was exposed to temperature swings which a Sr g f p. .he evaluations made by Teledyne, it was ranged from the hot saturation temperature of core f m ":". c the margin of safety for the structural water to the relatively cold entering W temperature. A frequency of I Hz could be reached during those inte;.rit; of the W nozzles was being maintained
- 1 required by the original acceptance stan-swings. Although there was some uncertainty that the at a ::
dards. It was anticipated that a visual reins;ection cracking was due to leakage flowMhe proposed fix" of the blend radit was to be conducted via uni.er-was based upon the assumption:that it was the cause. water camers during their next refuel'ng outage. CE believed that the nozzle ana could tolerate, N S** l= xvi. C. 247 for additional infor=atior.. without catastrophic fatture, a through.uati flaw (chl. chm.c=i) under 1000 psig pressure and 100*F N temp. 31. LOOSE SCREWS - FAILED JET PUMP TACl* k*ELOS. CE's reconsnended interia exam program was based upon the determination of the number of cycles to reach SUSPECTED 1/10 of the minimum metal pathhetween the point Vermont Yankee - Fall 1974 - refuelina shutdown. at which the nozzle is tangent tb the blend radius and the exterior surface of theJDV. CE believed During sisual inspection of the jet pumps, using that a. welded thermal sleeve and'sparger, with a borescope, 2 apparent discrepancies were fcund. A resultant zero bypass leakage, would be the solution lock plate screw on jet pump No.15 and a 1/2 in. to the problem. retainer cap screw on jet pump No. 6 each appeared to exhibit failed tack welds. The vessel cavity was An estimated maintenance. time su=sry was submitted reflooded i.,ediately af ter the inspection, which was applicable to all Bk'Rs with the sparger/ nozzle problem. An interference fit sparger design repair would require ~ 12 days and ~ 530 man-rem of exposure whereas replacement would require ~ 10 days e--ea Dec 29 .------iw nm s 4
- [.'.
T E ' P.ht um pre s su r ira t i n.- m e -- s .f.. rre c lw brittle failure with a pr..' t ni m. n wall fra an1 c. . -reri of eurosure. The welded-in sparrtr were calculated for reae to v. a ; pre suren ir-hwe. 51 reautre - 70 days and 2000 man-rem cluding inservice hydro testin-w ures. The ( of exrm ure. Mend rad!! indication grind-out more conservative of the min!ye pr m uriration vnule ree:.ste - 8 to 20 days and - 390 man-rem of temps determined from the c alc ul..t ions or the Tech-exposure. (cas, cat) nical Specs was to be used for in.wrvice hydro testing. 33. W r.N zt.E CRACKS Additional testing of spargers and nozzles was to be performed in late 1976 or early 1977 during a Dresden 2 - Jan 75 Mar & Apr 76 = refueling snubber inspection outage. (cmy, cow, caut) shutdown See item II. 23 for previous information. Upon 34 W N02ZLE CRACKS rT of the nozzle blend radii in Jan 75 a total of
- 400 linear indications were observed. All in.
Monticello - Oct 75 - shutdown dications were removed by grinding. The max grinding ecvity depth from the clad surface was 1/2 in, which During the Sept 75 outage, a crack was found in one wLs equivalent to a depth of 1/4 in. into base metal. of the 4 W spargers while performing an inservice Th2 crors-sectional area'of reinforcement rzmoved inspection. They elected to replace all 4 spargers at the worst grind location was 0.481 in.2 with spargers of a more advanced design. In the process of removing the spargers. cladding defects Subsequent inspection of the W nozzles took place were observed in all 4 of the reactor vessel'W la Mar 7ti. An underwater TV camera inspectior, of nozzle corners. PT of the nozzles indicated the all 4 noz:les was done froi inside the vessel as presence of ~ 180 cracks. The cracks were ** tight" well as UT from outside, utilizing a standard CE with some cracks penetrating as much as 1/4 it. Into precedure for the latter. the nozzle base material. The cracks were aftri-buted to thermal fatigue of the cladding caused Additional exams were performed in Apr 76. A 2nd by W temperature oscillations at the reactor UT was conducted utilizing the Breda technique vessle nozzle penetrations. Oscillations of 100'F which g ain showed no repcrtable indications. PT were estimated at the reactor vessel nozzle corners. was also performed on the accessible portion of the The installation of the new spargers, which had an lower half of one nozzle (240'F). The exam revealed interference fit with the "as le f t" nozzle, was 9 small linear cracks, the longest being 1/8 in. Of expected to reduce the by-pass W flow, ths 9 in'dications, 3 were located in the blend radii which appeared to be the major cause leading to of previous grinding. The 2 longest indications cladding defects. [, were ground out beforecreaching a depth of.07 in. \\ The re.aining 7 cracks were not ground out. Repair of the defects consisted of grinding out the cracks. In some instances it was necessary to grind in additica to the PT inspection, UT of all 4 W into the base metal to remove defects that had ne: les was performed again using the technique penetrated as much as 1/4 in. into the base metal. daveloped tv Catti. The inspection showed no in. PT was repeated af ter repair to ensure that all dications cf reportable magnitude. He results defects had been removed. provided furth:r assurance that no significant cracks existed in the nozzles. Confidence in the repair-replacement work was en-banced by amending the PT procedure to increase, only the !J.cr nalf of the nozzle was PT examined minimum penetrant and developer dwell times, removing bscause the highest stress in the nozzles occuired " tired material from the outer 1/32 to 1/16 in. of at the top t?.3 ho' tom; therefore, inspection of either clad surface and machining the thcrmal sleeves of ent would previde data concerning the severe cracking. the replacement spargers to final interference fit It was esties.*d that an additional day would have dir'ension rather than sizing by cold-workiexpansion. bsen required ta complete a PT of :e top half of [.1 A total whole body dnse of - 471 =an-Jem was dis-the no::!a and 5 days and 18 rem to 32 men had already ' tributed over ~ 450 individuals da. ring the sparger best expended on the job. De UT using the Catti replacement. The highest individual *aose was technique provided assurance that no cracks had ~ 2.5 rem. The initial max radiation levels in propagated into base material 1/4 in.on any of the the work area were ~ 7 R/hr. (cmf',diu) nsazles, and provided assurance that cracks were nit initiating in the root of previous grinding in base material. 35. CONTINUATION OF ITEM T1. If p& Unit 2 experienced 19 start up/ shutdown cycles between tha inspections done in Jan 75 and Mar 76. The total Cooper - Oc t 75 - shutdown number of cycles on Unit 2 was 120. It was their normal operating procedure to minimize the number of start up/shutdon cycles and minimize rapid t'emperature ' See item II. 26 for initial inforration. During changes in the final W temp. On the basis of con-the outage, all 192 fuel bundles adjacent to the cerns related to the discovery of cracks in the W in-c ore instrument and source tuta s were removed nozzle blend radli, CE and the ut!!ity were to review and the fuel channels were im. d.ted to determine operations during start up and shutdown to identify the eagnitude of the wear emed by the in-core changes in gerating procedures which could reduce tube rubbing and/or impacting against the fuel g channel corners. Of the 192 shannels inspected, the ragnitude of thermal cycles on the W nozzles. 125 channels or 657. of the total inspected were considered rejects. The reject charnels were re-
- mcced and new channels installed on these fuel bundles. The channel wall was perfotated on 4 of the channels.
s., t, ea.., h~ ]
- 11. h u. tor ti...
tm perf orated channst centainePa >;. ' .L t:" Pa 23 corner 2? in. in length, several bran.: eratas and a piece missing leaving a hole in tne side of the channel of a aise ~ 1 in. by 1 1/2 in. The piece 36. FAILED JET ITM.r f.ENINER TACK kT.LD* k of channel was located in a position flat on the core plate between the fuel support castings at a Incation , Pilgrie 1 - Mar 4 = refueling shutdown previously occupied by the fuel bundle. The piece was retrieved using an underwater vacuum cleaner. yr of certain reactor vessel interior surfaces with integrally welded internal supports revealed that 9 Another channel contained a split in the corner jet pump restrainer bolt keeper tack welds had - ~ 23 in. in length with several branch cracks, separated or 8 jet pumps. Se 3rd channel contained a split in the corner ~ The separations were the result of a small amount of 21 in. in length, several. branch cracks and a piece vibration. There was no indication of restrainer missing leaving a hole in the side of the channel bolt keeper rotation. There was no structural i of a size ~ 2 in. by 3 1/8 in. The piece of channel degradation of any other part of the jet pumps or was located in a position on edge on the core plate reactor vessel components. Since the restrainer between fuel support castings at a location pre-bolt did not move and there were no other components viously occupied by the failed channel. The piece affected by the separations, the probable appeared to have dropped directly down from the hole consequences of the event were considered to be in the channel. The piece was retrieved using a minimal (or none). L/ l special tool designed and built for removal of the V piece. A similar inspection of the jet pumps was also 4 performed during the 1974 refueling outage. No The 4th channel contained a split in the corner indications of reactor vessel jet pump restrainer ~ 26 in, in length, se*eral branch cracks and 3 bolt keeper tack weld degradation were noted during pieces were missing leaving holes in the side of the the 1974 inspection. A video tape of 8 jet pumps channel ~ 3 1/2 x 5 in., 2 x 5 in. and 1 1/2 x 1 3/4 in. was also made concurrent with the 1974 Jet Pump Two of the missing pieces were located and removed. inspection. A review of this videdape was performed The remaicing missing piece came from the channel and no anomalies were observed. hole ~ 1 1/2 x 1 3/4 in. Forty fuel bundles had The preload of a beam bolt was checked at Dresden 2 been removed and/or were removed in the area felt most probable for finding the remaining missing piece. and found to be acceptable although Dresden 2 also The entire area was carefully scanned by 2 independent had failed restrainer bolt keeper tack welds and had inspection teams using an underwater TV camera. no indicatiu s of major jet pump vibration. It was Neither team could locate the missing piece. A therefore concluded that tack weld separation was special crevice device was also used to vacuum around unrelated to preload on the beam bolts. The tack (- the 9 fuel support castings that had been vacated welds of the reactor vessel jet pump restrainer bolt by the removal of fuel bundles. The piece was not keepers were re-established. (cqf) located or retrieved. Frnm an evaluation of the hole sise in the channel. 37 OPERATINO PROCEDUKES REYTSED TO 1.IMIT FV it vas believed that the missing channel piece was TEMPE MTURE TRANS1ENTS TO MINIMIZE THERMA 1. ~ 0.08 in, thick and roughly the shape of an iso-CYCLING IN W N0ZZLES sceles triangle with a base of 1 1/2 in. and an altitude of 1 3/4 in. The area of the channel hole Dresden 2 1976 I v s 1.8 in.2 It was felt that the missing piece' we.a located on the core plate. However, since the A special study was initiated to determine the mGsing piece could not be located, CE was requested corrective action necessary to,sinimize the W to evaluate the safety implications of the missing nozzle crack problem (see items-II.27 and 33). i channel piece s maining in the vessel. Their eval
- It was determined that the probable cause of cracks untion concluded that the 2 areas of concern were in the N nozzle blend area vasitherinal cycling interference of control rod insertion and blockage of the 55 cladding. Sin =e the thermal expansion of coolant, flow to a fuel bundle. It was highly characteristics of the cla@ing and carbon steel unlikely that the missing piece could prevent in-base metal were different,*Jigh stresses could sertion or scram of a control rod. Control rod result during thermal transients. Eventually, l
insertion and scram checks would provide an indication the stresses initiate and further propagate cracks of the problem if'one did occur. With respect to through the clad and into the ferritic base metal. blockage of coolant flow to a fuel bundle, it was highly unlikely that the missing piece could N temperature data for ths last fuel cycle was attain the required attitude, stay in the required selected for review. The 11~rgest N temperature flow velocity paths and continue in a flow stream changes ( > 100'F/hr) occurred during plant startups such that the piece could get below the core plate and shutdowns. Out of 18 stMups during the fuel and be available to block coolant flow to a fuel cycle, there were 17 startupY that had recorded bundle. In the event the pie.ce did co=plete such at least one > 100*F/hr event. All of the 18 a path, CE concluded that no significant fuel damage shutdowns during the cycle had at least one instLnce would result. If serious fuel damage did occur, of 100*F/hr change in final IV temperature. main steam line radiation monitors would scram and isolate the reactor. (cnj) N' NThe lower t!' plates of all FA's were to be drilled during the Fall 76 refueling outage. (cit) ."o a w. p - ne g am June 77 w- -,,-,-,,-w ,m- e .i IN N flow / N-g
- ' Os ret!nt for long per -
This could bs , m ed. !. i e-to perature should ate ne - - seuel haat-up 'et ,oi9;
- 1) perform in..re.
' * ? r and minimise (' h done by: 2-the Tech Spec liett of ' l'1 ace the W ) 2 atthe total time at low p e r
- 3) For J
i nificant heaters in service as snon as tass 11,lt. il estate termerature also underwent s grenslants during a " flood-up" of the reac o t r vetul. W line, DVR's having cleanup return throt,;h theflow, and temperature brought uring this operation, the water level wasp rtptdly to the main steam lines, th maintain the RWCU system at e.anTne warla RVCU flow l l when operating at low power. in higher W tor vessel would mix with the cold W resulting When holding .hs c;&in steam lines were fi nozzles. 4) tavs1 wtc raised rapidly up to the reac temperature at the W less than fla gs, the only limit being the shelt to flangs at hot standby, consider operating at 50*F sad Since this operation was doned the valve rated temperature and pressure (e.g. at 3When pla 1T linitation. through the startup regulating line an heated densate water in the piping and heaters 120 pais). 5) cycled (open and closed), the W nossle was d from tsactor heat and then cooled from the col erThis transient operation c t re flush return h an avg teep-the condenser via minimum flow or switas, the avg rate beirg 280*F/hr witAfter evaluating the data, piping prior to reactor vessel injection, W. (dly) arctura change of 94*F.it w:s concluded'that the operations wh c i h had the re cutting greetsat af fect on final W temperature wela r.nd out W heaters and flooding the r EpOUS CRACKS W W SPARCER REQt! IRED 38. final EXTENSIVE RtFAIRS vsssal. Ta invsstigate a nossible method for limiting Oskarshanns 1. sucaner 74 W tenparature changes, a special operating pre. This cadure was prepared for cutting in W heaters. An See item 11.29 for additional information. i the final procedure directed the operator to lim tThis rate of k inspection of the reactor ti.ternals revealed a esa increase to 100'r/hr. i the floal W in the W sparger. The damaged sparger consiste,d of incracae was controlled by slowly open ng W teeperature extrterion steam valves and monitoringThis procedure limited the man rate f the number of cracks. a continuous ring located at the upper end o tr reactor downcomer, about 0.7 m above the reac o 320*F/hr of change of the final W temperature toduring cutting in of the 1st set of W heat hle core, serving to distribute the W uniformly terp3rature. nd ~ 1.5 hr. h ough 6 to en overall change of 80'r/hr over the bottom level of the vessel, flowing up f' h heat-up region. During previous startups, it was found t atof ~ 400-500*F/ hour with 90*r tempera s chtnges were common. The dose rate in the reactor service room ranged up rets: Therefore the d 100-300 mre:/hr. Th3 changes in W nozzle temperatures experienceduring a flood up of t was to a level of ~ control station for moving the lead working ca i wed ice room arranged on the floor below the reactor serv l changes to cin. The to datermine possible operationaIn light of the reactor using a transport 2pening of the butiding. lead reactor service room crane was used to c trausients. ified in the vessel te !!;,m t.T design Italts spec imite the=e i sly Tech spec., Q :.;peerd that the method prev oudsscrited of rapid floodtng from the control station and snervised by TV-cameras. h ly cage. d with telephone. TV moni_torg was to D e number of flood-ups (caq) i . he, ca_ge,was equ ppe Normally two men were T practical adod.be limited to the extent practic.able.electricity, fresh air etc. Conversation and all vital parts working in the cage. The l of the TV pictures were recorded on tape. working a operator was able to observe thehandle the e 2 ided CE Service Ishpation l4tter No. 208 prov i W no: ale C edditional r.a..:endations for minimiz ngOn-of f cold W flow cycling at low glove boxes on the cage. M F3 thermal cyc16.g. of the major contributors l To remove the W sparger ring it vaa necessary to d also power operatiun was one L' hen the IV was of f. rature (~550'T). 6 T/C tches, 6 riser piprs, ( link packets anA number of to W nourie thermal duty. le cooled to to cut the sparger ring into 6 parts.For, cutting T/C tubes ths nozzle heated up to reactor to:pe and when the W was turned on, the nossThis cycling could special tools were developed. ed. For and pipes, modified standard tools were usb t equip-W flow temperature (~100*T). utcoatic tow flow cutting the W ring, a plasma cutting ro oThe robot als , bs minimized on plants with an a W control valve by the proper use and maintenance 'the entire 6 tting ment was developed. On plants i.ithout autoestic low with a sludge extraction device.The tools were tested in l flow valves, it was recoanneoded that such va vesDypassig steam to the c of this valve. ld was per formed under water. W ring. a full scale model of a sectien of tg der j d that the Testing of plasma cutting was perfonned also uncarrieI be censidered when at low power (provide be added. d to keep the condenser was available), if requireW flow rate high enough to be controll using a full scale model of the case. Training of personnel was water. ) l and to avoid **on-off" behavior.the W flow valves and control ersl es and con-bjecta The core grid was protected against f alling oTo red( also the W heater and drain system va vdition to achieve insure that SS covering sheet. ide the RpV radiation level in the varking area ins by aL trollers were maintained in a conExcessive leakage ible. Also in the water level was kept as high as poss anufac-l e of ten required quality system performa ce. n special ingots covered with a SS sheet were m a f ull flow W flow control va vW pu p start /stop operation for low flow eskeup e tured and fitted to the front of the cag. m requirements, June 77,
- . ~ e
- - e---- = 1 s _
- _m_,
- 11. Reac tor le.t.
- e.' t.f p. ' '.
6 h,n of'th.
- e.. -.
] } -r 51.e -te sements. Ar. au -e.*' ure of the No repair p:sssit.st t t*- '.1 y,c ealdered w.re in- . -nioped to reduce the t h.
- on.r., t or 's arr~s. Extracting of t ~ s-
.tr fumes stalling a new sparPr/t%r.at sleeve design trin,, w.s developed by CE or re-,vir.; the spprger f rom the relve w o performed by a fan connected ts reac tor building ventilation syster. vessel and replacing the the rmal sleeve, grinding the nozzle area and returning the sparger to the (ddt) with a low water level, the radiation dose rate in vessel. the pressure vessel 25 cm above the cere grid sheet was 12 rea/hr. By the above mentioned censures the i radiation level was reduced to 1.5 ree/ar at the operator's hands outside the cage. The max dose l rate inside the cage was 10 ares /hr, although the i normal level was 1 aree/hr. The prolongstion of the summer shut down amounted to ~ 4 1/2 me. Only 1 1/2 no were used inside the 6 RPV. Preparing of tools, methods, shieldings, etc. l before the first cutting, demanded 2 no and clean up I Inside the vessel 1 me. The total number of man-hours required was about 74,000. About 120 men were directly engaged in the replacing of the W sparger. The total full body dose ta 107 persons was 43,000 ares. The man hand dose was 3360 ares and the mean value of the most exposed hand was 1460 mrem. (cyq) t l 39 POISON CURTAltl VIBRATION DAWACED ftEL CHANNELS 41. LOOSE JET PUMP RES1RAINER CLAMP SSLT KEEPERS 1 BROKEN TACK WELDS - VIBRATIONAt, MTICUE CRACKING Muhleberg - Aug 73 Vermont Yankee - Oct 73 Dresden 2 - Mar 76 - refueling s tdown Pilgrim - 1973 During jet pump inspection, loose restrainer clamp See item 1.12 bolt keepers were found on 19 of the 20 jet pumps. Of a total of 40 keepers, 30 were found to be loose as a result of broken tack welds. These keepers 40. W WOZZLE AND SPARCER CRACKS were tack-welded to the restrainer assembly to ensure that the gate clamp bolts remained tight. Inrsedia tely ~ Humboldt Bay 3 - July 76 - refueling shutdown af ter this inspection, a tension test was performed on one of the jet pump hold-down beams to determine An inspection of the W nozzle was perfomed uti-whether any further slackening had occurred. The 11:ing underwater TV. This inspection consisted hold-down beam tension was found to be 3S50 psi; of positioning the camera to view most of the cir-the minimum acceptable tension was 2S00 psi. cumfever.cc, of the nozzle as well as the thermat The original restrainer gate assembly and clamp bolt s1ce.r..it the point where it enters the nozzle. f The results of the inspection revealed the presence keepers from jet pump No. 5 were sent to CE for of cr. ding in the thermal sleeve where it enters analysis. Each restrainer assembly was found in its the W uozzle.,The cracks extended - 3/4 of the proper position, with both clamp bolts fully tightened. j circumference of the sleeve. In the region of the The keeper failures had no ef fect on jet pump operation ' 7 to 9 o' clock position on the thermal sleeve Broken tack welds on jet pump restrainer clamp (when viewed from the center of vessel) there bolt keepers had been found on 2 ' previous occasions. were several cracks which had resulted in the loss In a report entitled "Lah Exam otht Pump Restrainer } i of ~ a 1/4 in. by 1 1/2 in, portion of the thermal sleeve. The thermal sleeve to spargtr rin; weld-Assembly from Dresden 2," CE staYe'd that the keeper rent was examined and resulted in the discovery tack weld failures were prcMIly caused by vibrational of a crack extending - 50". of the circumference of fatigue cracking. Itwascoia]ecturedthat,with the weldment. The vessel nor:1e transition blend, a single tack weld, the keeper tended to be lif ted radius area exhibited aeveral (3 or more) crack off the gate surf ace as the tesult of weld shrinkage. indications in the vicinity of the hole in the with the keeper thus supported by the weld, jet pump thermal sleeve. assembly vibrations induced the keeper to vibrate, ultimately fatiguing the weld.& The W sparger ring was also examined and several cracks were observed radiating from the distribution .As corrective action, CE reco ended that 2 tack holes on the sparser (max length ~ 1/4 in.). These welds be placed 180' apart on Rch keeper. By were assumed to be from thermal fatigue with no securing the keeper in this mahner, what was termed further growth expected.. 'No of the 6 W sparger the " point support mode" would be eliminated, ac-support clamps (opposite the side of the thermal cording to the report. Thus the original objective sleeve) were observed to be tilted about 10', was to reinstall the keeptes with 2 tack welds but no cracks were found. 180' apart. However, difficulties were encountered in operating the welding, equip u nt, in obtaining a suitable welding arc (ground), and in seating the keeper rims. Nrthermore, there did not appear to be sufficient accessible keeper rim material to permit the placement of 2 diametrically opposed a
- -a= o --- ~-- s.- -- %. July 77'
--g. i-, ,_--m_. -,,. - _. .w.,_ 43. PADt Ci fN ? N"."**.t
- ~ "at M N"P T 191.
IN ptm% % <! df 7 9 D - BERYLI.I M t U M. 1 [_ [ g 15 ...L. Leesupe of thiss c$nsidstations, t tr --- keepers were rewelded to original spec; Big Rock Pt. - Mar A.lurw U. - re f uel tn; shut e - i... en. tack weld per keeper. These welds were to t. in--e ted during the next refueling outage, and Because the analysis of a creid samole obtained during C:' rare-mendation would be considered af ter this the 1973 refueling outage an< tic ated a.99*. berylliur oxide content, an inspection of the initial neutror. invec tion. (dir) sources was performed in Mar 74 The attempt to remove the source assemblies from their channels 42. CEACKID W SPAR ERS AND NOZZLES _ was initially unsuccessful. Only the' upper portion of the cylinders and the antimony pins came out. The Quad-Cities 1 - Jan 76 - refueling shutdown beryllium-containing cylinders had expanded to the extent that they could not be extracted and were FT revealed one indication of cracking on each of subsequently removed with the channel assembly. The 2 sparsers. The test was being performed upon re-channel sere noted to have large cracks in the middle j quist by CE, which was concerned over the occurrence where the beryllium cylinders were located. Both of cracking in several W sparsers of the design antimony pins were removed and VT with no anomalies utilized in Quad-Cities 1 and because of the cracking observed. One beryllius section was removed from ptsblens found previously on the sparsers in Quad-its SS cylinder and was found to be grossly expanded Cities 2 (see 11.25). and oxidized. On the 60* W sparger, there was a 1 1/2 in. linear The failure of the beryllium-containing cy'Ifnder indication which propagated through the junction was' attributed to excessive internal pressures, the b2x-te-thermal sleeve veld. On the 150* sparger, subsequent breach of a weld allowing water to enter a 2 it. linear indication which propagated through and resulting in the formation of an oxide accompanied thi junction box-to-thermal sleevp weld was located, by further vol expansion. Because of relatively Tha linear indications were interpreted as cracks close tolerances, the swelling of the cylinders.im-sud the inspection was terminated. A work request posed lateral stresses on the source channels fesulting was initiated to replace the apargers and to inspect in their failure. A limitation on the design,rlife of the W nuzzles. this type of neutron source was the pressure. buildup from helium and tritive gas formation from the (N.t) following the removal of the old sparsers, PT was and (N. 2N) reactions between the fast neutrons and conducted on the inner blend radtus of the W nozzles. beryllium. This pressure buildup, caused by prolonged Nunerous linear indication's were observed on all 4 use of the beryllium cylinders, might have been nozzles. The majority of these indications were assisted by stress corrosion cracking of the 304 segregated on the spper half of the nozzles. The SS cylinder. indications averaged 2 in, in length, and the length ( of the longest indication was 5 in. These cracks The SS source channels were replaced by-Zircaloy were ground out. On the 60* norile, 2 grinding areas channels. New beryllium-containing cylinders were protruded a man of 5/32 in. into the base metal. On also inserted, each containing an old irradiated the ISP nozzle, all grinding was confined to the antimony pin and one with a new unactivated antimony ci:f =tal. On the 240' nozzle, one grind area pin. The old antimony pins were to be removed when pr otrund 1/S in into the base metal. On the 330' they could be replaced by the irradiated new pins. nonor ~7 erhding areas protruted a max of 1/8 in. imo tav % w aetal. In June 74, many loose pieces of material, presumed to be beryllium oxide, were observed in the reactor Th= ensa nf '.is event was attributed to fatigue vessel and small pieces were noted lodged a=ong fuel Q W m rk ra caused by flow induced vibration, rods of several bundles. All vessel internals (fuel, c' f and cc pouroad by stresses induced by the thermal channels, control rod blades and incore monitors) gradh-ts ic'v-nt between the W piping and reactor were removed to f acilitate thorough vacuum cleaning vessel intercal:. Imakage between the sparger and of accessible portions of the reactor vessel. All the W no= 1. c.atributed significantly to vibration fuel bundles to be returned to the reactorc... vessel of the spargur assembly and also imposed thermal were cleaned with a water jet. Wedged p.ieces"were stresses on the nozzle. The new spargers were dislodged with special tools. The new ne. tron sources desi;ntd with an interference fit to eliminate leakage were to be inspected during esch refQling outage and thus reduce vibration and thermal induced to insure integrity. - 7 (dnd) s stress cycling. The new spargers were to be inspected duri:g their next refueling outage. (dn:s) M44 naoxts stuTaox sounct isstsst.v. prters lu coat N miistone 1. Oct u. refuelig shutdown 4 While performing a neutron sourse assendly inspection. with all the fuel removed from the reafter, a neutron source assembly was found to be broken *.* tiforts were made to retrieve the broken source assecbly and to locate pieces that were identified as missing. A majority of the broken soarce assembly was removed from the reactor and the remaining material identi-fled. This material consisted of a 7 in. dia, hollow SS tube, with a.020 in. wall thickness and an inner beryllium sleeve, sheathed in SS with the ( center portion filled with antimony. The total length of the material was estimated to be 76 in. 38 in, of the.7 in, dia tubing and 2,19 in. source e-e -- - s Det. 77 = b m W 1. T L*.
- 11. Reneter P
- Anunseterr.t]
.. e s.% e. M. i t.. dia. att of this ]
- p. U
'.Wr ut o 1d be sasn an the low-r'. 4 s-r pr t j plate. In addition to the lower rote support plate Fall 77 refuelins st< o d that p-veral af the to: inspections, the guide tube areas of tne control welds had crac .J. thus the sparger was to be re-rods in the area around the broken source were also placed. inspected. One piece, apparently a broken section of the source pin, was located in a control rod At Dresden 2, IIT vdP performed on the W nomsle guide tube. la Sept 77. There were no recordable indications found in the 4 nozzle bores. One recordable in. To preclude the possibility of the 4 other neutron dication was found in the inner radit. This in-source assemblies breaking apart, they were replaced dication was found in the 240' nomsle with an 801 along with the broken neutron source assembly with fSH amplitude and determined to be from a deep gried-new neutron source assemblies. The exact is11ure out cavity lef t fross the 1975 outage. In addition, sechanism of the neutron source assemblies was not all & W safe-ends and their associated welds were known. It was also decided to increase the surveil-IJT and found to be acceptable. lance on the 16 control rods in the 4 by 4 array surrounding the broken source assembly. The normal In Oct 77 the 240* W sparser was removed. De weekly surveillance of moving each non-fully inserted 240' nomsle inner radius and bore, as well as the control rod one notch, in both directions, was to be accessible areas of the other 3 nosales, were increased to daily, for the 16 control rods in ques-cleaned with a cleaning solution and very fine tion for I wk of full flow operation. (dup) (120 180 grit) flapper wheels. The subsequent FT revealed 9 linear indications in the 240' nossle. There were no indications found in the accessible areas of the other 3 nozzles. N o of the 9 indi-cations were found in the inner blend radius at S* and 30* and were S/16 in.,in length. Dese were located in the bottom of previous brind-outs, but were completely removed during the 1st cyaning of that area witi 180 grit flapper wheel. $he other 7 indications were located in the nostle; bore. They were situated a distance of 5 to 6 inches into the bore at ^ 180* nossle azimuth and from 1/2 to 3/8 in, long. Six of these indications were removed by recleaning with the flapper wheel. De recaining indication, the only one which required grinding, 45. BROKEN JET PUMP RESTRAINER ClJLMP PO!.T KEEPER was less than 1/16 in. deep. The UT and FT indicated that the interference fit, forged tee spargers had TACK WELDS been effective in reducing thermal f atigue cracking Dresden 2 - Sept 77 - refueling shutdown as evidenced by the significantly smaller crack depths than would be predicted by generic crack During an inspection, Jet Pump #4 outboard restrainer growth models. Dresden 2 had experienced a total clamp bolt keeper was found with a broken tack weld. of 35 start-up/ shutdown cycles since the last com-Tack welds on all other keepers were intact, with plete PT and grinding of the inner radii in Jan 75. The interference fit of the 240* sparger thermal each rcstrainer assembly properly positioned and both f clamp balts in place. Keeper tack weld failures / sleeve had remained intact as evidenced by the dif-were attributed to vibrational f atigue cracking. (/ ficulty with which the sparger was removed and Jet %f4 outboard keeper was revelded with 2 confirmed by measur ments taken of the sleeve dia. tacWJs ~ 180* apart. This would ensure that the keept,f uld not be lifted off the gate surface Inspection of the W spargers at Hatch 1 revealed dortr = reration. Similar problems were reported at least 1 fine hairline crack emanating from a in 1.7? and 41. (ebu) hole in one of the junction boxes. ;.Several visual indications which were possibly cracks in the periphery of the flow holes were observed. As a 46. W SpARCER AND N0ZZ1.E CkACKS result, they decided to replace alf'4 W spargers. (eal,eam) browns Ferry 1 - Nov 77 - refueling Brunswick 2 - Fall 77 - refueling During the 1st refueling at Pea (h Botton 3 in Jan **. Dresden 2 - Sept 77 - refueling an external UT was performed on all 6 W nostles. Hatch 1 - Mar 77 - refueling UT reflectors requiring further 4nvestigation were peach Bottom 3 - Jan 77 - refueling found on 2 of the nozzles (the "D" and "T" nozzles). peach Botton 2 - Apr 76 - refueling The W sparsers associated with the "D" and "T" nogales were removed, the nossle;(-surfaces were flapper wheel cleaned. and PT wa pe'rformed. The At Browns Ferry 1 the W sparsers and thermal sleeves were removed from all 6 W nozzles. Machining to PT showed 4 cinor indications onaphe "T" nossle, remove the cladding and heat-af fected base metal No indications were present on the "D" nossle. from the nozzle was initiated. The interim spargers The indications on the "T" norsle were removed by were procured and were to be installed. light grinding and did not penetrate the cladding into the base esterial. VT of both nogales showed CE had replaced the Brunswick 2 W sparger with evidence that a eetal to e,etal fit was still eatn-one of a slightly different design. The original tained at the time of sparger removal (a full cir-installation involved use of a sparger with over-cumferential band of unoxidized metal was visible e sine orifice holes. To correct the condition, in both safe ends). Both sparsers were reinstalled ~ ( orifice plates with the proper diameter holes were with an interference fit per the recommendations of tack welded over holes in the sparger to give the CE. UT subs.quent to the removal of the surface proper flow. Exarns of the sparger during the w o-w %= Mar 78 .h. G g ~ 3 T l'a* c ', r ht. j chennels vert M -** r d e-1er tne centinued E'
- cycle of operat !*
?:
- aen chan.els had no
~ f indientions showed that the reflectors had not been removed and their characteristics were not chann d. The last part of the Cf *r.,lution for reducing the Mont likely these indications were in the clad bond LPRM vibration problems required drilling holes in leterface. Even if these reflectors were physical the lower tie plate of the reload fuel. The new discontinuities, it was highly imlikely that they would fuel had already been drilled prior to receipt on praptgate due to ther=al cycling. site. Tne drilling was performed in the Fall 77 underwater by burning holes through the tie plate Acticas had been taken to altigate the W nozzle with electrodes (E.13.M. technique). Once the holes crccking. The startup procedures had been revised - were burned the assembly was moved to an underwater se es to provide increased turbine bypass flow to the inspection stand where both the inside and outside ceedinser during vessel heatup thus providing contin-of the holes were VT inspected for acceptance. The sous IV flow during the startup interval. An auto-electodes were replaced af ter completing drilling matic level control system was installed during the runs of about 20 bundles. At the beginning and end cuttge to position the 3 in. bypass valve which of a drilling run, a set of holes were drilled into cantrolled condensate delivery to the reactor vessel a replaceable blank in a duruny FA. The FA was priar to startup of the turbine-driven reactor feed lifted out of the water and the blanks removed and pumps. As pressure increased and the bypass valve inspected again. approached full open, a reactor feedpump was placed in ssrvice with its speed being automatically con-Equipment set up was begun by CE ~ 10 days prior to trolled via reactor level and its discharge valve when the drilling operation began following refueling. throttled via operating procedure. The bypass
- The drilling stand and inspection stand were located I
valve control system and the feedpump operating in the fuel pool with the power supplies and centrol I prscedure had eliminated batch feeding of condensate equipment along side on the floor. The equipment was and had provided continuous W flow thus minimizing checked out by drilling holes into a discharged'FA tha number and magnitude of W temperature cycles. from the reactor in additien to drilling holes [in dum y FA. e Cr. cencluded that in light of the results of the nszzle exam and considering that there had been only Originally 22 days were scheduled for drilling which 7 startups since the refueling outage (42 prior to was to be done in parallel with the fuel sipping f the cutage), there was no undue risk in deferring and invessel work. Due to the problems encountered nozzle surface exams until 1979, assuming no un-of having to repair some core spray elbows in the seceptable UT indications were found in the 1976 reactor vessel the invessel work required much longer than expected and the drilling never became critical Path. In spite of repeated drilling equipment At Peach Bottom 2 in Apr'76, 143 total indications allures and some support equipment problems the t vsre detected on the W nozzles using PT. The N r s W88 C mpleted in 21 days and M M. onh nozzles were reeleaned using a flapper wheel and a ^ " 2nd PT was performed. A total of 42 indications I st due to problems with the drilling equipment. were observed. The decrease in total number of The servo valves caused the most trouble, but pro-f ndf e mtions uns attributed to the fact that the blems were also experienced with the power supplies ra m rity of votestions were surface defects in the an ca es. E a as cany as 5 power supplies on eMing and could be removed during the flapper site either in repair or in operation and flev parts wheel cluaning process, prior to the 2nd PT. The and equipment in trying to maintain the schedule. 2no indicaticns tere recoved by grinding. Only 4 (*"*"*} crd.: pencerated the cladding into the base metal. All 0 laae metal penetrations occurred in I the Imr bler.r radius, or in the bore, very close A 48. W SPARCER CRACKS b - to the blend radius-nozzle bore tangent point. w 4; The drepest pen-tration into the base metal was ""d" mias'" - "8 r ' 6 - "'"* ""**G~ 1/S in. (eal,eam, ego.egr,egs,egt.eif) O Inspection of the vessel internals re7ealed cracks in the W spargers. Exns showed thaF it was Q 4?. AtDTTION TO TI.26 - 1.PRM VIBRATION - "'C'888rY to " place bo*h spargers with new compon-F1EL CHANNEi.LTEAR ents made of less crack sensitive steel. The cracks " ' ara'd by 1"***ctr$5*111a* C " **1*" 1" "*** rial i O brunswick 2 - spring 76, rall 77 - shutdown close to welded joints. Because of*high radiation & refueling levels inside the vessel, the work o (replacing the spargers was carried out as deep an>possible _ j in Mar 76 they shutdom to investigate possible underwater by divers in a protected _digt 1.sttle. fuel channel damage as reported in item II.26. Thepreparatoryworkauchasprocuremenfofmaterial, i The pro;; ram entailed the borescope inspection of fabrication and assembly, shielding, preparations for Itl of the 192 channels adjacent to in-core non-underveter plasma-cutting, etc. accounted for a I itors during operations. (Bundle BR 444 which was 18tte part of the repair apart from the actual i droppeo was not inspected for channel wear.) The rep acement. operations. The entire operation ex-J l f inspection results showed no through-wall wear on tended the outage by - 2700 hr and resulted in 08.1 any channels. Eighty-five channels were rejected. man-rem to 89 men. The replace =ent ef fort was This placed the rejection rate at 44.5., although similar to that performed at Oskarshamn 1 (see II.3S). bN'" 'N {\\ only 19 channels (107.)had calculated scar > $0. pen'ctration. There were 88 channels (46.1.), which were acceptable for future use in the reactor within restricted locations. Fourty-seven of these Aug 78 s o - + -- (
- =.
0 0 o Vol.14 4 .-.? n t.,;. . y gm.
- n.mt m
- 11. he i.e
- 5
) p.., ,e o r, *.9 .j.. j brunwick 2 - Tall 77 - rebeling 50. CPAC E D C04". D*.AY SPARCER See 11.46 for previous inin. 1*
- W spargers were Oyster Creek - Oct 78 - refueling of the " junction box" design whN a S3 thermal sleeve. During installation, the W sparger* thermal While performing ISI inside the reactor vessel, an sleeve was expanded ~.005 in. cversize and jacked indication was discovered on the core spray sparper into the W noaale. The flow holes on the original for System II. Tne core spray sparger consisted of W sparger were initially drilled oversized due to 31/2 in. schedule 40 type 304 SS pipe formed in a design error. To correct this probles, orifice 2 semi-circles held in place with brackets attached plates were tack welded into the incorrect W to the core shroud. Nossles were welded into the sparger flow holes. In the spring of 1976, the W bottom of the pipe about every 5 in to direct the sparsers underwent VT. This inspection revealed flow of water directly on the fuel bundles in a pre-several cracks on the flow hole orifice plate tack established pattern. The circumferenfiel crack welds. Weld repair of these cracks was attempted but fsee fleure) was,v1/32 in, vide at its widest point was unsuccessful.
and extendede* 2000 around the sparger. It was' located at an azimuth location of 2088 in the reactor W spargers with forged tees and Inconel thermal vessel. This was e 53' from the inlet and 32o sleeves were ordered from CE. It was decided to use free the end of the sparger arm. The crack was Chicago Bridge and Iron "as built" vessel curvature through the wall, as determined by pneumatic testina, dimensions for the W sparger.instead of using a and was smaller inside the pipe than on the outside. template in the vessel. This saved ~ 3 to 4 days The crack appeared to have initiated close to one critical path time and personnel exposure. The of the spray nozzles and was adjacent to one of the RPV was hydrolased with 6,000 psi water. The support brackets. Because of the design of the hydrolasing was successful and re=oved almost all sparger and the mounting brackets, it was concluded of the red oxide. An invessel work platforta supported that the sparger would have been held ja place if by the core shroud was installed. Lead blankets called upon for operation even if thagerack had pro-were hung over the vessel valls and laid over the pagated completely around the pipe circumference core spray header to reduce the radiation exposure. before {t was discovered. ?. The general area radiation in the RPV was 400-500 mR g. after all shielding was installed. All 4 spargers a were removed without any difficulty. They had
- " mW #"T extremely tight fits in the nozzle and required a special jacking device for removal. After removal, y
- f ac ms na prn/
each W sparger was given a VT. This exam revealed 1 / p no indications except the flow hole orifice tack 1.- I nerorro* "I-welds. / The W sparger brackets were given VT and no indi-ce a cations were found. Each W nezzle was prepared for () .) ~ PT by surface grinding with flapper wheels. A f special N nozzle mock-up was used to train each grinder prior to working in the vsssel. The W y,,, nomsles were then Fr examined. The extent of the PT. , p,c = = vi k z nwaan ***'73M was 360* around the blend radius and nozzle bore. .b The entire blend radius from the vessel shell to-l nozzle forging veld was included. The nozzle bore was examined up to the cladding / base metal inter-face. The PT was done in accordance with the require- ?g;2 ments of the ASME code. Interpretation of the FF g exam was made at least 25 min after application. The 45' W nozzle had 12 indications, the 225' W ] .m-4 nozzle had 5 indications, the 135' W nozzle had t h.N 7 indications, and the 315' W nczzle had 3 indica-m and < 1 in. long. They were all re=oved by light ' .Y tions. All the indications were s=all, hair-like 4 grinding without entering the base metal. W The most probable cause of the crack was due to 1 nozzle leakage land diameters were taken with CE's local cold work of the sparger and stresses imposed I during installation and fitting of the sparger witha,n l dial-a-bore. 'the W sparger ther=al sleeves were then machined.010 in larger than their respective the, shroud: Fitting the ytpe into position during Installation could have resub ed in stresses larr.c W nozzle and welded to the spargers. For installa. tion, the W sparger thermal sleeves were chilled enough to propagate a stress crosion crack. in liquid nitrogen and inserted into the nozzles. The upper core spray spargerirepair consisted of the All spargers were installed without any difficulty. Total task time was 162 hr. (eus) addition of a bracket assembly to, provide, axial support to the core spray piptog an the vacinity of the crack. No attempt would be made to seal the crack. The bracket asse=bly was constructed of Tyr" 304 solution annealed SS and was held in place b3 four 3/4 in. bolts that were pre-loaded and locked ( in place by Class A cype locking caps. The bracket assembly was fitted around the existing spray non.lcs I on both sides of the crack to provide axial support to the core spray sparge in the event the existina. i c - o - w a-Har 79 j l m. e ... e. .e ,e.
- rnt f r-ItM:
Pr Int. provide a descr!ptt.. its purchase dat e anc it is s t r. f..c t ien dar t?.. both normal and acciceni cmiditions. preparated completely around the pipe circum- -- A
- e re.
- 2) For each piece of ide ntified equipment. prov t.h the performance history associated with its usagt.
he addition of the bracket assembly would hold including the cause of any f at tures or nelf unctons tne sparger in such a way that if the crack were to and the frequency of such events. e proporate 3600, the man opening that could occur would be 1/16 in. This was based on the clearances
- 3) Provide information on the suppliers and receiver's el the bracket to nossles and took no credit for QA/QC program in ef fect at the time of purchase.
clerping forces. Structural evaluations had shown This information should be discussed in terms of that the bracket would limit the opening of the providing sufficient bases for judging that the crack to $ 1/16 in. and that the bracket would re-integrity of the equipment was sufficient to permit =ain in place. The choice of materials, design, plant operation during viormal and accident condittens. bolt preloading and use of locking caps assured the ggg,) bracket itself would not come loose. For the bracket to be forcibly removed from the sparger the bracket vsuld have to be deformed at least 1/4 in. and this Ns2* CS SpAncta rirlac caacxS would require a force of
- 3000 lb. There was no ronceivable mechanism that could apply this type N
Oystu Creek - Jan 80 - ufueling of farce to the bracket. While Performing ISI on the core spray sparser A 360* crack 1/16 in. vide on a sparger would con. piping for System II, two indications were dis-servatively divert < 21 (61 spa) of system core covered. Those indications were subsequently aprey flow throush it. This in itself would not evaluated and determined to be cracks. affect the ability of the core spray system to pro. ~* viss 3400 spa to the core at a reactor pressure The investigation revealed that the first indteetion of 110 psig, but some nozzle flow would be diverted was located at an azimuth ~ 152' and adjacent Jo the threugh the crack. GE calculated that the minimum sparger header branch veld. It appeared *.o extend flow f rom each norsle would be at least equivalent ~ 120' circumferential1y. The second indication to that obtained at 3400 gym from an uncracked was located at an azimuth ~ 170'. It ran at an sp<stger if the system flow was 3700 syn. In ad-angle of ~ 30'-40' to the centerline of the sparger ditisn, the installation of the bracket would and was ~ 2-3 in. In length with 2 branch indications not interfere with the core spray distribution pattern and would not allow water from,the crack running vertically for ~ 2 in, to directly impinge on spray from the sparger An air test performed on the sparger (p. F. Avery (". r.oz zles. Corp) indicated that the crack indications might T.e repair was performed remotely and the final' not have been through-wall as evidenced by lack of exa. to ensure proper fitup was recorded on video air penetration. Subsequently, a UT was perforred tap. The bracket assembly in place was examined on the areas in question. The first area was located ~ 152* and a through-wall crack was desceted in the at both ends and confirmatiort of proper fit was made. :34 recorded. (fng) HAZ of the header branch weld. The second area was located at ~ 170' where additional through-wall cracking was detected. I*' . NS*?tE W SPARCER AND THERMAt. St.EEVE OA The cause may have been due to installation rethods f g.,W mes . J and was being investigated. (gur god) .,b tys tr~ General =O e *:w.cond.2ed an inspection (Sept 79) at the nrvin Er.gir.e cing Co. to evaluate their overall M**' *
- . and CC programs. This company was a subcontractor 4
c.:! s :pp!!er to CE of BWR internal FW spargers and j ther=al sleeves. The results established that 7 serious uefic.lencies existed in the implementation of f tb OA progr:.:a relative to the manufacture of these ceeponents. D. ring this inspection, 27 deviations frcs appitephte codes, and contractual and regula-tery requirements were documented in the areas of t attrist identification and control, process control, velding and NUr exams. w All !Q. licensees and ennstruction permit holders vert to supply the following information within 90 days for operating plants and 120 days for plants under construction.
- 1) In:tertaine if reactor N spargers and thereal s1'c. eves canufactured and/or fabricated by the Marvin
( En; Co were purchased and/or installed. Since Nrvin Eng was principally a subcontracting corpany, de t (tr.ine if this equipment originated with the !brvin to and was eventually supplied through i e. tla r contractor / supplier. If any was identified, Par 60 s - e - *.-- L = l i ,. o N Vol. 1/4." 1 4 $3. JtT PUP BEAM BROKr*:. T*r U AT: M
- 11. kcact.e
- p. 31 i
l BWRs in Ceneral - 1980 I Cresden 3 - Feb 80 - 66 pcme r ( l Quad Cities 2 - Mar 80 - refueling At Quad Carias 2, a UT was performed on all. Pilgrim - Mar 80 - re fuelant jet pump hold down beam suczblies. A vidta N m11ston.1 - Oct 80 - refueling crack indication alonf the upper machined pu.- f ace near the beam bolt on jet pump No 16 The NRC issued a Bulletin (80-07) regarding was discovered. It was estimated in excess c,! BWR jet pump assembly failure af ter several plants 100 mils in depth and in the same location as reported jet pump problems. found at Dresden. At Dresden 3, alarms were received in the control Filgrim reported discontinuities on Jet Fump Seems room. Observed changes in plant parameters indi-1, 3 and 15. cated that a jet pump had failed. Shutdown was .A metallur-initiated to bring the Unit to cold shutdown within The failed assenblies were replaced.
- I 24 hr as per Tech Specs. The changes were 1) gical analysis was being performed by Cs and the '
generator output decrease from 539 to 511 NWe, 2) cause of the crack was identified as intergranular core thermal decrease,
- 3) core flow increase from stress corrosion cracking (ICSCC).
97.6 to 104.7 x 106 lb/hr. 4) core plate dp decrease from 16.1 to 13.8 paid, and 5) "a" re. Concerned with the potential for degradation of cire loop flow increase f rom 49 to 54 x 10 spa jet pump function, the NRC recommended the follow-3 while "A" loop flow remained at 49 x 10' gym. ing acticas at BWR-3 and B*.tR-4 facilities 8 It was then determined that jet pu=p No 13 had failed. Plants in their refueling outage, prior to startup, structures, the hold-down beam assembly,, jet pump were tot a) assess the integrity of the hold-Following vessel head removal and defueling. TV camera and VT of the jet pumps and vessel annulus downs wedge and restrainer assembly, 4) conduct revealed the hold.down beam assembly of the sus-UT to assess the integrity of the jet. jump hold-Feet jet pump had broken across its ligament sec-down beams at the mid length ligament areas tions at the mean diameter of the bolt thread area. bounding the beam bolt, and c) when startup began Failure of the beam assembly resulted in pump initiate the surveillance described below. decoupling at the diffuser connection. Subsequent in situ UT of all other jet pump hold-down beses, Plants in operation were to perform daily jet pump using a special UT technique developed by CE re-operability surveillance. Individual jet pump dp vealed ultrasonic indications of cracking at the readings should be recorded and used to establish same location in 6 of the remaining 19 beams em. a data base for expected characteristics for each amined. A sketch of*the typical jet pump assembly jet pump. Record and evaluate the following de-is shown in Tigs.. The indications ranged from 6 to. viationst a) the recirc pump flow differed by core % 20 alls. 3he beams were machined from s'tock than IC: f rom the data base for that pump, b) Inconel X-750 material, (hat) the total core flow was more than 101 greater thac 'the core flow value derived from established power-core flow relationships, and c) the diffuser to lower plenum dp reading on an individual jet pu=p fw,/ exceeded the expected characteristics established s..a s.n y..p., i 9 f.: A.a for that pump. rT : If it was determined that a jet pump was inoper- .T able or significantly degraded, the reactor was o q to be. shutdown in accordance with Tech Spec re- , god.gwo,bbs) ( v -a j quirements. .7, y Inve'stigations by CE indicated that-beam failure vas preceded by a long period,ofcMow crack prepa-l gation followed by a short period of rapid crack During this ti'e, the loaded beam p ,,,/, propagation. m deflects to accomodate the lead. Using data from v ses m, s.. :wn, Dresden and from another reactor which earlier experienced siellar failure, CE showed that beam deflection during the rapid propagation period allowed significant jet pump leakage. By monitoring loop flow and recire pump speed during this period, .u..n 8"t* *"*' icpendine beam failure couldibe detected during the rapau crack propagation pe'riod of a week to 10 days before actual failura g f r% (o .a: CE studies also indicated that initiation of ICSCC v s l in beams was unlikely before 2.6 yr of operation. Af ter crack initiation, a calculated period of 2.7
- 0 go h
yr of crack growth preceded actual failure. 3e-A cause of such slow crack growth rate, and assu=In f
- 2r N/
p a reasonable nargin of error, UT of all bea:s during
- Y * *
)'.. a re fueling outage would reasonably assure that no .u... beam f ailure would occur during the following op-( n ua.neau w erating cycle. e e -,.- o - Jun 81 e + -, -, - e
- ?.6 m
z var. krom 1 to 8 in. si.d the i spat. .i M. e.. t o r 1..t. 8 and 3436 1M .ere 1.#, 195 ar isun. *
- t. 3:
indicati*- v ?' ?-P.fareed as cracks after hydro-Cr was in the process of develonwa a permanent fix to lasing aw. nre r. c u antng. An evaluation indicated [ , the jet pump holddown beam problem. probably involvang that the pars ' Wid* retain its integrity through-reduction of stress on the beam through redesign out the next cycle althouph CS flow distribution and modified heat treatment to resist ICSCC, or use could be af f eered due to through-wall cracks; of a dif ferent material. (hai) however, flos delivery to the shroud interior was not expected to cecrease. It was believed that a
- * * ' " 'a d "' '" *" * ' ' "' 12 * * * "
- 1 3 ' '
" ^ ' " ' ' " i=" "c * '*a "*"1d 6* i=P*** d d"'i"8 'h' N ***n"beams indicated that 1 beam was rejectable and next cycle to compensate for the assumption of no pur p a beams were in marginal condition. The cause of CS heat transler. Based on results free other sparger inspections and previous pipe cracking }st pump beam degradation was believed to be exPerlence eold work and sensitization during intergranular stress corrosion cracking. The jet j fabrication and installation stresses were considered pump beams 3, 4, 5, 6 and 7 were to be replaced to be the major factors.in causing the observed during the refueling outage then in progress: all were to be inspected and/or replaced during the cracks at Pilgri=. The cracks were hypothesised to next refueling outage. Rectre loop flows were to be initiated and propagated by ICSCC. be monitored during the next fuel cycle, as per recorranendations set forth in IE fiull 50 07. (1sk) Because the cause of cracking was not yet confirmed by metallurgical analysis, CE was developing tooling to extract sparger samples, and was evaluating methods of improving sparger inspection techniques. In addition, the NRC informed the BWR facilities to take the following actionas 1) At' the next scheduled and each following ref'ueling wtage un$ Mhr notice, perfum a doual b SOURCE H3LDERS BROKEM spection of the SC Sparsers and the segment oF P Ping between the inlet nosale and the vesse?' I Peach Botton 2 - 1977 - refueling shroud. Remote underwater TV exams were acceptable
- S"***"
In May, during the removal of source holders, the n u 0.M in. Ma h des was fourth holder to be removed came apart. The upper ce sidered an acceptable means of demonstrating ssetion was removed but the lower section remained e itable resolution of the TV examinations. Such in the reactor. A TV inspection of the remaining techgiques as the use of oblique lighting, and the portion, as well as the other 3 source holders a ty to t hos ead sW Wepe@ndy me revealed significant cracking of the SS sleeve
- """ E
- E' A special tool which gripped the holders from the bottom was ordered from the NSSS. The re= oval of the remaining source holders and the debris
- 2) If cracks were identified, the location and lef t above the fuel support pieces from the extent of the indications were to be reported.
broken holders was co=pleted in July. (hao) Supple =entary ext:s using volumetric methods could be performed to aid in characterizing the extent of cracking in non-visible locations. (hal) 65. CS SPARCER CRACKS IDE*;TIFIED EURs in General - 1980 Oyater Creek - 1978, Jan 80 Pilgric - Feb SO The NRC issued a Bulletin (No. 80-13) in May regarding core spray cracking reported by 2 utilities (19 of M *. 21 plants inspected observed no cracking). o.WY ~, C In 78, Oyster Creek identified a crack in the CS Sparger System II (see II.50). An evaluation postu-lated that deformation of the sparger had occurred during f abrication and installation which led to Y cracking by Intergranular Stress corrosion Cracking (ICSCC). A clamp asse=bly was installed over the crack and operation continued until refueling In Jan 80, the sparger was inspected and further cracking was discovered (see II.52). A total of 2$ cracks, 0.001 to 0.002 in. in width and of varying lengths, were identified in both CS spargers. It was believed g ,s that the cracks were undetected during the 1978 Inspection due to inspection equipment ll=itation. Nine additional clamp asse:blies were installed and were expected to maintain the sparrer physical in-tegrity. The ERC determined that the repsir measures were adequate until an improved replactn nt system ( ceuld be installed during the next refueling outage. An exam at Pilgrim revealed 5 indications in the upper CS sparger and 2 indications on the lovet 05 Jun.$1 s o=- -a N -